Peter Szolovits

What Is a Knowledge Representation

Although knowledge representation is one of the central and, in some ways, most familiar concepts in AI, the most fundamental question about it -- What is it? -- has rarely been answered directly. Numerous papers have lobbied for one or another variety of representation, other papers have argued for various properties a representation should have, and still others have focused on properties that are important to the notion of representation in general. In this article, we go back to basics to address the question directly. We believe that the answer can best be understood in terms of five important and distinctly different roles that a representation plays, each of which places different and, at times, conflicting demands on the properties a representation should have. We argue that keeping in mind all five of these roles provides a usefully broad perspective that sheds light on some longstanding disputes and can invigorate both research and practice in the field.

Evaluating the State-of-the-Art in Automatic De-identification

To facilitate and survey studies in automatic de-identification, as a part of the i2b2 (Informatics for Integrating Biology to the Bedside) project, authors organized a Natural Language Processing (NLP) challenge on automatically removing private health information (PHI) from medical discharge records. This manuscript provides an overview of this de-identification challenge, describes the data and the annotation process, explains the evaluation metrics, discusses the nature of the systems that addressed the challenge, analyzes the results of received system runs, and identifies directions for future research. The de-indentification challenge data consisted of discharge summaries drawn from the Partners Healthcare system. Authors prepared this data for the challenge by replacing authentic PHI with synthesized surrogates. To focus the challenge on non-dictionary-based de-identification methods, the data was enriched with out-of-vocabulary PHI surrogates, i.e., made up names. The data also included some PHI surrogates that were ambiguous with medical non-PHI terms. A total of seven teams participated in the challenge. Each team submitted up to three system runs, for a total of sixteen submissions. The authors used precision, recall, and F-measure to evaluate the submitted system runs based on their token-level and instance-level performance on the ground truth. The systems with the best performance scored above 98% in F-measure for all categories of PHI. Most out-of-vocabulary PHI could be identified accurately. However, identifying ambiguous PHI proved challenging. The performance of systems on the test data set is encouraging. Future evaluations of these systems will involve larger data sets from more heterogeneous sources.

Causal understanding of patient illness in medical diagnosis

First generation AI in Medicine programs have clearly demonstrated the usefulness of AI techniques However, il has also been recognized that the use. of notions such as causal relationships, temporal patterns, and aggregate disease categories in these programs has been too weak From our study of clinicians behavior we realized that a diagnostic or therapeutic program must consider a case at various levels of detail to integrate overall understanding with detailed knowledge, To explore these issues, we have undertaken a study of the problem of providing expert consultation for electrolyte and acid-base disturbances We have partly completed an implementation of ABEL, the diagnostic component of the overall effort. In this paper we concentrate on ABLL.s mechanism for describing a patient. Called the patient-specific model, this description includes data about the patient as well as the programs hypothetical interpretations of these data in a multilevel causal network. The lowest level of this description consists of pathophysiological knowledge about the patient, which is successively aggregated into higher level concepts and relations, gradually shifting the content from pathophysiological to syndromic knowledge The aggregate level of this description summarizes the patient data providing a global perspective for efficient exploration of the diagnostic possibilities The pathophysiological level description provides the ability to handle complex clinical situations arising in illnesses with multiple etiologies, to evaluate the physiological validity of diagnostic possibilities being explored, and to organize large amounts of seemingly unrelated facts into coherent causal descriptions.

Electronic medical records for discovery research in rheumatoid arthritis

Electronic medical records (EMRs) are a rich data source for discovery research but are underutilized due to the difficulty of extracting highly accurate clinical data. We assessed whether a classification algorithm incorporating narrative EMR data (typed physician notes) more accurately classifies subjects with rheumatoid arthritis (RA) compared with an algorithm using codified EMR data alone.

Genetic Misdiagnoses and the Potential for Health Disparities

BACKGROUND For more than a decade, risk stratification for hypertrophic cardiomyopathy has been enhanced by targeted genetic testing. Using sequencing results, clinicians routinely assess the risk of hypertrophic cardiomyopathy in a patients relatives and diagnose the condition in patients who have ambiguous clinical presentations. However, the benefits of genetic testing come with the risk that variants may be misclassified. METHODS Using publicly accessible exome data, we identified variants that have previously been considered causal in hypertrophic cardiomyopathy and that are overrepresented in the general population. We studied these variants in diverse populations and reevaluated their initial ascertainments in the medical literature. We reviewed patient records at a leading genetic-testing laboratory for occurrences of these variants during the near-decade-long history of the laboratory. RESULTS Multiple patients, all of whom were of African or unspecified ancestry, received positive reports, with variants misclassified as pathogenic on the basis of the understanding at the time of testing. Subsequently, all reported variants were recategorized as benign. The mutations that were most common in the general population were significantly more common among black Americans than among white Americans (P<0.001). Simulations showed that the inclusion of even small numbers of black Americans in control cohorts probably would have prevented these misclassifications. We identified methodologic shortcomings that contributed to these errors in the medical literature. CONCLUSIONS The misclassification of benign variants as pathogenic that we found in our study shows the need for sequencing the genomes of diverse populations, both in asymptomatic controls and the tested patient population. These results expand on current guidelines, which recommend the use of ancestry-matched controls to interpret variants. As additional populations of different ancestry backgrounds are sequenced, we expect variant reclassifications to increase, particularly for ancestry groups that have historically been less well studied. (Funded by the National Institutes of Health.).

Artificial Intelligence in Medical Diagnosis

In an attempt to overcome limitations inherent in conventional computer-aided diagnosis, investigators have created programs that simulate expert human reasoning. Hopes that such a strategy would lead to clinically useful programs have not been fulfilled, but many of the problems impeding creation of effective artificial intelligence programs have been solved. Strategies have been developed to limit the number of hypotheses that a program must consider and to incorporate pathophysiologic reasoning. The latter innovation permits a program to analyze cases in which one disorder influences the presentation of another. Prototypes embodying such reasoning can explain their conclusions in medical terms that can be reviewed by the user. Despite these advances, further major research and developmental efforts will be necessary before expert performance by the computer becomes a reality.

Normalization of plasma 25-hydroxy vitamin D is associated with reduced risk of surgery in Crohn's disease.

Background:Vitamin D may have an immunologic role in Crohn’s disease (CD) and ulcerative colitis (UC). Retrospective studies suggested a weak association between vitamin D status and disease activity but have significant limitations. Methods:Using a multi-institution inflammatory bowel disease cohort, we identified all patients with CD and UC who had at least one measured plasma 25-hydroxy vitamin D (25(OH)D). Plasma 25(OH)D was considered sufficient at levels ≥30 ng/mL. Logistic regression models adjusting for potential confounders were used to identify impact of measured plasma 25(OH)D on subsequent risk of inflammatory bowel disease–related surgery or hospitalization. In a subset of patients where multiple measures of 25(OH)D were available, we examined impact of normalization of vitamin D status on study outcomes. Results:Our study included 3217 patients (55% CD; mean age, 49 yr). The median lowest plasma 25(OH)D was 26 ng/mL (interquartile range, 17–35 ng/mL). In CD, on multivariable analysis, plasma 25(OH)D <20 ng/mL was associated with an increased risk of surgery (odds ratio, 1.76; 95% confidence interval, 1.24–2.51) and inflammatory bowel disease–related hospitalization (odds ratio, 2.07; 95% confidence interval, 1.59–2.68) compared with those with 25(OH)D ≥30 ng/mL. Similar estimates were also seen for UC. Furthermore, patients with CD who had initial levels <30 ng/mL but subsequently normalized their 25(OH)D had a reduced likelihood of surgery (odds ratio, 0.56; 95% confidence interval, 0.32–0.98) compared with those who remained deficient. Conclusion:Low plasma 25(OH)D is associated with increased risk of surgery and hospitalizations in both CD and UC, and normalization of 25(OH)D status is associated with a reduction in the risk of CD-related surgery.

Genetic Basis of Autoantibody Positive and Negative Rheumatoid Arthritis Risk in a Multi-ethnic Cohort Derived from Electronic Health Records

Discovering and following up on genetic associations with complex phenotypes require large patient cohorts. This is particularly true for patient cohorts of diverse ancestry and clinically relevant subsets of disease. The ability to mine the electronic health records (EHRs) of patients followed as part of routine clinical care provides a potential opportunity to efficiently identify affected cases and unaffected controls for appropriate-sized genetic studies. Here, we demonstrate proof-of-concept that it is possible to use EHR data linked with biospecimens to establish a multi-ethnic case-control cohort for genetic research of a complex disease, rheumatoid arthritis (RA). In 1,515 EHR-derived RA cases and 1,480 controls matched for both genetic ancestry and disease-specific autoantibodies (anti-citrullinated protein antibodies [ACPA]), we demonstrate that the odds ratios and aggregate genetic risk score (GRS) of known RA risk alleles measured in individuals of European ancestry within our EHR cohort are nearly identical to those derived from a genome-wide association study (GWAS) of 5,539 autoantibody-positive RA cases and 20,169 controls. We extend this approach to other ethnic groups and identify a large overlap in the GRS among individuals of European, African, East Asian, and Hispanic ancestry. We also demonstrate that the distribution of a GRS based on 28 non-HLA risk alleles in ACPA+ cases partially overlaps with ACPA- subgroup of RA cases. Our study demonstrates that the genetic basis of rheumatoid arthritis risk is similar among cases of diverse ancestry divided into subsets based on ACPA status and emphasizes the utility of linking EHR clinical data with biospecimens for genetic studies.

Building National Electronic Medical Record Systems via the World Wide Web

Electronic medical record systems (EMRSs) currently do not lend themselves easily to cross-institutional clinical care and research. Unique system designs coupled with a lack of standards have led to this difficulty. The authors have designed a preliminary EMRS architecture (W3-EMRS) that exploits the multiplatform, multiprotocol, client-server technology of the World Wide Web. The architecture abstracts the clinical information model and the visual presentation away from the underlying EMRS. As a result, computation upon data elements of the EMRS and their presentation are no longer tied to the underlying EMRS structures. The architecture is intended to enable implementation of programs that provide uniform access to multiple, heterogeneous legacy EMRSs. The authors have implemented an initial prototype of W3-EMRS that accesses the database of the Boston Childrens Hospital Clinicians Workstation.

Global conditioning for probabilistic inference in belief networks

In this paper we propose a new approach to probabilistic inference on belief networks, global conditioning, which is a simple generalization of Pearls (1986b) method of loop-cutset conditioning. We show that global conditioning, as well as loop-cutset conditioning, can be thought of as a special case of the method of Lauritzen and Spiegelhalter (1988) as refined by Jensen et al (1990a; 1990b). Nonetheless, this approach provides new opportunities for parallel processing and, in the case of sequential processing, a tradeoff of time for memory. We also show how a hybrid method (Suermondt and others 1990) combining loop-cutset conditioning with Jensens method can be viewed within our framework. By exploring the relationships between these methods, we develop a unifying framework in which the advantages of each approach can be combined successfully.