iSlide: a 'big picture' interactive teledermatopathology e-learning system.

BACKGROUND: Dermatopathology training is often limited by facilities and a dearth of specialists. Advancements in information and communication technologies have made possible the adoption of innovative learning techniques, especially in places where specialists are lacking. OBJECTIVES: To implement and evaluate the performance of the iSlide system, which is an interactive dermatopathology training platform (http://scope.tmu.edu.tw/islide2/). METHODS: Fifty-two cases representing a variety of dermatopathology conditions and complications were used to set up the iSlide system, and virtual slides of these cases were produced. Medical students from the Dermatology Department of Taipei Medical University were taught to use the system. Performance of the system was evaluated and validated using questionnaires, the first comprising 20 questions and the second a shorter, six-question telephone-based survey on 15 of the 96 interns. Twenty cases prepared by the iSlide system were also presented at an international dermatopathology conference and evaluated by conference participants. RESULTS: Ninety-six students and 72 experts participated in the study. Ninety-two per cent of the students and 98% of the experts found the iSlide system to be a useful tool for learning dermatopathology. Of these, 82% of the students and 63% of the experts felt that iSlide was easy to use. CONCLUSIONS: iSlide is useful for dermatopathology. As only 82% of the student evaluators and 63% of the expert evaluators found the system easy to use, further work has to be done to improve the iSlide interface to make the system more user friendly.

Co-occurrence of second primary malignancy in patients with thyroid cancer.

BACKGROUND: The discovery of asynchronous or synchronous double or multiple malignancies in patients is not uncommon. The co-occurrence of second primary malignancy (SPM) could be randomly occurring or association with risk factors such as environmental, genetic predisposition and therapy-related. MATERIALS AND METHODS: We retrieved ∼782 million claim records consisting of 10.8 million males and 10.6 million females from Taiwan's National Health Insurance, which were collected for a period of 3 years (January 2000-December 2002). All the patient records were stratified by gender and ages at a 20-year interval with SPMs and specific groups. Interestingness or Q-value was used to measure strength of the disease-disease associations. RESULTS: A total of 9423 thyroid cancer (female: 7483, male: 1940), 276 184 SPM (female: 141 023, male: 135 161) and 861 co-occurrence cases (female: 583, male: 278) were recorded. The co-occurrence incidence rate of head and neck, breast, digestive system and lung was 1.93%, 1.59%, 1.44% and 1.18%, respectively. Malignancy of salivary glands, laryngx, sarcoma, lymphoid tissue, mouth, central nervous system and lungs found Q-value >10. Malignancies with intermediate Q-values (5.0-9.9) were observed in nasopharynx, kidney and ureter, breast, stomach and skin. Prostate, leukemia, urinary bladder, ovary, colon, liver and uterine cervix cancer have lower Q-values (1.0-4.9). CONCLUSION: Co-occurrence ratio of thyroid cancer and SPM was high, occurred in all organ systems. We postulated that the aggressive use of modern diagnostic modalities, aggressive radioiodine treatment, pre-existing molecular oncogen mutations, and thyroid hormone for simultaneously supple-mentary and suppressive therapies were responsible.

Electronic health record goes personal world-wide.

OBJECTIVE: Increasing patient demand for convenient access to their own healthcare data has led to more personal use of the Electronic Health Record (EHR). With "consumer empowerment" being an important issue of EHR, we are seeing a more "patient-centric" approach of EHR from countries around the world. Researchers have reported on issues in EHR sharing including concerns on privacy and security, consumer empowerment, competition among providers, and content standards. This study attempts to analyze prior research and to synthesize comprehensive, empirically-based conceptual models of EHR for personal use. METHODS: We use "B2C(2B) " to represent this new behavior of EHR sharing and exchange, with "consumer" in the center stage. RESULTS: Based on different information sharing mechanisms, we summarized the "B2C(2B) " behavior into three models, namely, the Inexpensive data media model, the Internet patient portal model and the Personal portable device model. Models each have their own strengths and weaknesses in their ways to share patient data and to address privacy and security concerns. CONCLUSION: Personal use of EHR under the B2C(2B) model does look promising based on our study. We started to observe a trend that governments around the world are embarking on related projects. With multiple stake-holders involved, we are only beginning to understand the complexity of such undertakings.

Building a generic architecture for medical information exchange among healthcare providers.

Due to the inability to exchange clinical information among hospitals, continuity of care cannot be maintained and a tremendous amount of medical resource has been wasted. This paper describes an architecture that would facilitate exchange of clinical information among heterogeneous hospital information systems. It is dubbed 'Medical Information Exchange Center' or MIEC as part of a six-year Health Information Network Project hosted by the Department of Health. MIEC was designed so that it is innovative yet technically feasible today. It is convenient for authorized users yet secure enough so people can trust and has minimal impact to participated hospitals. Authorized users will be able to access information through two web-based interfaces directed to physician and non-physician users respectively. Hospitals are connected through a virtual private network to exchange patient information and users need to obtain a private key from the certificate authority in order to securely connect to MIEC. A pilot project was conducted to demonstrate the feasibility of this architecture and the problems encountered were discussed.

Consumer health information websites in Taiwan.

Developers of consumer health information (CHI) websites and pages in Taiwan have included: 1) individuals; 2) general hospitals; 3) disease-oriented associations; 4) government health administrative agencies; and 5) commercial corporations. General hospitals have developed the largest number of sites, but the quality of these sites differs widely. The unprecedented impact of health-oriented websites on medical care and the health of the public in general necessitates the establishment of credible instruments for the evaluation and rating of the quality of such websites and pages. In this paper, we describe the current status of CHI websites and pages in Taiwan. Suggestions for improving the content of CHI websites and pages are also proposed.

Neural network modeling for surgical decisions on traumatic brain injury patients.

Computerized medical decision support systems have been a major research topic in recent years. Intelligent computer programs were implemented to aid physicians and other medical professionals in making difficult medical decisions. This report compares three different mathematical models for building a traumatic brain injury (TBI) medical decision support system (MDSS). These models were developed based on a large TBI patient database. This MDSS accepts a set of patient data such as the types of skull fracture, Glasgow Coma Scale (GCS), episode of convulsion and return the chance that a neurosurgeon would recommend an open-skull surgery for this patient. The three mathematical models described in this report including a logistic regression model, a multi-layer perceptron (MLP) neural network and a radial-basis-function (RBF) neural network. From the 12,640 patients selected from the database. A randomly drawn 9480 cases were used as the training group to develop/train our models. The other 3160 cases were in the validation group which we used to evaluate the performance of these models. We used sensitivity, specificity, areas under receiver-operating characteristics (ROC) curve and calibration curves as the indicator of how accurate these models are in predicting a neurosurgeon's decision on open-skull surgery. The results showed that, assuming equal importance of sensitivity and specificity, the logistic regression model had a (sensitivity, specificity) of (73%, 68%), compared to (80%, 80%) from the RBF model and (88%, 80%) from the MLP model. The resultant areas under ROC curve for logistic regression, RBF and MLP neural networks are 0.761, 0.880 and 0.897, respectively (P < 0.05). Among these models, the logistic regression has noticeably poorer calibration. This study demonstrated the feasibility of applying neural networks as the mechanism for TBI decision support systems based on clinical databases. The results also suggest that neural networks may be a better solution for complex, non-linear medical decision support systems than conventional statistical techniques such as logistic regression.

Curriculum for building medical resources on the Internet--experience in Taiwan.

This paper describes our experience with a one-semester course on how to build a medically-related homepage on the Internet, designed for fourth-year medical students. Based on the assumption of limited technical knowledge of the Internet, this class taught students Internet-exploring skills, multimedia authoring and HTML (Hypertext Markup Language) in the first eight-week period. Students were then divided into four-person teams and asked to select a topic for a homepage. These teams were then asked to collect the necessary resources for the development of the homepage both through individual work and, consultation with an advisor. Each group project for building medically-related homepage was accomplished and presented in the remaining four-week period. The resultant projects of these fourth-year medical students were of surprisingly good content and high quality. Medical students rapidly learned to use the software tools, and through proper instruction and provision of equipment, they were able to build significant medical resources on the Internet that can potentially be useful in education, clinical applications and research.