Visualization of dynamic simulations of muscle thin filaments.

Following a novel computational formalism, the thin filament of muscle can be modeled by a computational machine containing a large number of finite automata that have one-to-one correspondence with the constituent protein molecules. Computer graphics can be used to visualize the correspondence between the states of finite automata and the configurations of protein molecules according to the structural data. The dynamic simulation of the muscle filament that corresponds to the concurrent state transitions of finite automata can be represented as a sequence of video images. The kinetic and structural knowledge of individual protein molecules is, therefore, integrated into a coherent and functional system. This type of computation and visualization can also be useful for the investigation of molecular structure, function, and interaction in various complex biological systems.

IAIMS development at Baylor College of Medicine.

At Baylor College of Medicine, we are developing the technical and intellectual resources needed to realize the Integrated Academic Information Management System (IAIMS) concept fully. The substantial technical, organizational, and financial commitments involved demand that we align our efforts with the strategic purposes of the college. The support of science, therefore, has become the principal, but not exclusive, focus of Baylor's IAIMS effort. Even so, the information technology architecture we have created for biomedical research is proving valuable in other settings as well. And the infrastructure we are creating--the communications architecture and the linkages to information resources--serves many purposes in addition to those of research. The architecture accommodates a diversity of workstations, networks, and informational and computational servers. This will be the greatest possible chance of transferring the fruits of our Phase III development to other academic medical centers.

The MEDLINE Retriever.

Baylor College of Medicine has developed the MEDLINE Retriever, a tool to query MEDLINE, the data-base of medical literature at the National Library of Medicine. The MEDLINE Retriever communicates via the Internet to achieve excellent response time for MEDLINE queries. It uses the X Window System and the Motif toolkit, and employs the Knowbot Operating Environment developed by the Corporation for National Research Initiatives. We discuss the architecture of the MEDLINE Retriever, focusing on the graphical user interface that we have developed, as well as our experiences in developing and deploying the MEDLINE Retriever at Baylor. The MEDLINE Retriever is an extension of Baylor's IAIMS design concept that brought forth the Virtual Notebook System, and fits well with Baylor's aims with regard to the High Performance Computing Initiative.

Medical informatics and institutional strategy.

Programs in medical informatics can help academic medical centers make effective use of information technology. But to achieve the greatest strategic benefits from these programs, an institution must forge proper linkage between informatics and its overall effort to deploy computing in research, education and patient care. Here we explore this linkage and the ways in which it can be established and managed.

A Virtual Notebook for biomedical work groups.

During the past several years, Baylor College of Medicine has made a substantial commitment to the use of information technology in support of its corporate and academic programs. The concept of an Integrated Academic Information Management System (IAIMS) has proved central in our planning, and the IAIMS activities that we have undertaken with funding from the National Library of Medicine have proved to be important extensions of our technology development. Here we describe our Virtual Notebook system, a conceptual and technologic framework for task coordination and information management in biomedical work groups. When fully developed and deployed, the Virtual Notebook will improve the functioning of basic and clinical research groups in the college, and it currently serves as a model for the longer-term development of our entire information management environment.

Toward a theory of clinical expertise.

We believe that the outline of a promising approach to the study of expert clinical problem-solving is emerging. This process involves collecting as verbal transcripts the thoughts of experienced clinicians engaged in simulated clinical tasks and analyzing these transcripts to formulate theories of problem-solving tactics and strategies. In turn, these theories are embodied in a computer program that simulates the problem-solving process. The computer program's response to fresh clinical problems can be compared with physician behavior in order to revise, refute, or confirm the problem-solving theories. A scientific understanding of the expert clinician's reasoning strategies could prove exceptionally useful in improving the way we teach and practice medicine and in developing practical computer aids for medical decision-making.

A new approach for evaluating risk factors in coronary artery disease: a study of lipid concentrations and severity of disease in 1847 males.

Using recently developed statistical techniques, we estimated joint probability density functions for plasma cholesterol and plasma triglyceride concentrations in 12 subgroups of 1847 males with angiographically demonstrated coronary artery disease. The subjects were grouped by patterns of age, history of hypertension, history of smoking and severity of disease. From the estimated joint probability density functions, we determined the way in which the risk of severe vs moderate coronary artery disease varied with the joint variation of plasma lipids for given combinations of the other risk factors. The patterns of risk that we found are evidence of the complicated way in which the relationship between plasma lipids and the severity of coronary artery disease is affected by other risk factors.

The process of ambulatory care: a comparison of the hospital and the community health center.

In a study of 300 chronically ill patients who were referred from an outpatient department to community health centers in a public hospital district, we found that the number of visits made, the amount of medicine prescribed, and the number of tests ordered all increased markedly for those patients in the 6 months immediately following their referral. As a result, there was substantial increase in the charges incurred for their care. The care itself changed as well. The physicians at the centers on the whole differed from those at the hospital in the emphasis they placed upon various types of visits, medicines, and tests. There were also large differences among the community health centers along these dimensions. The findings indicate the difficulty of providing comparable care at a comparable cost within a large health care system.

Scientific publications of a medical school faculty.

To characterize patterns of scientific productivity in a medical school faculty, the authors assessed the quantity and apparent scientific quality of publications by a large sample of the full-time faculty at Baylor College of Medicine. The apparent quality of an article in a given journal was taken to be the average extent to which articles in that journal are cited. Productivity varied greatly among the faculty members; a few prolific authors produced 10 times as much as the average of their colleagues, and a substantial fraction of the faculty published nothing in a four-year period. When productivity was related to length of academic career, two peaks of scientific accomplishment were noted. The first was for young faculty members whose relatively few publications were of high apparent quality. The second was for more mature faculty whose high productivity was attributable to more publications of less apparent quality.

Clinical problem solving: a behavioral analysis.

To extend the understanding of the clinical problem-solving process, we have analyzed the tape-recorded behavior of experienced clinicians engaged in "taking the history of the present illness" from a simultated patient. We showed that specific diagnostic hypotheses were generated often with little more information than presenting complaints, that testing of diagnostic hypotheses consisted of various case-building strategies for corroborating and discrediting hypotheses, and that the process of information gathering included techniques to evaluate the validity of data and assess the need for immediate action. Overall strategies were more difficult to discern but included a focused approach, a systemic exploration method, and a chronologic technique. The data have potential value in medical education and in developing computer programs to simulate the diagnostic process.

The diagnostic importance of the normal finding.

When a diagnostic test is reported as normal, the clinician generally uses it only to rule out certain diseases. However, if properly interpreted, the normal value may help to differentiate among diagnoses that yield normal results with different frequencies. A simple method permits the extraction of such information. The physician estimates the probability of various diagnoses and then combines these estimates with the anticipated frequency of negative results for each disease under consideration.