Initial evaluation of a continuous speech recognition program for radiology

The aims of this work were to measure the accuracy of one continuous speech recognition product and dependence on the speaker's gender and status as a native or nonnative English speaker, and evaluate the product's potential for routine use in transcribing radiology reports. IBM MedSpeak/Radiology software, version 1.1 was evaluated by 6 speakers. Two were nonnative English speakers, and 3 were men. Each speaker dictated a set of 12 reports. The reports included neurologic and body imaging examinations performed with 6 different modalities. The dictated and original report texts were compared, and error rates for overall, significant, and subtle significant errors were computed. Error rate dependence on modality, native English speaker status, and gender were evaluated by performing ttests. The overall error rate was 10.3 +/- 3.3%. No difference in accuracy between men and women was found; however, significant differences were seen for overall and significant errors when comparing native and nonnative English speakers (P = .009 and P = .008, respectively). The speech recognition software is approximately 90% accurate, and while practical implementation issues (rather than accuracy) currently limit routine use of this product throughout a radiology practice, application in niche areas such as the emergency room currently is being pursued. This methodology provides a convenient way to compare the initial accuracy of different speech recognition products, and changes in accuracy over time, in a detailed and sensitive manner.     

Use of Computed Tomography Scans for Cochlear Implants

While 3-dimensional (3D) imaging by computed tomography has long been desirable for research and treatment of cochlear-implant patients, technical challenges have limited its wide application. Recent developments in scanner hardware and image processing techniques now allow image quality improvements that make clinical applications feasible. Validation experiments were performed to characterize a new methodology and its imaging performance.     

Hemodynamics and blood flow distribution following prolonged circulation at 5 degrees C.

Circulation was maintained in profoundly hypothermic dogs for 8 h at temperatures below 10 degrees C. During cooling to 5 degrees C cardiac output and blood pressure decreased and peripheral resistance rose. Thereafter, circulatory dynamics remained relatively stable over the next 6 h with a gradual decline in blood pressure. The proportion of blood flow to the heart and brain increased with cooling and remained elevated throughout the hypothermia period. Despite continued circulation, pulmonary edema developed after 5-7 h and the dogs were nonviable when rewarmed.     

The adaptation of the plaque reduction assay for measuring specific cytotoxic cells in limiting dilution cultures

The use of the original haemolytic plaque reduction technique to measure cytotoxic T lymphocytes (CTL) has been developed further as a rapid screening assay, particularly suitable for limiting dilution analyses. Using hybridoma cells as targets, the cytotoxicity has been measured by the loss of haemolytic plaque formation and by the reduction of the amount of haemolytic monoclonal antibody secreted from viable target cells into the assay supernatants. The assessment of large numbers of cytotoxic samples has been greatly facilitated by quantitating the amount of haemoglobin released in the assay with an automated microELISA multiscanner and by scoring visually using a modification of the spot test. Using these new techniques, relatively high frequency estimates of cytotoxic cell precursors in an allogeneic response (1 in 462 spleen cells) and an anti-fluorescein response (1 in 3970 spleen cells) were obtained.     

An immunocytochemical study of malignant melanoma and its differential diagnosis from other malignant tumours.

A series of 41 fresh and 36 routinely processed malignant melanomas were immunostained with a panel of 12 monoclonal antibodies reactive against a range of epithelial, lymphoid, and melanoma associated antigens. The aim of the study was to determine whether this panel of antibodies would be useful in diagnostically difficult cases for differentiating melanomas from other tumours, particularly carcinomas and lymphomas. The results confirmed that most unequivocal malignant melanomas can be identified by positivity for S100 protein and for the antigen recognised by antibody NK1/C3, and by negativity for epithelial and lymphoid antigens. The incidence of melanomas expressing cytokeratin antigens was higher, however, particularly in cryostat sections than has previously been reported. It is therefore suggested that a panel of antibodies with more than one marker in each category should be used for identifying melanomas in clinical practice.     

Preventing alcohol problems: survey of Canadian medical schools.

In preparation for a national conference on medical education in the prevention of alcohol problems, a survey of conference participants was conducted. Participants were undergraduate and postgraduate representatives from each Canadian medical school and representatives from 11 provincial and territorial alcohol and other drug agencies. There was agreement that physicians and medical schools have important roles in prevention and treatment of alcohol problems, with "traditional" medical roles seen as the most important. Current training is variable and was seen as inadequate, with more time devoted to treatment than prevention. To correct this situation, renewed priorities and faculty leadership are needed. Respondents felt that there should be uniform standards for assessing undergraduate students' skills in dealing with alcohol problems. Provincial alcohol and other drug agencies are underused in medical education in the prevention and treatment of alcohol problems.     

What do stochastic frontier cost functions tell us about inefficiency?

Recent studies using the stochastic frontier approach have suggested high levels of inefficiency in nursing home and hospital facilities. This paper makes two observations. First, one could falsely estimate high levels of inefficiency if a statistical assumption--zero skewness of the random component of the cost residual--is violated. Second, an inefficient industry may be difficult to distinguish either statistically or visually from an industry free of inefficiency. If available, panel data yields more robust estimates of cost differences among nursing homes or hospitals.