Showing papers in "Journal of Medical Systems in 1991"

Using neural networks to diagnose cancer

Abstract: While artificial brains are in the realm of science fiction, artificial neural networks (ANNs) are scientific facts. An artificial neural network is a computational structure modeled somewhat on the neural structure of the brain; both have many highly interconnected processing elements. These biologically inspired processing elements are taught by feeding examples until the results are acceptable. In the past 5 years, neural networks have become successful in providing meaningful second opinions in clinical diagnosis. In our research, a prototype artificial neural network was trained on numeral ultrasound data of 52 actual cases and then correctly identified renal cell carcinoma from renal cysts and other conditions without diagnostic errors. Our nonlinear artificial neural network was trained on software using the standard backpropagation paradigm on a 80386 microcomputer. Our ANN learned from ultrasound data in 52 cases (17 malignant, 30 cysts, and 5 other) at a Memphis hospital. The trained prototype performed without error on 47 cases which were not in the data used for training. This prototype must be validated by extending this study to more cases.

Iliad training enhances medical students' diagnostic skills

Abstract: Iliad is a computerized, expert system for internal medical diagnosis. The system is designed to teach diagnostic skills by means of simulated patient case presentations. We report the results of a controlled trial in which junior students were randomly assigned to received Iliad training on one of two different simulated case mixes. Each group was subsequently tested in both their "trained" and "untrained" case domain. The testing consisted of computerized, simulated patient cases for which no training feedback was provided. Outcome variables were designed to measure the students' performance on these test cases. The results indicate that students made fewer diagnostic errors and more conclusively confirmed their diagnostic hypotheses when they were tested in their trained domain. We conclude that expert systems such as Iliad can effectively teach diagnostic skills by supplementing trainees' actual case experience with computerized simulations.

Significant efficiency findings while controlling for the frequent confounders of CAI research in the PlanAlyzer project's computer-based, self-paced, case-based programs in anemia and chest pain diagnosis.

Abstract: Richard E. Clark in his widely published comprehensive studies and meta-analyses of the literature on computer assisted instruction (CAI) has decried the lack of carefully controlled research, challenging almost every study which shows the computer-based intervention to result in significant post-test proficiency gains over a non-computer-based intervention. We report on a randomized study in a medical school setting where the usual confounders found by Clark to plague most research, were carefully controlled. PlanAlyzer is a microcomputer-based, self-paced, case-based, event-driven system for medical education which was developed and used in carefully controlled trials in a second year medical school curriculum to test the hypothesis that students with access to the interactive programs could integrate their didactic knowledge more effectively and/or efficiently than with access only to traditional textual “nonintelligent” materials. PlanAlyzer presents cases, elicits and critiques a student's approach to the diagnosis of two common medical disorders: anemias and chest pain. PlanAlyzer uses text, hypertext, images and critiquing theory. Students were randomized, one half becoming the experimental group who received the interactive PlanAlyzer cases in anemia, the other half becoming the controls who received the exact same content material in a text format. Later in each year there was a crossover, the controls becoming the experimentals for a similar intervention with the cardiology PlanAlyzer cases. Preliminary results at the end of the first two full trials shows that the programs have achieved most of the proposed instructional objectives, plus some significant efficiency and economy gains. 96 faculty hours of classroom time were saved by using PlanAlyzer in their place, while maintaining high student achievement. In terms of student proficiency and efficiency, the 328 students in the trials over two years were able to accomplish the project's instructional objectives, and the experimentals accomplished this in 43% less time than the controls, achieving the same level of mastery. However, in spite of these significant efficiency findings, there have been no significant proficiency differences (as measured by current factual and higher order multiple choice post-tests) between the experimental and control groups. Very careful controls were used to avoid what Clark has found to be the most common confounders of CAI research. Accordingly, this research proved Clark's rival hypothesis, that the computer, in itself, does not appear to contribute to proficiency gains, at least as measured by our limited post-testing. Clark's position is that the computer is primarily a vehicle—as is either a pill or a hypodermic needle for delivering a drug. The hypodermic needle can deliver the drug more efficiently than can the pill, (as can the computer deliver the subject matter content more efficiently, as our research indicates), but the same content is delivered. At the same time, we proved our own hypothesis, as far as efficiency gains resulting from the computer are concerned. However, going beyond Clark's research, we may be teaching processes both more effectively and efficiently with the computer (experience in problem-solving or clinical reasoning and pattern recognition) which our current post-tests do not adequately measure. Our on-going research suggests additional inquiry in several areas: better evaluation instruments to measure the clinical reasoning skills PlanAlyzer was designed to teach; the addition of more advanced cases to determine if this might transform efficiency gains of the computer group into proficiency gains; the addition of enhanced graphic decision support tools and other pedagogical enhancements including cognitive feedback to strengthen PlanAlyzer's power to teach complex concepts of medical decision-making.

A model for designing intelligent tutoring systems

Abstract: A model for the design of an intelligent tutoring system is presented using artificial intelligence techniques and cognitive processing theories. The model of cognitive processing known as the Knowledge Acquisition and Recall Theory is derived from Anderson's Adaptive Control of Thought Theory. The model is used as a basis for a methodology for the design of an intelligent tutoring system that teaches problem-solving strategies for blood grouping discrepancies. The system developed was tested to determine the effectiveness of the methodology developed and to provide support for the concepts in the model of cognition developed. The preliminary results provide some evidence that an intelligent tutoring system designed using the methodology developed may aid in the knowledge acquisition process. (knowledge acquisition, artificial intelligence, computer-assisted instruction, cognition, intelligent tutoring system).

Computer Rx: more harm than good?

Abstract: Clinical information systems (CIS) are health care technologies that can assist clinicians and clinical managers to improve the performance of health care organizations. However, failure to consider scientific evidence of efficacy, effectiveness, and efficiency when selecting CISs is one factor explaining the adoption of systems that do not improve either the quality or efficiency of patient care. This paper discusses a technology assessment framework that can assist decision-makers to evaluate alternative CISs. Existing methodologies developed to evaluate diagnostic and therapeutic technologies can be used by researchers to provide evidence needed by decision-makers at each step of the framework. The rigorous evaluation of CISs prior to their implementation can help decision-makers to avoid adopting “white elephants.”

A very large PC LAN as the basis for a hospital information system.

Abstract: Brigham and Women's Hospital is converting its financial, administrative and clinical information systems from a mini-computer environment to a platform based on MUMPS and a network of several thousand personal computers. This article describes the project rationale and status and provides an overview of the architecture of the new system. The initial results of the project indicate that the personal computer network can provide large amounts of processor power and storage at costs per unit of power and storage that are several times less expensive than the minicomputer environment. The performance of the converted Accounts Payable system indicates that the architecture can deliver acceptable performance.

AUTRES—the Johns Hopkins Hospital automated resume

Abstract: The Johns Hopkins Hospital has developed AUTRES, an electronic discharge summary. This computer generated report was designed to provide improved continuity in the care of patients by offering a legible, standardized summary of the events of hospitalization. It can be sent by mail, FAX, or electronically (computer-to-computer) to the next care giver who will see the patient. It is available from any network attached terminal at the hospital and School of Medicine to support re-admission to the hospital, treatment in the emergency room, or outpatient visits. AUTRES first ran in fully supported production serving the Department of Medicine in 1988. Re-engineered to take advantage of cooperative processing, it has been implemented in Neurology, Internal Medicine, Obstetrics and Gynecology and Pediatrics, with plans for phased installation in the remaining clinical departments of The Johns Hopkins Hospital.

A random allocation system with the minimization method for multi-institutional clinical trials

Abstract: This paper describes the random allocation system used to perform precise and rapid treatment assignments in multi-institutional clinical trials. This system is based on sophisticated randomization procedures, according to Pocock and Simon's minimization method and Zelen's method for institution balancing. The major advantage of randomized treatment assignments with this system is to balance treatment numbers for each level of various prognostic factors over the entire trial and at the same time balance the allocation of treatments within an institution. Therefore, the randomized treatment assignments by this system can prevent degrading of the statistical power of a particular treatment factor. This system is designed to run on a small-sized notebook computer and therefore can be set up beside a telephone for registration, without occupying a large space. At present, this system is conveniently being used in two clinical trials.

Making the transition from information systems of the 1970s to medical information systems of the 1990s: the role of the physician's workstation

Abstract: Many hospitals today have implemented widely disparate information systems on mainframe and mini-computer hardware. The advent of network technology in hospitals has made it possible to access information in these systems. Unfortunately, the user interfaces to applications on these systems are unique and difficult to learn, which makes them unsuitable for use by clinical services. In this paper we describe the development of a Physician's Workstation which integrates information from multiple existing information systems and discuss how the workstation makes it possible to move from the departmental systems of the present to the computer-based medical record system of the future.

A distributed approach to integrated inquiry and display for radiology

Abstract: Picture archive and communications (PACS) systems should be flexible and modular in design so that new advances in storage, computation, and display technology can be introduced into the system without a significant redesign of existing software. The acquisition, storage, and management of radiologic images must be carefully integrated with a radiology information system. Our architecture is based on a four-level data model: (1) patient information, (2) examination information and reports, (3) image information, and (4) instances of images. The PACS being developed at the Mallinckrodt Institute of Radiology within the Electronic Radiology Laboratory consists of three primary components: application clients, database servers and image servers. One type of application client is an image-capable workstation that supports a radiology image viewing application. The application client queries the database server for information regarding patient and examination data in response to user-level requests. The database server responds to the request by retrieving the appropriate patient demographics and examination information, along with a pointer to the image/instance data from a central database. The client then uses the image data pointer to query the image server for the actual pixel data. The image server responds by transmitting the pixel data to the requesting application client or a designated auxiliary display device. Other clients act as image data acquisition nodes. Queries to the database servers are made via a library of callable subroutines. Software integrity is maintained throughout the system by dynamically loading software from a code-control database. Inquiry and display transactions, supported on a local-area network (Ethernet), have been measured and analyzed. Results and observations are presented.

A new teaching tool for dental and medical schools: multimedia CD-ROM on Dental and skeletal development

Abstract: Multimedia teaching tools in Medicine are still very rare. Their presentation on a CD-ROM format is part of the emerging technologies in Medical Informatics. The Multimedia material presented here on Dental and Skeletal Development is mostly a Database and a Tutorial. The Database, which contains more than 2000 hand/wrist radiographs, is mostly intended for Pediatricians, Endocrinologists, Forensic Dentists and Physicians, as well as for Orthodontists and Pedodontists. The material is the result of 20-year of Longitudinal Growth study of children in Montreal. In this Multimedia CD-ROM, the research results are used for Medical teaching purposes.

Health managers' attitudes toward robotics and artificial computer intelligence: an empirical investigation

Abstract: An empirical investigation of the perceived role of robotics and artificial computer intelligence in the future of health care reveals factors favoring a positive attitude by health administrators. The study employed a two-part survey administered in late 1989 and early 1990 to health care managers in hospitals and nursing homes. Part One of the survey asked about the participant, his or her work habits and work environment. Part Two obtained a psychological profile of rationality vs. intuition in problem solving. Through bivariate and multivariate post hoc statistical tests, we discovered the following variables which significantly determined attitudes toward robotics and artificial computer intelligence: sex; number of employees supervised; perceptions of waste and inefficiency in the workplace; perceptions of time-consuming personnel problems; perceived need to make more efficient use of time, money, and facilities; and perceived favorable climate for innovation. Among the factors which did not have an effect on attitudes toward advanced technology were three measures of rationality vs. intuition in problem solving.

A multicenter study of data collection and communication and communication at primary health care centers

Abstract: Health care delivery is information intensive. As computer applications make information available to the decision maker with speed and accuracy, informatics applications will strengthen the infrastructure. This paper is the second part of a multicenter systems analysis study to design a common application software to support primary health care focused on information flow. We present the questionnaire analysis and observations from a field study of a district health site. Analyses using contingency tables revealed differences, some statistically significant. The field study confirmed that minor differences exist even within a district health site. Development of a common application software on the basis of information flow studies is feasible. However, to make optimum use of computer implementation, revision of the health information systems was recommended. It was suggested that application software be developed with thecore data set required by the care providers to deliver and administrators to manage a vertical health program.

A new microcomputer software system evaluation paradigm: the medical perspective

Abstract: The new fourth-generation software has enormously eased the burden of computing for users, but they have also created a confusing, difficult problem in software evaluation and selection. Therefore it is argued that a sound and complete evaluation paradigm is a key element in an efficient and effective software system design and use process. While much has been written about software evaluation in general, in the medical field, the guidances and recommendations previously provided are too general to be of practical use. Considering also the other weaknesses of conventional evaluation paradigms we have decided to develop a more adequate one which will have a solid theoretical framework, specific guidances, strict and well-defined taxonomic space, and a fair ranking approach. In the present paper we will therefore introduce our new evaluation paradigm and show its applicability in evaluation and selection of medical software systems according to their usability.

Pathologic elements in the planning of information systems: The case of hospitals in Quebec

Abstract: The existing usage of information technology for information systems in hospitals draws attention to the fact that: (a) health care units which operate sophisticated numerical instrumentation to support medical activities rarely employ computers for the management of their operations and (b) despite the availability of affordable information technologies (microcomputers, networks, etc.), few hospitals include such elements as health care unit management in the development of their information systems portfolio. Is this paradoxical situation a result of the planning methodologies or of factors affecting their use? Or can it be attributed to the failure of present planning methodologies to incorporate the diversity of computerization approaches of the various stakeholders involved? If this is the case, untapped potential computerization possibilities exist that could lead to the development of hospital services and increased organizational efficiency.

An inventory model for optimizing purchasing of intravenous fluids for hospitals: a case study.

Abstract: This paper deals with determining the optimal level of purchase of intravenous fluids for hospitals by utilizing inventory control concepts. Using the conceptual framework of the economic order quantity (EOQ) model, the hospital administrators can elicit the efficient materials management, thus reducing both the space and capital requirements without compromising the quality of services rendered.

Migrating a clinical laboratory information system between technologies

Abstract: The technical revolution that has strongly driven events in the clinical laboratory for the last thirty years is now threatening to make obsolete what has become the central pillar of operation in many laboratories, the minicomputer-based laboratory information system. Some of its functions could easily be absorbed by the personal computers which are proliferating in the laboratory, but any single step leap between systems risks replacing order with chaos. Appropriate use of networking tools, together with essential software development, can provide a systematic migrational path for both the administrative and technical computer support from one environment to another without the trauma of a massive replacement step.

MDA-image: an environment of networked desktop computers for teleradiology/pathology.

Abstract: MDA-Image, a project of The University of Texas M. D. Anderson Cancer Center, is an environment of networked desktop computers for teleradiology/pathology. Radiographic film is digitized with a film scanner and histopathologic slides are digitized using a red, green, and blue (RGB) video camera connected to a microscope. Digitized images are stored on a data server connected to the institution's computer communication network (Ethernet) and can be displayed from authorized desktop computers connected to Ethernet. Images are digitized for cases presented at the Bone Tumor Management Conference, a multidisciplinary conference in which treatment options are discussed among clinicians, surgeons, radiologists, pathologists, radio-therapists, and medical oncologists. These radiographic and histologic images are shown on a large screen computer monitor during the conference. They are available for later review for follow-up or representation.

The use of X-terminals as clinical workstations

Abstract: The Medical Computer Facility at the Fox Chase Cancer Center has installed X-terminals in patient examination rooms and at nursing stations for clinical data acess by physicians and nurses. The X-terminals are connected to UNIX operating system RISC processors via Ethernet. The RISC processors communicate with databases on a minicomputer cluster. Simultaneous presentation of textual (e.g., pathology and radiology reports) and graphical (e.g., clinical laboratory results) clinical data is provided under X-Windows. CT and MRI images can also be displayed in windows. Our experiences implementing X-terminal clinical workstations in a production environment will be discussed.

Designing clinical protocols for optimal use: measuring attributes of treatment and cancer control trials

Abstract: The design of clinical protocols is an important factor in assuring patient accrual and compliance. This paper describes the methods used to assess the attributes of treatment and cancer control research protocols. Measures are presented for assessing three attributes considered important: relative advantage, complexity, and compatibility.

Micro-mainframe-link personal clinical research system

Abstract: We constructed a micro-mainframe-link clinical research system for personal use (Personal Clinical Research System). This system was developed with both a mainframe computer and a personal computer (PC). The prepared programs included a database manager (on the mainframe computer), a user interface program (on the PC), and a communication control program that connected the mainframe computer with the PC. The database on the mainframe computer was constructed by two methods. The first method was to transmit data from the PC to the mainframe computer. The second method was to extract data from the patient information database. Using this system, a physician is able to construct a personal research database that contains interesting data for the physician. In addition, the physician is able to accumulate data on a special field using this system. A discharge summary system is now in operation as an example of this system.

BioSYNTHESIS: access to a knowledge network of health sciences databases.

Abstract: Users of the IAIMS Knowledge Network at the Georgetown University Medical Center have access to multiple in-house and external databases from a single point of entry through BioSYNTHESIS. The IAIMS project has developed a rich environment of biomedical information resources that represent a medical decision support system for campus physicians and students. The BioSYNTHESIS system is an information navigator that provides transparent access to a Knowledge Network of over a dozen databases. These multiple health sciences databases consist of bibliographic, informational, diagnostic, and research systems which reside on diverse computers such as DEC VAXs, SUN 490, AT&T 3B2s, Macintoshes, IBM PC/PS2s and the ATT ISN and SYTEK network systems. Ethernet and TCP/IP protocols are used in the network architecture. BioSYNTHESIS also provides network links to the other campus libraries and to external institutions. As additional knowledge resources and technological advances have become available, BioSYNTHESIS has evolved from a two phase to a three phase program. Major components of the system including recent achievements and future plans are described.

Quality assurance tracking or "finding the needle in a haystack".

Abstract: This paper will present a picture of the steps required from conception, through development, and leading to implementation of an automated Quality Assurance Incident Tracking System within in a multihospital system utilizing general purpose software (WordPerfect and Lotus Agenda). This automated tracking system provides the manager with an effective and efficient method to control multiple staff assignments, tasks, and manage multiple objectives over the resolution cycle associated with hospital incident reporting in the Quality Assurance arena. It also provides the QA managers and hospital executives with an individualized record of the Incident's progress and helps identify useful and dysfunctional patterns in the resolution of the Incident. The identification of these patterns can assist hospital management in revising and developing protocols and procedures that minimize hospital exposure to state sanctions and provide a mechanism to improve the quality of care in the institution: QA Tracking, Quality Assurance, Incident, Automated Tracking. ty of care in the institution: QA Tracking, Quality Assurance, Incident, Automated Tracking.

An assessment of the radiological module of NEONATE as an aid in interpreting chest x-ray findings by nonradiologists

Abstract: NEONATE is a prototype of an expert system for the Newborn Intensive Care Unit developed at the Primary Children's Medical Center in Salt Lake City, Utah. A pilot study was undertaken to see if the addition of radiological frames to the NEONATE software could aid attending Neonatologists to interpret chest X-ray films. A set of radiological frames was created from rules generated by a radiologist. The performance of these radiological frames was compared to the performance of other radiologists using the Kappa statistic to measure agreement. There is a good agreement between the computer's decisions and the radiologists' decisions. The radiological frames were also tested to see if they help physicians who are not trained in radiology. A system that compares the residents' interpretation and the computer's interpretation to a gold standard interpretation was developed. It shows that the computer helps the first and second year residents, but not the third year residents. This article suggests that NEONATE's interpretation of chest X-ray findings are close to the radiologists' interpretations. While NEONATE's radiological frames help novice physicians in reaching better chest X-ray interpretation, the current study suggests that they are not likely to help a Neonatologist.

The use of graphic design in an interactive computer teaching program.

Abstract: The widespread diffusion of affordable computers into the scientific and educational community has provided the opportunity to design medical and scientific teaching programs illustrated either by hand or by utilizing commercially available software and manipulating existing computer generated images. The medical illustrator can provide the ideal aesthetic link between text format information and the visual representation of such knowledge in a concise presentation format. The availability of interactive multimedia programs has given the medical illustrator an environment to create and enhance Hypermedia designed specifically for the parpose of medical education. This paper will focus on the incorporation of illustration and screen design into “CT The Came,” an experimental medical teaching program currently being developed in the Johns Hopkins Body CT Imaging Laboratory. The program is designed to provide an enjoyable approach to learning Computed Tomography (CT), and is directed toward an audience of medical students, residents, and fellows.

A stochastic model for determining the necessary staff level in a service industry

Abstract: Determining sufficient staff levels and providing quality, economical and efficient service is a problem that is difficult for some staffing managers to solve. The objectives in this paper are to derive a stochastic model for determining the necessary staff levels in various service environments such as hospitals, banks, or in manufacturing settings where the demand for a good staff is erratic. In order to assess the proper staff levels, an equation must be used which involves previous service demand hours that are considered as a distribution integrated within the statistical capability scale distribution. The most important property of this staffing model is that the results are independent of the mix of the jobs, the variation of the in-service time for jobs, and the stationary or nonstationary job arrival rate.

Automated analysis of medical text. II.: Cognitive strategy

Abstract: In a previous paper, the authors described three paradigms applicable to automated medical text analysis. In this paper, the relative importance of the three paradigms is discussed, viz. the relative value of the linguistic word categorization, the semantic paradigm, and the medical fact delineation. The strategy adopted was to limit the linguistic disambiguation and apply probabilistic rules, in order to speed up the analytic process.

Designing an outcome-oriented computer decision-support system for cardiovascular ICU—a preliminary report

Abstract: This paper describes the conceptual framework and preliminary results of an outcome-oriented decision-support system prototype for the cardiovascular intensive care unit (CVICU). The major characteristics of this design include: (1) its problem-based approach to solving clinical problems; (2) an integrated structure with the hospital information system in terms of its data, model and knowledge bases; (3) proposed alternative modes of interaction that include monitoring and critiquing; (4) and research modules that design, manage, and analyze outcome-based clinical studies. At present, an initial prototype has been implemented on a PC as a set of modules accessible from a main menu. The structural framework of the overall system is fairly well defined but only limited quantitative, statistical and expert knowledge has been captured. The second phase of the project involves porting the prototype to a Unix workstation environment, refining and adding models to the model base, expanding its knowledge bases, reasoning capability, and testing the prototype with actual clinical cases in a real-time fashion.

Experience using the CQI model in health care management

Abstract: This article describes practical applications of the Continuous Quality Improvement (CQI) concept applied to management of the Information Services Departments of a health care institution. The article is valuable for two reasons: (1) it is valuable as a successful application of CQI in an award-winning information services division of an award-winning health care institution, and (2) since the work of the information services function is so complex, has major interdependencies with many other functions throughout the entire organization, and is driven by fast-paced change, successful use in that function can be a model for other major functions of the hospital.

Developing a management information system for a hospital: a case study on vendor selection

Abstract: This study examines the decision-making process that a medium-sized hospital took to develop a management information system. Since system developers in this particular hospital were all users without system expertise, a substantial amount of time was devoted to their search for relevant information. Outside consulting, literature review, and a market survey were conducted for the developers to understand the characteristics of systems products on the market. Developers also invited extended user participation in vendor evaluation. The authors analyzed the reasoning process behind the successful selection of a vendor-supported information system. The findings provide hospitals of similar characteristics a decision model to follow in their search for a vendor-supported information system.