Toward a Reference Architecture for Intelligent Systems in Clinical Care

A Software Architecture for Precision Medicine

Intelligent systems in clinical care leverage the latest innovations in machine learning, real-time data stream mining, visual analytics, natural language processing, ontologies, production rule systems, and cloud computing to provide clinicians with the best knowledge and information at the point of care for effective clinical decision making. In this post, I propose a unified open reference architecture that combines all these technologies into a hybrid cognitive system for clinical decision support. Indeed, truly intelligent systems are capable of reasoning. The goal is not to replace clinicians, but instead to provide them with cognitive support during clinical decision making. Furthermore, Intelligent Personal Assistants (IPAs) such as Apple's Siri, Google's Google Now, and Microsoft's Cortana have raised our expectations on how intelligent systems interact with users through voice and natural language.

In the strict sense of the term, a reference architecture should be abstracted away from concrete technology implementation. However in order to enable a better understanding of the proposed approach, I take liberty in explaining how available open source software can be used to realize the intent of the architecture. There is an urgent need for an open and interoperable architecture which can be deployed across devices and platforms. Unfortunately, this is not the case today with solutions like Apple's HealthKit and ResearchKit.

The specific open source software mentioned in this post can be substituted with other tools which provide similar capabilities. The following diagram is a depiction of the architecture (click to enlarge).

 

Clinical Data Sources

Clinical data sources are represented on the left of the architecture diagram. Examples include electronic medical record systems (EMR) commonly used in routine clinical care, clinical genome databases, genome variant knowledge bases, medical imaging databases, data from medical devices and wearable sensors, and unstructured data sources such as biomedical literature databases. The approach implements the Lambda Architecture enabling both batch and real-time data stream processing and mining.

Predictive Modeling, Real-Time Data Stream Mining, and Big Data Genomics

The back-end provides various tools and frameworks for advanced analytics and decision management. The analytics workbench includes tools for creating predictive models and data streaming mining. The decision management workbench includes a production rule system (providing seamless integration with clinical events and processes) and an ontology editor.

The incoming clinical data likely meet the Big Data criteria of volume, velocity, and variety (this is particularly true for physiological time series from wearable sensors). Therefore, specialized frameworks for large scale cluster computing like Apache Spark are used to analyze and process the data. Statistical computing and Machine Learning tools like R are used here as well. The goal is knowledge and patterns discovery using Machine Learning model builders like Decision Trees, k-Means Clustering, Logistic Regression, Support Vector Machines (SVMs), Bayesian Networks, Neural Networks, and the more recent Deep Learning techniques. The latter hold great promise in applications such as Natural Language Processing (NLP), medical image analysis, and speech recognition.

These Machine Learning algorithms can support diagnosis, prognosis, simulation, anomaly detection, care alerting, and care planning. For example, anomaly detection can be performed at scale using the k-means clustering machine learning algorithm in Apache Spark. In addition, Apache Spark allows the implementation of the Lambda Architecture and can also be used for genome Big Data analysis at scale.

In another post titled How Good is Your Crystal Ball?: Utility, Methodology, and Validity of Clinical Prediction Models, I discuss quantitative measures of performance for clinical prediction models.

Visual Analytics

Visual Analytics tools like D3.js, rCharts, ploty, googleVis, ggplot2, and ggvis can help obtain deep insight for effective understanding, reasoning, and decision making through the visual exploration of massive, complex, and often ambiguous data. Of particular interest is Visual Analytics of real-time data streams like physiological time series. As a multidisciplinary field, Visual Analytics combines several disciplines such as human perception and cognition, interactive graphic design, statistical computing, data mining, spatio-temporal data analysis, and even Art. For example, similar to Minard's map of the Russian Campaign of 1812-1813 (see graphic below), Visual Analytics can help in comparing different interventions and care pathways and their respective clinical outcomes over a certain period of time by displaying causes, variables, comparisons, and explanations.

Production Rule System, Ontology Reasoning, and NLP

The architecture also includes a production rule engine and an ontology editor (Drools and Protégé respectively). This is done in order to leverage existing clinical domain knowledge available from clinical practice guidelines (CPGs) and biomedical ontologies like SNOMED CT.  This approach complements machine learning algorithms' probabilistic approach to clinical decision making under uncertainty. The production rule system can translate CPGs into executable rules which are fully integrated with clinical processes (workflows) and events. The ontologies can provide automated reasoning capabilities for decision support.

NLP includes capabilities such as:

• Text classification, text clustering, document and passage retrieval, text summarization, and more advanced clinical question answering (CQA) capabilities which can be useful for satisfying clinicians' information needs at the point of care; and
• Named entity recognition (NER) for extracting concepts from clinical notes.

The data tier supports the efficient storage of large amounts of time series data and is implemented with tools like Cassandra and HBase. The system can run in the cloud, for example using the Amazon Elastic Compute Cloud (EC2). For real-time processing of distributed data streams, cloud-based solutions like Amazon Kinesis and Lambda can be used.

 

Clinical Decision Services

The clinical decision services provide intelligence at the point of care typically using deployed predictive models, clinical rules, text mining outputs, and ontology reasoners. For example, Machine Learning algorithms can be exported in predictive markup language (PMML) format for run-time scoring based on the clinical data of individual patients, enabling what is referred to as Personalized Medicine. Clinical decision services include:

• Diagnosis and prognosis
• Simulation
• Anomaly detection 
• Data visualization
• Information retrieval (e.g., clinical question answering)
• Alerts and reminders
• Support for care planning processes.

The clinical decision services can be deployed in the cloud as well. Other clinical systems can consume these services through a SOAP or REST-based web service interface (using the HL7 vMR and DSS specifications for interoperability) and single sign-on (SSO) standards like SAML2 and OpenID Connect.

Intelligent Personal Assistants (IPAs)

Clinical decision services can also be delivered to patients and clinicians through IPAs. IPAs can accept inputs in the form of voice, images, and user's context and respond in natural language. IPAs are also expanding to wearable technologies such as smart watches and glasses. The precision of speech recognition, natural language processing, and computer vision is improving rapidly with the adoption of Deep Learning techniques and tools. Accelerated hardware technologies like GPUs and FPGAs are improving the performance and reducing the cost of deploying these systems at scale.

Hexagonal, Reactive, and Secure Architecture

Intelligent Health IT systems are not just capable of discovering knowledge and patterns in data. They are also scalable, resilient, responsive, and secure. To achieve these objectives, several architectural patterns have emerged during the last few years:

• Domain Driven Design (DDD) puts the emphasis on the core domain and domain logic and recommends a layered architecture (typically user interface, application, domain, and infrastructure) with each layer having well defined responsibilities and interfaces for interacting with other layers. Models exist within "bounded contexts". These "bounded contexts" communicate with each other typically through messaging and web services using HL7 standards for interoperability.


• The Hexagonal Architecture defines "ports and adapters" as a way to design, develop, and test an application in a way that is independent of the various clients, devices, transport protocols (HTTP, REST, SOAP, MQTT, etc.), and even databases that could be used to consume its services in the future. This is particularly important in the era of the Internet of Things in healthcare.


• Microservices consist in decomposing large monolithic applications into smaller services following good old principles of service-oriented design and single responsibility to achieve modularity, maintainability, scalability, and ease of deployment (for example, using Docker).


• CQRS/ES: Command Query Responsibility Segregation (CQRS) and Event Sourcing (ES) are two architectural patterns which consist in the use of event-driven messaging and an Event Store for separating commands (write-side) from queries (read-side) relying on the principle of Eventual Consistency. CQRS/ES can be implemented in combination with microservices to deliver new capabilities such as temporal queries, behavioral analysis, complex audit logs, and real-time notifications and alerts.


• Functional Programming: Functional Programming languages like Scala have several benefits that are particularly important for applying Machine Learning algorithms on large data sets. Like functions in mathematics, functions in Scala have no side effects. This provides referential transparency. Machine Learning algorithms are in fact based on Linear Algebra and Calculus. Scala supports high-order functions as well. Variables are immutable witch greatly simplifies concurrency. For all those reasons, Machine Learning libraries like Apache Mahout have embraced Scala, moving away from the Java MapReduce paradigm.


• Reactive Architecture: The Reactive Manifesto makes the case for a new breed of applications called "Reactive Applications". According to the manifesto, the Reactive Application architecture allows developers to build "systems that are event-driven, scalable, resilient, and responsive."  Leading frameworks that support Reactive Programming include Akka and RxJava. The latter is a library for composing asynchronous and event-based programs using observable sequences. RxJava is a Java port (with a Scala adaptor) of the original Rx (Reactive Extensions) for .NET created by Erik Meijer.
  
  Based on the Actor Model and built in Scala, Akka is a framework for building highly concurrent, asynchronous, distributed, and fault tolerant event-driven applications on the JVM. Akka offers location transparency, fault tolerance, asynchronous message passing, and a non-deterministic share-nothing architecture. Akka Cluster provides a fault-tolerant decentralized peer-to-peer based cluster membership service with no single point of failure or single point of bottleneck.
  
  Also built with Scala, Apache Kafka is a scalable message broker which provides high-throughput, fault-tolerance, built-in partitioning, and replication  for processing real-time data streams. In the reference architecture, the ingestion layer is implemented with Akka and Apache Kafka.

• Web Application Security: special attention is given to security across all layers, notably the proper implementation of authentication, authorization, encryption, and audit logging. The implementation of security is also driven by deep knowledge of application security patterns, threat modeling, and enforcing security best practices (e.g., OWASP Top Ten and CWE/SANS Top 25 Most Dangerous Software Errors) as part of the continuous delivery process.

An Interface that Works across Devices and Platforms

The front-end uses a Mobile First approach and a Single Page Application (SPA) architecture with Javascript-based frameworks like AngularJS to create very responsive user experiences. It also allows us to bring the following software engineering best practices to the front-end:

• Dependency Injection
• Test-Driven Development (Jasmine, Karma, PhantomJS)
• Package Management (Bower or npm)
• Build system and Continuous Integration (Grunt or Gulp.js)
• Static Code Analysis (JSLint and JSHint), and 
• End-to-End Testing (Protractor). 

For mobile devices, Apache Cordova can be used to access native functions when desired. The main goal is to provide a user interface that works across devices and platforms such as iOS, Android, and Windows Phone.

Interoperability

Interoperability will always be a key requirement in clinical systems. Interoperability is needed between all players in the healthcare ecosystem including providers, payers, labs, knowledge artifact developers, quality measure developers, and public health agencies like the CDC. These standards exist today and are implementation-ready. However, only health IT buyers have the leverage to demand interoperability from their vendors.

Standards related to clinical decision support (CDS) include:

• The HL7 Fast Healthcare Interoperability Resources (FHIR)
• The HL7 virtual Medical Record (vMR)
• The HL7 Decision Support Services (DSS) specification
• The HL7 CDS Knowledge Artifact specification
• The DMG Predictive Model Markup Language (PMML) specification.

Overcoming Barriers to Adoption

In a previous post, I discussed a practical approach to addressing challenges to the adoption of clinical decision support (CDS) systems.

Ontologies for Addiction and Mental Disease: Enabling Translational Research and Clinical Decision Support

In a previous post titled Why do we need ontologies in healthcare applications, I elaborated on what ontologies are and why they are different from information models of data structures like relational database schemas and XML schemas commonly used in healthcare informatics applications. In this post, I discuss two interesting applications of ontology engineering related to addiction and mental disease treatment. The first is the use of ontologies for achieving semantic interoperability in  translational research. The second is the use of ontologies for modeling complex medical knowledge in clinical practice guidelines (CPGs) for the purpose of automated reasoning during execution in clinical decision support systems (CDS) at the point of care.

Why Semantic Interoperability is needed in biomedical translational research?

In order to accelerate the discovery of new effective therapeutics for mental health and addiction treatment, there is a need to integrate data across disciplines spanning biomedical research and clinical care delivery [1]. For example, linking data across disciplines can facilitate a better understanding of treatment response variability among patients in addiction treatment. These disciplines include:

• Genetics, the study of genes.
• Chemistry, the study of chemical compounds including substances of abuse like heroin.
• Neuroscience, the study of the nervous system and the brain (addiction is a chronic disease of the brain)
• Psychiatry which is focused on the diagnosis, treatment, and prevention of addiction and mental disorders.

Each of these disciplines has its own terminology or controlled vocabularies. In the clinical domain for example, DSM5 and RrxNorm are used for documenting clinical care. In biomedical research, several ontologies have been developed over the last few years including:

• The Gene Ontology (GO)
• The Chemical Entities of Biological Interest Ontology (CHEBI)
• NeuroLex, an OWL ontology covering major domains of neuroscience: anatomy, cell, subcellular, molecule, function, and dysfunction.

To facilitate semantic interoperability between these ontologies, there are best practices established by the Open Biomedical Ontology (OBO) community. An example of best practice is the use of an upper-level ontology called the Basic Formal Ontology (BFO) which acts as a common foundational ontology upon which  new ontologies can be created. OBO ontologies and principles are available on the OBO Foundry web site.

Among the ontologies available on the OBO Foundry is the Mental Functioning Ontology (MF) [2, 3]. The MF is being developed as a collaboration between the University of Geneva in Switzerland and the University at Buffalo in the United States. The project also includes a Mental Disease Ontology (MD) which extends the MF and the Ontology for General Medical Science (OGMS). The Basic Formal Ontology (BFO) is an upper-level ontology for both the MF and the OGMS. The picture below is a view of the class hierarchy of the MD showing details of the class "Paranoid Schizophrenia" in the right pane of the windows of the beta release of Protege 5, an open source ontology development environment (click on the image to enlarge it).

The following is a tree view of the "Mental Disease Course" class (click on the image to enlarge it):

Ontology constructs defined by the OWL2 language can help establish common semantics (meaning) and relationships between entities across domains. These constructs provide automated inferencing capabilities such as equivalence (e.g., owl:sameAs and owl:equivalentClass) and subsumption (e.g., rdfs:subClassOf) relationships between entities.

In addition, publishing data sources following Linked Open Data (LOD) principles and semantic search using federated SPARQL queries can help answer new research questions. Another application is semantic annotation for natural language processing (NLP) applications.

 

Ontologies as knowledge representation formalism for clinical decision support (CDS)

As knowledge representation formalism, ontologies are well suited for modeling complex medical knowledge and can facilitate reasoning during the automated execution of clinical practice guidelines (CPGs) and Care Pathways (CPs) based on patient data at the point of care. Several approaches to modelling CPGs and CPs have been proposed in the past including PROforma, HELEN, EON, GLIF, PRODIGY, and SAGE. However, the lack of free and open source tooling has been a major impediment to a wide adoption of these knowledge representation formalisms. OWL has the advantage of being a widely implemented W3C Recommendation with available mature open source  tools.

In practice, the medical knowledge contained in CPGs can be manually translated into IF-THEN statements in most programming languages. Executable CDS rules (like other complex types of business rules) can be implemented with a production rule engine using forward chaining. This is the approach taken by OpenCDS and some large scale CDS implementations in real world healthcare delivery settings. This allows CDS software developers to externalize the medical knowledge contained in clinical guidelines in the form of declarative rules as opposed to embedding that knowledge in procedural code. Many viable open source business rule management systems (BRMS) are available today and provide capabilities such as a rule authoring user interface, a rules repository, and a testing environment.

However, production rule systems have a limitation. They do not scale because they require writing a rule for each clinical concept code (there are more than 311,000 active concepts in SNOMED CT alone). An alternative is to exploit the class hierarchy in an ontology so that subclasses of a given superclass can inherit the clinical rules that are applicable to the superclass (this is called subsumption). In addition to subsumption, an OWL ontology also support reasoning with description logic (DL) axioms [4].

An ontology designed for a clinical decision support (CDS) system can integrate the clinical rules from a CPG, a domain ontology like the Mental Disorder (MD) ontology, and the patient medical record from an EHR database in order to provide inferences in the form of treatment recommendations at the point of care. The OWL API [5] facilitates the integration of ontologies into software applications. It supports inferencing using reasoners like Pellet and HermiT. OWL2 reasoning capabilites can be enhanced with rules represented in SWRL (Semantic Web Rule Language) which is implemented by reasoners like Pellet as well as the Protege OWL development environement. In addition to inferencing, another benefit of an OWL-based approach is transparency: the CDS system can provide an explanation or justification of how it arrives at the treatment recommendations.

Nonetheless, these approaches are not mutually exclusive: a production rule system can be integrated with business processes, ontologies, and predictive analytics models. Predictive analytics models provide a probabilistic approach to treatment recommendations to assist in the clinical decision making process.

References

[1]  Janna Hastings, Werner Ceusters, Mark Jensen, Kevin Mulligan and Barry Smith. Representing mental functioning: Ontologies for mental health and disease. Proceedings of the Mental Functioning Ontologies workshop of ICBO 2012, Graz, Austria.

[2]  Ceusters, W. and Smith, B. (2010a). Foundations for a realist ontology of mental disease. Journal of Biomedical Semantics, 1(1), 10.

[3] Hastings, J., Smith, B., Ceusters, W., and Mulligan, K. (2012). The mental functioning ontology. http://code.google.com/p/mental-functioning-ontology/, last accessed August 24, 2014

[4] Sesen MB, Peake MD, Banares-Alcantara R, Tse D, Kadir T, Stanley R, Gleeson F, Brady M. 2014 Lung Cancer Assistant: a hybrid clinical decision support application for lung cancer care. J. R. Soc. Interface 11: 20140534.

[5] Matthew Horridge, Sean Bechhofer. The OWL API: A Java API for OWL Ontologies Semantic Web Journal 2(1), Special Issue on Semantic Web Tools and Systems, pp. 11-21, 2011.

Natural Language Processing (NLP) for Clinical Decision Support: A Practical Approach

A significant portion of the electronic documentation of clinical care is captured in the form of unstructured narrative text like psychotherapy and progress notes. Despite the big push to adopt structured data entry (as required by the Meaningful Use incentive program for example), many clinicians still like to document care using free narrative text. The advantage of using narrative text as opposed to coded entries is that narrative text can tell the story of the patient and the care provided particularly in complex cases. My opinion is that free narrative text should be used to complement coded entries when necessary to capture relevant information.

Furthermore, medical knowledge is expanding very rapidly. For example, PubMed has more than 24 millions citations for biomedical literature from MEDLINE, life science journals, and online books. It is impossible for the human brain to keep up with that amount of knowledge. These unstructured sources of knowledge contain the scientific evidence that is required for effective clinical decision making in what is referred to as Evidence-Based Medicine (EBM).

In this blog, I discuss two practical applications of Natural Language Processing (NLP). The first is the use of NLP tools and techniques to automatically extract clinical concepts and other insight from clinical notes for the purpose of providing treatment recommendations in Clinical Decision Support (CDS) systems. The second is the use of text analytics techniques like clustering and summarization for Clinical Question Answering (CQA).

The emphasis of this post is on a practical approach using freely available and mature open source tools as opposed to an academic or theoretical approach. For a theoretical treatment of the subject, please refer to the book Speech and Language Processing by Daniel Jurafsky and James Martin.

Clinical NLP with Apache cTAKES

Based on the Apache Unstructured Information Management Architecture (UIMA) framework and the Apache OpenNLP natural language processing toolkit, Apache cTAKES provides a modular architecture utilizing both rule-based and machine learning techniques for information extraction from clinical notes. cTAKES can extract named entities (clinical concepts) from clinical notes in plain text or HL7 CDA format and map these entities to various dictionaries including the following Unified Medical Language System (UMLS) semantic types: diseases/disorders, signs/symptoms, anatomical sites, procedures, and medications.

cTAKES includes the following key components which can be assembled to create processing pipelines:

• Sentence boundary detector based on the OpenNLP Maximum Entropy (ME) sentence detector.
• Tokenizor
• Normalizer using the National Library of Medicine's Lexical Variant Generation (LVG) tool
• Part-of-speech (POS) tagger
• Shallow parser
• Named Entity Recognition (NER) annotator using dictionary look-up to UMLS concepts and semantic types. The Drug NER can extract drug entities and their attributes such as dosage, strength, route, etc.
• Assertion module which determines the subject of the statement (e.g., is the subject of the statement the patient or a parent of the patient) and whether a named entity or event is negated (e.g., does the presence of the word "depression" in the text implies that the patient has depression).

Apache cTAKES 3.2 has added YTEX, a set of extensions developed at Yale University which provide integration with MetaMap, semantic similarity, export to Machine Learning packages like Weka and R, and feature engineering.

The following diagram from the Apache cTAKES Wiki provides an overview of these components and their dependencies (click to enlarge):

Source: Apache cTAKES wiki

Massively Parallel Clinical Text Analytics in the Cloud with GATECloud

The General Architecture for Text Engineering (GATE) is a mature, comprehensive, and open source text analytics platform. GATE is a family of tools which includes:

• GATE Developer: an integrated development environment (IDE) for language processing components with a comprehensive set of available plugins called CREOLE (Collection of REusable Objects for Language Engineering). 
• GATE Embedded: an object library for embedding services developed with GATE Developer into third-party applications.
• GATE Teamware: a collaborative semantic annotation environment based on a workflow engine for creating manually annotated corpora for applying machine learning algorithms. 
• GATE Mímir: the "Multi-paradigm Information Management Index and Repository" which supports a multi-paradigm approach to index and search over text, ontologies, and semantic metadata.
• GATE Cloud: a massively parallel clinical text analytics platform (Platform as a Service or PaaS) built on the Amazon AWS Cloud.

What makes GATE particularly attractive is the recent addition of GATECloud.net PaaS which can boost the productivity of people involved in large scale text analytics tasks.

 

Clustering, Classification, Text Summarization, and Clinical Question Answering (CQA)

 

An unsupervised machine learning approach called Clustering can be used to classify large volumes of medical literature into groups (clusters) based on some similarity measure (such as the Euclidean distance). Clustering can be applied at the document, search result, and word/topic levels. Carrot2 and Apache Mahout are open source projects that provide several methods for document clustering. For example, the Latent Dirichlet Allocation learning algorithm in Apache Mahout automatically clusters words into topics and documents into mixtures of topics. Other clustering algorithms in Apache Mahout include: Canopy, Mean-Shift, Spectral, K-Means and Fuzzy K-Means. Apache Mahout is part of the Hadoop ecosystem and can therefore scale to very large volumes of unstructured text.

Document classification essentially consists in assigning predefined set of labels to documents. This can be achieved through supervised machine learning algorithms. Apache Mahout implements the Naive Bayes classifier.

Text summarization techniques can be used to present succinct and clinically relevant evidence to clinicians at the point of care. MEAD (http://www.summarization.com/mead/) is an open source project that implements multiple summarization algorithms. In the biomedical domain, SemRep is a program that extracts semantic predications (subject-relation-object triples) from biomedical free text. Subject and object arguments of each predication are concepts from the UMLS Metathesaurus and the relation is from the UMLS Semantic Network (e.g., TREATS, Co-OCCURS_WITH). The SemRep summarization provides a short summary of these concepts and their semantic relations.

AskHermes (Help clinicians to Extract and aRrticulate Multimedia information for answering clinical quEstionS) is a project that attempts to implement these techniques in the clinical domain. It allows clinicians to enter questions in natural language and uses the following unstructured information sources: MEDLINE abstracts, PubMed Central full-text articles, eMedicine documents, clinical guidelines, and Wikipedia articles.

The processing pipeline in AskHermes includes the following: Question Analysis, Related Questions Extraction, Information Retrieval, Summarization and Answer Presentation. AskHermes performs question classification using MMTx (MetaMap Technology Transfer) to map keywords to UMLS concepts and semantic types. Classification is achieved through supervised machine learning algorithms such as Support Vector Machine (SVM) and conditional random fields (CFRs). Summarization and answer presentation are based on clustering techniques. AskHermes is powered by open source components including: JBoss Seam, Weka, Mallet , Carrot2 , Lucene/Solr, and WordNet (a lexical database for the English language).

Enabling Scalable Realtime Healthcare Analytics with Apache Spark

Modern and massively parallel computing platforms can process humongous amounts of data in real time to obtain actionable insights for effective clinical decision making. In this blog, I discuss an emerging Big Data platform called Apache Spark and its application to remote real-time healthcare monitoring using data from medical devices and wearable sensors. The goal is to provide effective remote care for an increasingly aging population as well as public health surveillance.

The Apache Spark Framework

Apache Spark has emerged during the last couple of years as an innovative platform for Big Data and in-memory cluster computing capable of running programs up to 100x faster than traditional Hadoop MapReduce. Apache Spark is written in Scala, a functional programming language (see my previous post titled Navigating in Scala land). Spark also offers a Java and a Python APIs. The Scala API allows developers to interact with Spark by using very concise and expressive Scala code.

The Spark stack also includes the following integrated tools:

• Spark SQL which allows relational queries expressed in SQL, HiveQL, or Scala to be executed using Spark through a data abstraction called SchemaRDD. Supported data sources include Parquet files (a columnar storage format for Hadoop), JSON datasets, or data stored in Apache Hive.


• Spark Streaming which enables fault-tolerant stream processing of live data streams. Data can be ingested from many sources like Kafka, Flume, Twitter, ZeroMQ or plain old TCP sockets. The ingested data can be directly processed with Spark built-in Machine Learning algorithms.


• MLlib (Machine Learning Library) provides a library of practical Machine Learning algorithms including support vector machines (SVM), logistic regression, decision trees, naive Bayes, and k-means clustering.


• GraphX which provides graph-parallel computation for graph-analytics application like social networks.

Apache Spark can also play nicely with other frameworks within the Hadoop ecosystem. For example, it can run standalone or on a Hadoop 2's YARN cluster manager, on Amazon EC2 or a Mesos cluster manager. Spark can also read data from HFDS, HBase, Cassandra or any other Hadoop data source. Other noteworthy integrations include:

• SparkR, an R package allowing the use of Spark from R, a very popular open source software environment for statistical computing with more that 5800 packages including Machine Learning packages; and


• H2O-Sparkling which provides an integration with the H2O platform through in-memory sharing with Tachyon, a memory-centric distributed file system for data sharing across cluster frameworks. This allows Spark applications to leverage advanced distributed Machine Learning algorithms supported by the H2O platform like emerging Deep Learning algorithms.

 

Wearable Sensors for Remote Healthcare Monitoring 

Three factors are contributing to the availability of massive amounts of clinical data: the rising adoption of EHRs by providers thanks in part to the Meaningful Use incentive program; the increasing use of medical devices including wearable sensors used by patients outside of healthcare facilities; and medical knowledge (for example in the form of medical research literature).

One promising area in Healthcare Informatics where Big Data architectures like the one provided by Apache Spark can make a difference is in applications using data from wearable health monitoring sensors for anomaly detection, care alerting, diagnosis, care planning, and prediction. For example, anomaly detection can be performed at scale using the k-means clustering machine learning algorithm in Spark.

These sensors and devices are part of a larger trend called the "Internet of Things". They enable new capabilities such as remote health monitoring for personalized medicine and chronic care management for an increasingly aging population as well as public health surveillance for outbreaks and epidemics.

Wearable sensors can collect vital signs data like weight, temperature, blood pressure (BP), heart rate (HR), blood glucose (BG), respiratory rate (RR), electrocardiogram (ECG), oxygen saturation (SpO2), and Photoplethysmography (PPG). Spark Streaming can be used to perform real-time stream processing on sensors data and the data can be processed and analyzed using the Machine Learning algorithms available in MLlib and the other integrated frameworks like R and H2O. What makes Spark particularly suitable for this type of applications is that sensor data meet the Big Data criteria of volume, velocity, and variety.

Researchers predict that internet use on mobile phones will increase 20-fold in Africa in the next five years. The number of mobile subscriptions in sub-Saharan Africa is expected to reach 635 millions by the end of this year. This unprecedented level of connectivity (fueled in part by the historical lack of land line infrastructure) provides opportunities for effective public health surveillance and disease management in the developing world.

Apache Spark is the type of open source computing infrastructure that is needed for distributed, scalable, and real-time healthcare analytics for reducing healthcare costs and improving outcomes.

Improving the quality of mental health and substance use treatment: how can Informatics help?

According to the 2012 National Survey on Drug Use and Health, an estimated 43.7 million adults aged 18 or older in the United States had mental illness in the past year. This represents 18.6 percent of all adults in this country. Among those 43.7 million adults, 19.2 percent (8.4 million adults) met criteria for a substance use disorder (i.e., illicit drug or alcohol dependence or abuse). In 2012, an estimated 9.0 million adults (3.9 percent) aged 18 or older had serious thoughts of suicide in the past year.

Mental health and substance use are often associated with other issues such as:

• Co-morbidity involving other chronic diseases like HIV, hepatitis, diabetes, and cardiovascular disease.


• Overdose and emergency care utilization.


• Social issues like incarceration, violence, homelessness, and unemployment.

It is now well established that addiction is a chronic disease of the brain and should be treated as such from a health and social policy standpoint.

The regulatory framework

• The Affordable Care Act (ACA) requires non-grandfathered health plans in the individual and small group markets to provide essential health benefits (EHBs) including mental health and substance use disorder benefits.  


• Starting in 2014, insurers can no longer deny coverage because of a pre-existing mental health condition.


• The ACA requires health plans to cover recommended evidence-based prevention and screening services including depression screening for adults and adolescents and behavioral assessments for children.


• On November 8, 2013, HHS and the Departments of Labor and Treasury released the final rules implementing the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act of 2008 (MHPAEA). 


• Not all behavioral health specialists are eligible to the Meaningful Use EHR Incentive program created by the Health Information Technology for Economic and Clinical Health Act (HITECH) of 2009.

 

Implementing Clinical Practice Guidelines (CPGs) with Clinical Decision Support (CDS) systems

 

Clinical Decision Support (CDS) can help address key challenges in mental health and substance use treatment such as:

• Shortages and high turnover in the addiction treatment workforce.


• Insufficient or lack of adequate clinician education in mental health and addiction medicine.


• Lack of implementation of available evidence-based clinical practice guideline (CPGs) in mental health and addiction medicine.

For example, there are a number of scientifically validated CPGs for the Medication Assisted Treatment (MAT) of opioid addiction using methadone or buprenorphine. These evidence-based CPGs can be translated into executable CDS rules using business rule engines. These executable clinical rules should also be seamlessly integrated with clinical workflows.

The complexity and costs inherent in capturing the medical knowledge in clinical guidelines and translating that knowledge into executable code remains an impediment to the widespread adoption of CDS software. Therefore, there is a need for standards that facilitate the sharing and interchange of CDS knowledge artifacts and executable clinical guidelines. The ONC Health eDecision Initiative has published specifications to support the interoperability of CDS knowledge artifacts and services.

Ontologies as knowledge representation formalism are well suited for modeling complex medical knowledge and can facilitate reasoning during the automated execution of clinical guidelines based on patient data at the point of care.

The typical Clinical Practice Guideline (CPG) is 50 to 150 pages long. Clinical Decision Support (CDS) should also include other forms of cognitive aid such as Electronic Checklists, Data Visualization, Order Sets, and Infobuttons.

The issues of human factors and usability of CDS systems as well as CDS integration with clinical workflows have been the subject of many research projects in healthcare informatics. The challenge is to be bring these research findings into the practice of developing clinical systems software.

Learning from Data

Learning what works and what does not work in clinical practice is important for building a learning health system. This can be achieved by incorporating the results of Comparative Effectiveness Research (CER) and Patient-Centered Outcome Research (PCOR) into CDS systems. Increasingly, outcomes research will be performed using observational studies (based on real world clinical data) which are recognized as complementary to randomized control trials (RCTs). For example, CER and PCOR can help answer questions about the comparative effectiveness of pharmacological and  psychotherapeutic interventions in mental health and substance abuse treatment. This is a form of Practice-Based Evidence (PBE) that is necessary to close the evidence loop.

Three factors are contributing to the availability of massive amounts of clinical data: the rising adoption of EHRs by providers (thanks in part to the Meaningful Use incentive program), medical devices (including those used by patients outside of healthcare facilities), and medical knowledge (for example in the form of medical research literature). Massively parallel  computing platforms such as Apache Hadoop or Apache Spark can process humongous amounts of data (including in real time) to obtain actionable insights for effective clinical decision making.

The use of predictive modeling for personalized medicine (based on statistical computing and machine learning techniques) is becoming a common practice in healthcare delivery as well. These models can predict the health risk of patients (for pro-active care) based on their individual health profiles and can also help predict which treatments are more likely to lead to positive outcomes.

Embedding Visual Analytics capabilities into CDS systems can help clinicians obtain deep insight for effective understanding, reasoning, and decision making through the visual exploration of massive, complex, and often ambiguous data. For example, Visual Analytics can help in comparing different interventions and care pathways and their respective clinical outcomes for a patient or population of patients over a certain period of time through the vivid showing of causes, variables, comparisons, and explanations.

Genomics of Addiction and Personalized Medicine

Advances in genomics and pharmacogenomics are helping researchers understand treatment response variability among patients in addiction treatment. Clinical Decision Support (CDS) systems can also be used to provide cognitive support to clinicians in providing genetically guided treatment interventions.

Quality Measurement for Mental Health and Substance Use Treatment

An important implication of the shift from a fee-for-service to a value-based healthcare delivery model is that existing process measures and the regulatory requirements to report them are no longer sufficient.

Patient-reported outcomes (PROs) and patient-centered measures include essential metrics such as mortality, functional status, time to recovery, severity of side effects, and remission (depression remission at six and twelve months). These measures should take into account the values, goals, and wishes of the patient. Therefore patient-centered outcomes should also include the patient's own evaluation of the care received.

Another issue to be addressed is the lack of data elements in Electronic Medical Record (EMR) systems for capturing, reporting, and analyzing PROs. This is the key to accountability and quality improvement in mental health and substance use treatment.

Using Natural Language Processing (NLP) for the automated processing of clinical narratives

Electronic documentation in mental health and substance use treatment is often captured in the form of narrative text such as psychotherapy notes. Natural Language Processing (NLP) and machine learning tools and techniques (such as named entity recognition) can be used to extract clinical concepts and other insight from clinical notes.

Another area of interest is Clinical Question Answering (CQA) that would allow clinicians to ask questions in natural language and extract clinical answers from very large amounts of unstructured sources of medical knowledge. PubMed has more than 23 millions citations for biomedical literature from MEDLINE, life science journals, and online books. It is impossible for the human brain to keep up with that amount of knowledge.

Computer-Based Cognitive Behavioral Therapy (CCBT) and mHealth

According to a report published last year by the California HealthCare Foundation and titled The Online Couch: Mental Health Care on the Web:

"Computer-based cognitive behavioral therapy (CCBT) cost-effectively leverages the Internet for coaching patterns in self-driven or provider-assisted programs. Technological advances have enabled computer systems designed to replicate aspects of cognitive behavior therapy for a growing range of mental health issues".

An example of a successful nationwide adoption of CCBT is the online behavioral therapy site Beating the Blues in the United Kingdom which has been proven to help patients suffering from anxiety and mild to moderate depression. Beating the Blues has been recommended for use in the NHS by the National Institute for Health and Clinical Excellence (NICE).

In addition, there is growing evidence to support the efficacy of mobile health (mHealth) technologies for supporting patent engagement and activation in health behavior change (e.g., smoking cessation).

 

Technologies in support of a Collaborative Care Model

There is sufficient evidence to support the efficacy of the collaborative care model (CCM) in the treatment of chronic mental health and substance use conditions.The CCM is based on the following principles:

• Coordinated care involving a multi-disciplinary care team.


• Longitudinal care plan as the backbone of care coordination.


• Co-location of primary care and mental health and substance use specialists.


• Case management by a Care Manager. 

Implementing an effective collaborative care model will require a new breed of advanced clinical collaboration tools and capabilities such as:

• Conversations and knowledge sharing using tools like video conferencing for virtual two-way face-to-face communication between clinicians (see my previous post titled Health IT Innovations for Care Coordination).


• Clinical content management and case management tools.


• File sharing and syncing allowing the longitudinal care plan to be synchronized and shared among all members of the care team.


• Light-weight and simple clinical data exchange standards and protocols for content, transport, security, and privacy. 

 

Patient Consent and Privacy

Because of the stigma associated with mental health and substance use, it is important to give patients control over the sharing of their medical record. Patients consent should be obtained about what type information is shared, with whom, and for what purpose. The patient should also have access to an audit trail of all data exchange-related events. Current paper-based consent processes are inefficient and lack accountability. Web-based consent management applications facilitate the capture and automated enforcement of patient consent directives (see my previous post titled Patient privacy at web scale).

Addressing Challenges to the Adoption of Clinical Decision Support (CDS) Systems: A Practical Approach

Laptop and stethoscope by jfcherry is licensed under CC BY-SA 2.0

This post has been updated on February 15, 2015.

Despite its potential to improve the quality of care, CDS is not widely used in health care delivery today. In technology marketing parlance, CDS has not crossed the chasm. There are several issues that need to be addressed including:

• Clinicians' acceptance of the concept of automated execution of evidence-based clinical practice guidelines.


• Integration into clinical workflows and care protocols.


• Usability and human factors issues including alert fatigue and the human factors that influence alert acceptance.


• With the expanding use of clinical prediction models for diagnosis and prognosis, there is a need for a better understanding of the probabilistic approach to clinical decision making under uncertainty.


• Standardization to enable the interoperability and reuse of CDS knowledge artifacts and executable clinical guidelines.


• The challenges associated with the automatic concurrent execution of multiple clinical practice guidelines for patients with co-morbidities.


• Integration with modern information retrieval capabilities which allow clinicians to access up-to-date biomedical literature. These capabilities includes text mining, Natural Language Processing (NLP), and more advanced Clinical Question Answering (CQA) tools.  CQA allows clinicians to ask clinical questions in natural language and extracts answers from very large amounts of unstructured sources of medical knowledge. PubMed has more than 23 millions citations for biomedical literature from MEDLINE, life science journals, and online books. The typical Clinical Practice Guideline (CPG) is 50 to 150 pages long. It is impossible for the human brain to keep up with that amount of knowledge.


• The use of mathematical simulations in CDS to explore and compare the outcomes of various treatment alternatives.


• Integration of genomics to enable personalized medicine as the cost of whole-genome sequencing (WGS) continues to fall.
  


• Integration of outcomes research in the context of a shift to a value-based healthcare delivery model. This can be achieved by incorporating the results of Comparative Effectiveness Research (CER) and Patient-Centered Outcome Research (PCOR) into CDS systems. Increasingly, outcomes research will be performed using observational studies (based on real world clinical data) which are recognized as complementary to randomized control trials (RCTs) for discovering what works and what doesn't work in practice. This is a form of Practice-Based Evidence (PBE) that is necessary to close the evidence loop.


• Support for a shared decision making process which takes into account the values, goals, and wishes of the patient.


• The use of Visual Analytics in CDS to facilitate analytical reasoning over very large amounts of structured and unstructured data sources.


• Finally, the challenges associated with developing hybrid decision support systems which seamlessly integrate all the technologies mentioned above including: machine learning predictive algorithms, real-time data stream mining, visual analytics, ontology reasoning, and text mining.

In response to a paper titled Grand challenges in clinical decision support by Sittig et al. [1], Fox et al. [2] outlined four theoretical foundations for the design and implementation of CDS systems: decision theory, theories of knowledge representation, process design, and organizational modeling. The practical approach discussed in this post is grounded in those four theoretical foundations.

CDS Interoperability

The complexity and cost inherent in capturing the medical knowledge in clinical guidelines and translating that knowledge into executable code remain a major impediment to the widespread adoption of CDS software. Therefore, there is a need for standards for the interchange and reuse of CDS knowledge artifacts and executable clinical guidelines.

Different formalisms, methodologies, and architectures have been proposed over the years for representing the medical knowledge in clinical guidelines. Examples include but are not limited to the following:

• The Arden Syntax
• GLIF (Guideline Interchange Format)
• GELLO (Guideline Expression Language Object-Oriented)
• GEM (Guidelines Element Model)
• The Web Ontology Language (OWL)
• PROforma
• EON
• PRODIGY
• Asbru
• GUIDE
• SAGE.

More recently, HL7 has published the Clinical Decision Support (CDS) Knowledge Artifact Specification which provides guidance on shareable CDS knowledge artifacts including event-condition-action rules, order sets, and documentation templates.

The HL7 Context-Aware Knowledge Retrieval (Infobutton) specifications provide a standard mechanism for clinical information systems to request context-specific clinical knowledge to satisfy clinicians and patients' information needs at the point of care.

Enabling the interoperability of executable clinical guidelines requires a standardized domain model for representing the medical information of patients and other contextual clinical information. The HL7 Virtual Medical Record (vMR) is a standardized domain model for representing the inputs and outputs of CDS systems. The ability to transform an HL7 CCDA document into an HL7 vMR document means that EHR systems that are Meaningful Use Stage 2 certified can consume these standard-compliant decision support services.

Because of the complexity and cost of developing CDS software, CDS software capabilities can be exposed as a set of services (part of a service-oriented architecture [16]) which can be consumed by other clinical systems such as EHR and Computerized Physician Order Entry (CPOE) systems. When deployed in the cloud, these CDS software services can be shared by several health care providers to reduce costs. The HL7 Decision Support Service (DSS) specification defines a REST and a SOAP web service interfaces using the vMR as message payload for accessing interoperable decision support services.

In practice, executable CDS rules (like other complex types of business rules) can be implemented with a production rule system using forward chaining. This is the approach taken by OpenCDS and some other large scale CDS implementations in real-world healthcare delivery settings. This allows CDS software developers to externalize the medical knowledge contained in clinical practice guidelines in the form of declarative rules as opposed to embedding that knowledge in procedural code. Many viable open source business rule management systems (BRMS) are available today and provide capabilities such as a rule authoring user interface, a rules repository, and a testing environment. Furthermore, a rule execution environment can be integrated with business processes, ontologies, and predictive analytics models (more on that later).

The W3C Rule Interchange Format (RIF) specification is a possible solution to the interchange of executable CDS rules. The RIF Production Rule Dialect (RIF PRD) is designed as a common XML serialization syntax for multiple rule languages to enable rule interchange between different BRMS. For example, RIF-PRD would allow the exchange of executable rules between existing BRMS like JBoss Drools, IBM ILOG JRules, and Jess. RIF is currently a W3C Recommendation and is backed by several BRMS vendors. Consentino et al. [3] described a model-driven approach for interoperability between IBM's proprietary ILOG rule language and RIF.

Seamless Integration into Clinical Workflows and Care Pathways

One of the main complaints against CDS systems is that they are not well integrated into clinical workflows and care protocols. Existing business process management standards like the Business Process Modeling Notation (BPMN) can provide a proven, practical, and adaptable approach to the integration of CDS rules and clinical pathways and protocols. Some existing open source and commercial BRMS already provide an integration of business rules and business processes out-of-the box and there are well-known patterns for integrating rules and processes [4, 5, 6] in business applications.

In 2014, the Object Management Group (OMG) released the Decision Model and Notation (DMN) specification which defines various constructs for modeling decision logic. The combination of BPMN and DMN [7, 8] provides a practical approach for modeling the decisions in clinical practice guidelines while integrating these decisions with clinical workflows. BPMN and DMN also support the modeling of decisions and processes that span functional and organizational boundaries.

Human Factors in the Use of Clinical Decision Support Systems

We need to do a better job in understanding the human factors that influence alert acceptance by clinicians in CDS. We also need clear and proven usability guidelines (backed by scientific research) that can be implemented by CDS software developers. Several research projects have sought to understand why clinicians accept or ignore alerts in medication-related CDS [9, 10]. Zacharia et al. [11] developed and validated an Instrument for Evaluating Human-Factors Principles in Medication-Related Decision Support Alerts (I-MeDeSA). I-MeDeSA measures CDS alerts on the following nine human factors principles: alarm philosophy, placement, visibility, prioritization, color learnability and confusability, text-based information, proximity of task components being displayed, and corrective actions.

The British National Health Service (NHS) Common User Interface (CUI) Program has created standards and guidance in support of the usability of clinical applications with inputs from user interface design specialists, usability experts, and hundreds of clinicians with a diversity of background in using health information technology. The program is based on a rigorous development process which includes: research, design, prototyping, review, usability testing, and patient safety assessment by clinicians. In the US, the National Institute of Standards and Technology (NIST) has also published some guidance on the usability of clinical applications.

Studies have also shown that like in aviation, checklists can provide cognitive support to clinicians in the decision making process.

Integrating Genomic Data with CDS

The costs of whole-genome sequencing (WGS) and whole-exome sequencing (WES) continue to fall. Increasingly, both WGS and WES will be used in clinical practice for inherited disease risk assessment and pharmacogenomic findings [21]. There is a need for a modern CDS architecture that can support and facilitate the introduction and use of WGS and WES in clinical practice.

In a paper titled Technical desiderata for the integration of genomic data with clinical decision support [14], Welch et al. proposed technical requirements for the integration of genomic data with clinical decision support.  In another paper titled A proposed clinical decision support architecture capable of supporting whole genome sequence information [15], Welch et al. proposed the following clinical decision support architecture for supporting whole genome sequence information (click to enlarge):

Proposed service-oriented architecture (SOA) for whole genome sequence (WGS)-enabled CDS by Brandon M. Welch, Salvador Rodriguez Loya, Karen Eilbeck, and Kensaku Kawamoto is licensed under CC BY 3.0

The proposed architecture includes the following components: the genome variant knowledge base, the genome database, the CDS knowledge base, a CDS controller and the electronic health record (EHR). The authors suggest separating the genome data from the EHR data.

Practice-Based Evidence (PBE) needed for closing the evidence loop

Prospective randomized controlled trials (RCTs) are still considered the gold standard in evidence-based medicine. Although they can control for biases, RCTs are costly, time consuming, and must be performed under carefully controlled conditions.

The retrospective analysis of existing clinical data sources is increasingly recognized as complementary to RCTs for discovering what works and what doesn't work in real world clinical practice [23]. These retrospective studies will allow the creation of clinical prediction models which can provide personalized absolute risk and treatment outcome predictions for patients. They also facilitate what has been referred to as "rapid learning health care." [24]

Toward Data-Driven Clinical Decision Support (CDS)

Williams Osler (1849-1919) [19] famously said that "Medicine is a science of uncertainty and an art of probability."

The use of clinical prediction models for diagnosis and prognosis is becoming common practice in clinical care. These models can predict the health risks of patients based on their individual health data. Clinical Prediction Models provide absolute risk and treatment outcome prediction for conditions such as diabetes, kidney disease, cancer, cardiovascular disease, and depression.  These models are built with statistical learning techniques and introduce new challenges related to their probabilistic approach to clinical decision making under uncertainty [20]. In his book titled Super Crunchers: Why Thinking-By-Numbers Is The New Way To be Smart, Ian Ayres wrote:

Traditional experts make better decisions when they are provided with the results of statistical prediction. Those who cling to the authority of traditional experts tend to embrace the idea of combining the two forms of knowledge by giving the experts 'statistical support'. The purveyors of diagnostic software are careful to underscore that its purpose is only to provide support and suggestions. They want the ultimate decision and discretion to lie with the doctor. [12]

Furthermore, in order to leverage existing clinical domain knowledge from clinical practice guidelines and biomedical ontologies [22], machine learning algorithms' probabilistic approach to decision making under uncertainty must be complemented by technologies like production rule systems and ontology reasoners. Sesen et al. [18] designed a hybrid CDS for lung cancer care based on probabilistic reasoning with a Bayesian Network model and guideline-based recommendations using a domain ontology and an ontology reasoner.

Fox et al. [2] proposed an argumentation approach based on the construction, summarization, and prioritization of arguments for and against each generated candidate decision. These arguments can be both qualitative or quantitative in nature. On the importance of presenting evidence-based rationale in CDS systems, Fox et al. wrote:

In short, to improve usability of clinical user interfaces we advocate basing design around a firm theoretical understanding of the clinician’s perspective on the medical logic in a decision, the qualitative as well as quantitative aspects of the decision, and providing an evidence-based rationale for all recommendations offered by a CDS. [2]

In a paper titled A canonical theory of dynamic decision-making [13], Fox et al. proposed a canonical theory of dynamic decision-making and presented the PROforma clinical guideline modeling language as an instance of the canonical theory.

Clinical predictive model presentation techniques include traditional score charts, nomograms, and clinical rules [17]. However Clinical Prediction Models are easier to use and maintain when deployed as scoring services (part of a service-oriented software architecture) and integrated into CDS systems. The scoring service can be deployed in the cloud to allow integration with multiple client clinical systems [20]. The Predictive Model Markup Language (PMML) specification published by the Data Mining Group (DMG) supports the interoperable deployment of predictive models in heterogeneous software environments.

Visual Analytics or data visualization techniques can also play an important role in the effective presentation of Clinical Prediction Models to nonstatisticians particularly in the context of shared decision making.

Concurrent execution of multiple guidelines for patients with co-morbidities

According to the Medicare 2012 chart book, "over two-thirds of beneficiaries having two or more chronic conditions and 14% having 6 or more chronic conditions." [25]

In Grand Challenges in Clinical Decision Support [1], Sittig et al. wrote:

The challenge is to create mechanisms to identify and eliminate redundant, contraindicated, potentially discordant, or mutually exclusive guideline-based recommendations for patients presenting with co-morbid conditions.

Michalowski et al. [26] proposed a mitigation framework based on a first-order logic (FOL) approach.

A CDS Architecture for the era of Precision Medicine

I proposed a scalable CDS architecture for Precision Medicine in another post titled Toward a Reference Architecture for Intelligent Systems in Clinical Care.

 

References

[1] Sittig D, Wright A, Osheroff JA, Middletone B, Jteich J, Ash J, et al. Grand challenges in clinical Decision support. J Biomed Inform 2008;41(2):387–92.

[2] Fox, J., Glasspool, D.W., Patkar, V., Austin, M., Black, L., South, M., et al. (2010). Delivering clinical decision support services: there is nothing as practical as a good theory. J. Biomed. Inform. 43, 831–843

[3] Valerio Cosentino, Marcos Didonet del Fabro, Adil El Ghali. A model driven approach for bridging ILOG Rule Language and RIF. RuleML, Aug 2012, Montpellier, France.

[4] Mauricio Salatino. (Processes & Rules) OR (Rules & Processes) 1/X. http://salaboy.com/2012/07/19/processes-rules-or-rules-processes-1x/. Retrieved February 15, 2015.

[5] Mauricio Salatino. (Processes & Rules) OR (Rules & Processes) 2/X. http://salaboy.com/2012/07/28/processes-rules-or-rules-processes-2x/. Retrieved February 15, 2015.

[6] Mauricio Salatino. (Processes & Rules) OR (Rules & Processes) 3/X. http://salaboy.com/2012/07/29/processes-rules-or-rules-processes-3x/. Retrieved February 15, 2015.

[7] Sylvie Dan. Modeling Clinical Rules with the Decision Modeling and Notation (DMN) Specification. http://sylviedanba.blogspot.com/2014/05/modeling-clinical-rules-with-decision.html. Retrieved February 15, 2015.

[8] Dennis Andrzejewski, Eberhard Beck, Eberhard Beck, Laura Tetzlaff. The transparent representation of medical decision structures based on the example of breast cancer treatment. 9th International Conference on Health Informatics.

[9] Phansalkar S, Zachariah M, Seidling HM, Mendes C, Volk L, Bates DW. Evaluation of medication alerts in electronic health records for compliance with human factors principles. J Am Med Inform Assoc. 2014 Oct;21(e2):e332-40. doi: 10.1136/amiajnl-2013-002279.

[10] Seidling HM, Phansalkar S, Seger DL, et al. Factors influencing alert acceptance: a novel approach for predicting the success of clinical decision support. J Am Med Inform Assoc 2011;18:479–84.

[11] Zachariah M, Phansalkar S, Seidling HM, et al. Development and preliminary evidence for the validity of an instrument assessing implementation of human-factors principles in medication-related decision-support systems--I-MeDeSA. J Am Med Inform Assoc 2011;18(Suppl 1):i62–72.

[12] Ayres I. Super Crunchers: Why Thinking-By-Numbers Is The New Way To be Smart. Bantam.

[13] Fox J., Cooper R. P., Glasspool D. W. (2013). A canonical theory of dynamic decision-making. Front. Psychol. 4:150 10.3389/fpsyg.2013.00150.

[14] Welch, B.M.; Eilbeck, K.; Del Fiol, G.; Meyer, L.; Kawamoto, K. Technical desiderata for the integration of genomic data with clinical decision support. 2014,

[15] Welch BM, Loya SR, Eilbeck K, Kawamoto K. A proposed clinical decision support architecture capable of supporting whole genome sequence information. J Pers Med. 2014 Apr 4;4(2):176-99. doi: 10.3390/jpm4020176.

[16] Loya SR, Kawamoto K, Chatwin C, Huser V. Service oriented architecture for clinical decision support: a systematic review and future directions. J Med Syst. 2014 Dec;38(12):140. doi: 10.1007/s10916-014-0140-z.

[17] Ewout W. Steyerberg. Clinical Prediction Models. A Practical Approach to Development, Validation, and Updating. New York: Springer, 2010.

[18] Sesen MB, Peake MD, Banares-Alcantara R, Tse D, Kadir T, Stanley R, Gleeson F, Brady M. 2014 Lung Cancer Assistant: a hybrid clinical decision support application for lung cancer care. J. R. Soc. Interface 11: 20140534. http://dx.doi.org/10.1098/rsif.2014.0534

[19] Bean RB, Bean, BW. Sir William Osler: aphorisms from his bedside teachings and writings. New York; 1950.

[20] Joel Amoussou. How good is your crystal ball?: Utility, Methodology, and Validity of Clinical Prediction Models. http://efasoft.blogspot.com/2015/01/how-good-is-your-crystal-ball-utility.html. Retrieved February 15, 2015.

[21] Dewey FE, Grove ME, Pan C, et al. Clinical Interpretation and Implications of Whole-Genome Sequencing. JAMA. 2014;311(10):1035-1045. doi:10.1001/jama.2014.1717.

[22] Joel Amoussou. Ontologies for Addiction and Mental Disease: Enabling Translational Research and Clinical Decision Support. http://efasoft.blogspot.com/2014/08/ontologies-for-addiction-and-mental.html. Retrieved February 2015.

[23] Future radiotherapy practice will be based on evidence from retrospective interrogation of linked clinical data sources rather than prospective randomized controlled clinical trials. Dekker, Andre L. A. J. and Gulliford, Sarah L. and Ebert, Martin A. and Orton, Colin G., Medical Physics, 41, 030601 (2014), DOI:http://dx.doi.org/10.1118/1.4832139

[24] Lambin, Philippe et al. 'Rapid Learning health care in oncology' – An approach towards decision support systems enabling customised radiotherapy. Radiotherapy and Oncology , Volume 109 , Issue 1 , 159 - 164.

[25] Centers for Medicare & Medicaid Services. Chronic Conditions Among Medicare Beneficiaries, Chartbook: 2012 Edition. http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/Downloads/2012Chartbook.pdf. Accessed Feb. 15, 2015.

[26] Szymon Wilk, Martin Michalowski, Xing Tan, Wojtek Michalowski: Using First-Order Logic to Represent Clinical Practice Guidelines and to Mitigate Adverse Interactions. KR4HC@VSL 2014: 45-61.

Statistical Computing and Machine Learning with R

The use of predictive risk models for personalized medicine is becoming a common practice in healthcare delivery. These models can predict the health risk of patients based on their individual health profiles. Examples include models for predicting breast cancer, stroke, cardiovascular disease, Alzheimer's disease, chronic kidney disease, diabetes, hypertension, and operative mortality for patients undergoing cardiac surgery. These predictive models are created through data analysis using statistical computing.

Predictive risk modeling can be used to identity at-risk populations and provide them with pro-active care including early screening and prevention. For example, predictive risk modeling can help identify patients at risk of hospital re-admission, an important Accountable Care Organization (ACO) quality measure.

Another important challenge in healthcare is to discover what works and what does not work in clinical practice. Comparative Effectiveness Research (CER), an emerging trend in Evidence Based Practice (EBP), has been defined by the Federal Coordinating Council for CER as "the conduct and synthesis of research comparing the benefits and harms of different interventions and strategies to prevent, diagnose, treat, and monitor health conditions in 'real world' settings."

Despite their inherent methodological challenges (lack of randomization leading to possible bias and confounding), observational studies (using real world clinical data) are increasingly recognized as complementary to Randomized Control Trials (RCTs) and an important tool in clinical decision making and health policy.

Statistical Computing and Machine Learning are essential components of intelligent health IT systems. Over the last few years, the free and open source R Project for Statistical Computing has emerged as one the most popular tools for data analysis. This poll by kdnuggets.com shows the breakdown in popularity of various data mining and analytic tools.

R supports several Machine Learning algorithms including:

• Nearest Neighbor
• Naive Bayes
• Decision Trees
• Logistic Regression
• Neural Networks
• Support Vector Machines
• Association Rules
• k-Means Clustering

A technique called "Ensemble Methods" which consists in combining multiple models into one can be used to achieve a higher level of accuracy than its components. There are also R packages for niche methods like the Latent Class Causal Analysis (LCCA) Package for R. LCA is used in behavioral health research.

The following are very useful resources for doing statistical computing and data mining with R:
 

• RStudio: an Integrated Development Environment (IDE) for R


• ggplot2: statistical graphics and plotting system for R


• sqldf: a package for manipulating R data frames using SQL


• RMySQL: R interface to the MySQL database


• RMongo: MongoDB Database interface for R
  


• RHIPE: Big Data analysis using R and Hadoop. RHIPE stands for R and Hadoop Integrated Programming Environment. This approach is referred to as D&R (Divide and Recombine) Analysis of Large Complex Data (see this tech report on D&R from the RHIPE team)


• RHadoop:  Big Data analysis using R and Hadoop. This tool provides Hadoop MapReduce functionality in R


• Rattle: A Graphical User Interface for Data Mining using R. This tool can export predictive models in Predictive Model Markup Language (PMML) format.

Automated Clinical Question Answering: The Next Frontier in Healthcare Informatics

In a previous post, I predicted that 2013 will be the year Intelligent Health IT Systems (iHIT) go mainstream.  I based my prediction on a number of factors, notably the transformation of healthcare to a value-based delivery system driven by the latest scientific evidence (evidence-based practice and practice-based evidence).

Last week, IBM together with health insurer WellPoint Inc., and New York’s Memorial Sloan-Kettering Cancer Center announced the commercialization of Watson (the supercomputer which beat human champions in "Jeopardy!" on February 16, 2011) for question answering (QA) in the clinical domain. The following are some interesting facts released by IBM as part of this announcement:

• The supercomputer has ingested 1,500 lung cancer cases from Sloan-Kettering records, plus 2 million pages of text from journals, textbooks and treatment guidelines. This is what I called Big Data in medicine.
• In 2012, Watson became 240 percent faster and 75 percent smaller so it can run on a single server. No surprise here and I expect this trend to continue.

The following YouTube video entitled Oncology Diagnosis and Treatment explains how IBM envisions using Watson for Clinical Question Answering (CQA):

The User Experience in the Watson Demo

 

• Clinical questions can be posed in natural language (spoken or typed in by the clinician using a keyboard).
• The sources used for answering clinical questions include both structured (EMR databases) and unstructured information (journal articles, clinical guidelines, etc.).
• Personalized medicine: the proposed interventions are driven by the data in the patient's medical record and the system can prompt the clinician for additional information on the patient if necessary. The displayed evidence and recommendations are updated to reflect changes in the patient's clinical data.
• Human Factors: the clinician is always in the loop. She can ask Watson how it arrives at a specific care recommendation and can even remove a specific evidence (if deemed irrelevant or not appropriate).
• The use of confidence scoring and evidence highlighting.
• Patient-centeredness and shared decision making: the treatment plans take into account the values, goals, and wishes of the patient (patient preferences). Treatment options are discussed with the patient.
• Comparative effectiveness is used to compare the benefits and harms of different interventions.
• Information is displayed using data visualization (dashboard) to help meet key performance indicators in the context of a value-based payment model.

The Science Behind Watson

The real question is how do we make intelligent health IT systems like Watson widely available to all patients. A landmark report published by the Institute of Medicine in 2001 and titled Crossing the Quality Chasm - A New Health System for the 21st Century contained the following recommendation:

Patients should receive care based on the best available scientific knowledge. Care should not vary illogically from clinician to clinician or from place to place.

For the scientifically (and Artificial Intelligence) inclined, the following are some pointers on the science behind Watson:

• The AI Behind Watson - The Technical Article
• Building Watson: An Overview of the DeepQA Project
• Unstructured Information Management (The IBM Journal of Research and Development)
• Semantic Web Technology in Watson.

The picture below represents a high level architecture of Watson (click on the image to enlarge it).

AskHermes and MiPACQ

IBM Watson is not the only effort to develop automated CQA capabilities.  Some earlier CQA efforts used the PICO framework (Problem/Population, Intervention, Comparison, Outcome) to facilitate processing. More recent efforts have focused on the use of clinical questions posed in natural language.

AskHermes (Help clinicians to Extract and aRrticulate Multimedia information for answering clinical quEstionS) allows clinicians to enter questions in natural language and uses the following unstructured information sources: MEDLINE abstracts, PubMed Central full-text articles, eMedicine documents, clinical guidelines, and Wikipedia articles.

The processing pipeline in AskHermes includes the following: Question Analysis, Related Questions Extraction, Information Retrieval, Summarization and Answer Presentation. AskHermes performs question classification using MMTx (MetaMap Technology Transfer) to map keywords to UMLS concepts  and semantic types. Classification is also achieved through supervised machine learning algorithms such as Support Vector Machine (SVM) and conditional random fields (CFRs). Summarization and answer presentation are based on clustering techniques.

MiPACQ (Multi-source Integrated Platform for Answering Clinical Questions) is based on Natural Language Processing (NLP) and Information Retrieval (IR) and utilizes data sources such as Electronic Medical Record (EMR) databases and online medical encyclopedia like Medpedia. MiPACQ uses a processing pipeline based on UIMA (Unstructured Information Management Architecture) and machine learning-based as well as rule-based scoring. NLP capabilities are provided by ClearTK and cTakes (clinical Text Analysis and Knowledge Extraction System).

The Road Ahead

Automated Clinical Question Answering (CQA) is really hard. However, that is the future of computing: intelligent machines we can have meaningful conversations with. CQA is a multidisciplinary field which combines disciplines like statistical computing, information retrieval, natural language processing, machine learning, rule engines, semantic web technologies, knowledge representation and reasoning, visual analytics, and massively parallel computing. There are several open source projects that provide the building blocks. Many EHR software today are glorified data entry systems. We need to move to the next level and that will require technical leadership.

Toward Intelligent Health IT (iHIT) Systems: Getting Out of the Box

In this post, I describe a new type of application that I refer to as iHIT. iHIT stands for Intelligent Health IT.

The Architecture of Traditional Health IT systems

Traditional software architectures for health IT systems typically include the following:

• Dependency Injection (DI)
  
  
• Object Relational Mapping (ORM)
  
  
• An architectural pattern for the presentation layer such as the Model View Controller (MVC) pattern
  
  
• HTML5, CSS3, and a JavaScript library like JQuery/Mobile
  
  
• Other architectural patterns including GoF Design Patterns, SOLID Principles, and Domain Driven Design (DDD)
  
  
• Structured Query Language (SQL)
  
  
• Enterprise Integration Patterns (EIPs) implemented through an Enterprise Service Bus (ESB) using HL7 messages as the "Published Language"
  
  
• REST or SOAP-based web services.
  

An entire generation of developers has been trained in these techniques. They represent proven best practices accumulated over several decades of object-oriented design and relational data management. Although pervasive in today's clinical systems, these applications lack basic intelligent features such as the ability to capture and execute expert knowledge, make inferences, or make predictions about the future based on the analysis of historical data. Some of these systems actually look like glorified data entry systems.

With the availability and explosion of medical knowledge and real world observational EHR data, these intelligent features will become important in assisting clinicians in the medical decision making process at the point of care by reducing their cognitive load.

Intelligent Health IT (iHIT) Systems

iHIT systems process huge quantities of both structured and unstructured data to provide clinicians with specific recommendations. iHIT systems play an important role in translating Comparative Effectiveness Research (CER) findings into clinical practice. Comparative effectiveness Research (CER), an emerging trend in Evidence Based Medicine (EBM), has been defined by the Federal Coordinating Council for CER as "the conduct and synthesis of research comparing the benefits and harms of different interventions and strategies to prevent, diagnose, treat and monitor health conditions in 'real world' settings." For example, based on the clinical profile of a patient, CER can help determine the best treatment option for breast cancer among the various options available such as: chemotherapy, radiation therapy, and surgery (Masectomy and Lumpectomy).

The following are examples of key characteristics displayed by iHIT systems:

• The ability to analyze patient data as well as very large historical observational data sets in order to make probability-based predictions about the future and recommend specific actions that can yield the best clinical outcomes given the clinical profile of a patient.
  
  
• The ability to capture and execute expert knowledge such as the medical knowledge contained in Clinical Practice Guidelines (CPGs). This includes the ability to mediate between different CPGs to arrive at a specific recommendation by merging and reconciling the medical knowledge in multiple CPGs as is the case with patients with comorbidities.
  
  
• The ability to perform automated reasoning by inferring new implicit clinical facts from existing explicit facts and by exploiting semantic relationships between concepts and entities.
  
  
• The ability to retrieve knowledge from unstructured data sources such as the biomedical research literature from sources like PubMed in order to answer clinical questions sometimes posed in natural language.
  
  
• The ability to learn over time (and hence become smarter) as the amount of processed data continues its exponential growth.
  
  
• Very fast response time to queries over very large data sets.
  

Sounds like Artificial Intelligence (AI)? I believe we are indeed witnessing the resurgence of AI and even the ideas of the Semantic Web in the healthcare industry. As healthcare costs and quality become national priorities for many countries around the world, the boundaries of computing will continue to be pushed further. Actually, some of the underlying principles of intelligent systems were originally developed decades and even centuries ago in the field of biomedical research. Williams Osler (1849-1919) famously said:

Medicine is a science of uncertainty and an art of probability.

Technologically advanced and competitive industries like financial services (e.g., credit eligibility and fraud detection), online retail (e.g., recommendation engine), and logistics (e.g., delivery route optimization) have adopted some of these technologies. Health IT developers now need to embrace them as well. This will require thinking out of the box.


The Ingredients of iHIT Systems

iHIT systems represent not one, but the integration of many different technologies. Mathematical Models, Statistical Analysis, and Machine Learning algorithms play an important role in iHIT systems. Examples include:

• Logistic Regression models
  
  
• Decision Trees
  
  
• Association Rules
  
  
• Bayesian Network
  
  
• Neural Networks
  
  
• Random Forests
  
  
• Time Series for temporal reasoning
  
  
• k-means Clustering
  
  
• Support Vector Machines (SVM)
  
  
• Probabilistic Graphical Models (PGMs) based on methods such as Bayesian networks and Markov Networks for making clinical decisions under uncertainty.
  

These algorithms can be used not only for making therapeutic predictions (e.g., the future hospitalization risk of a patient with Asthma), but also for dividing a population into subgroups based on the clinical profile of patients in order to achieve the best treatment outcomes.

Clinical Practice Guidelines (CPGs) are usually-based on Systematic Reviews (SRs) of Randomized Controlled Trials (RCTs) which are essentially scientific experiments. According to a report titled "Clinical Practice Guidelines (CPGs) We Can Trust" which was published last year by the Institute Of Medicine (IoM):

However, even when studies are considered to have high internal validity, they may not be generalizable to or valid for the patient population of guideline relevance. Randomized trials commonly have an under representation of important subgroups, including those with comorbidities, older persons, racial and ethnic minorities, and low-income, less educated, or low-literacy patients. Many RCTs and observational studies fail to include such "typical patients" in their samples; even when they do, there may not be sufficient numbers of such patients to assess them separately or the subgroups may not be properly analyzed for differences in outcomes.

On the other hand, observational studies using statistical analysis and machine learning algorithms operate on large real world observational data and can therefore provide feedback on the effectiveness of the actual use of different therapeutic interventions. Although very costly, RCTs are still considered the strongest form of evidence in EBM. Despite their inherent methodological challenges (lack of randomization leading to possible bias and confounding), observational studies are increasingly recognized as complementary to RCTs and an important tool in clinical decision making and health policy. iHIT systems play an important role in translating Comparative Effectiveness Research (CER) findings into clinical practice in the form of clinical decision support (CDS) interventions at the point of care.

iHIT systems also use business rules engines to capture and execute expert knowledge such as the medical knowledge contained in Clinical Practice Guidelines (CPGs). Examples include rules engines based on forward chaining inference, also known as production rule systems. These rules engines can be combined with Complex Event Processing (CEP) and Business Process Management (BPM) for intelligent decision making.

iHIT systems support ontologies such as those represented by the web ontology language (OWL) providing reasoning capabilities as well as the ability to navigate semantic relationships between concepts and entities.

More advanced iHIT systems have Natural Language Processing (NLP) and Automatic Speech Recognition (ASR) capabilities in order to answer clinical questions posed in natural language. They rely on Information Retrieval techniques like probabilistic methods for scoring the relevance of a document given a query and the application of supervised machine learning classification methods such as decision trees, Naive Bayes, K-Nearest Neighbors (kNN), and Support Vector Machines (SVM).

In some cases, the responsibilities of an iHIT system are performed by Intelligent Agents which are autonomous entities capable of observing the clinical environment and acting upon those observations.

For scalability and performance, iHIT systems often sit on NoSQL databases and run on massively parallel computing platforms like Apache Hadoop while leveraging the elasticity of the cloud.

Integrating these technologies is the main challenge posed by iHIT systems. An example is the integration between statistical and machine learning models, business rules, ontologies, and more traditional forms of computing such as object-oriented programming. Various solutions to these challenges have been proposed and implemented.

Human-Centered Design

Finally, iHIT systems fully embrace a human-centered design approach. They provide a seamless integration between automated decision logic and clinical workflows. They provide the clinician with detailed explanations of the rationale behind the actions they recommend. In addition, they use techniques like Visual Analytics to enhance human cognitive abilities in order to facilitate analytical reasoning over very large data sets.