Biomedical subjects
V West
Publications and source records attributed to V West.
Model selection for a medical diagnostic decision support system: a breast cancer detection case.
There are a number of different quantitative models that can be used in a medical diagnostic decision support system (MDSS) including parametric methods (linear discriminant analysis or logistic regression), non-parametric models (K nearest neighbor, or kernel density) and several neural network models. The complexity of the diagnostic task is thought to be one of the prime determinants of model selection. Unfortunately, there is no theory available to guide model selection. Practitioners are left to either choose a favorite model or to test a small subset using cross validation methods. This paper illustrates the use of a self-organizing map (SOM) to guide model selection for a breast cancer MDSS. The topological ordering properties of the SOM are used to define targets for an ideal accuracy level similar to a Bayes optimal level. These targets can then be used in model selection, variable reduction, parameter determination, and to assess the adequacy of the clinical measurement system. These ideas are applied to a successful model selection for a real-world breast cancer database. Diagnostic accuracy results are reported for individual models, for ensembles of neural networks, and for stacked predictors.
Improving diagnostic accuracy using a hierarchical neural network to model decision subtasks.
A number of quantitative models including linear discriminant analysis, logistic regression, k nearest neighbor, kernel density, recursive partitioning, and neural networks are being used in medical diagnostic support systems to assist human decision-makers in disease diagnosis. This research investigates the decision accuracy of neural network models for the differential diagnosis of six erythematous-squamous diseases. Conditions where a hierarchical neural network model can increase diagnostic accuracy by partitioning the decision domain into subtasks that are easier to learn are specifically addressed. Self-organizing maps (SOM) are used to portray the 34 feature variables in a two dimensional plot that maintains topological ordering. The SOM identifies five inconsistent cases that are likely sources of error for the quantitative decision models; the lower bound for the diagnostic decision error based on five errors is 0.0140. The traditional application of the quantitative models cited above results in diagnostic error levels substantially greater than this target level. A two-stage hierarchical neural network is designed by combining a multilayer perceptron first stage and a mixture-of-experts second stage. The second stage mixture-of-experts neural network learns a subtask of the diagnostic decision, the discrimination between seborrheic dermatitis and pityriasis rosea. The diagnostic accuracy of the two stage neural network approaches the target performance established from the SOM with an error rate of 0.0159.
The rotation of maxillary first molars, mandibular first molars, and maxillary first premolars in acceptable occlusions.
The rotation of the maxillary molars is considered important in the orthodontic treatment of malocclusions. In this study, a computer analysis program was developed to examine the rotations of maxillary molars, mandibular molars, and maxillary first premolars in casts of permanent dentitions with acceptable occlusions. Ninety-three sets of untreated 'acceptable occlusion' models from the collection of the Foundation for Orthodontic Research (FOR) were scanned on a flat bed scanner. The images were analysed using custom software. Measurements were made by relating maxillary first permanent molars to the midline, archform, opposite canine, and mandibular first permanent molars. The mandibular first molars and maxillary first premolars were also analysed and their rotations measured. The mean rotations of the maxillary first molars, measured as the angle between a line joining the tips of the buccal cusps and a line tangent to the appropriate archwire form (from Ricketts' Pentamorphic Arches) at the first molars, were 0.59 and -0.72 degree (positive values represent mesio-lingual rotations) for the right and left, respectively. For the mandibular molars, these means were 6.34 and 8.40 degrees, respectively. The mean differences in rotation between buccal cusp tips of maxillary and mandibular first molars in occlusion were 5.75 and 9.12 degrees for the right and left, respectively, with the mandibular being more mesio-lingually rotated. The differences between left and right were significant for all measurements. The present study brings into question the suitability of our present "straight wire" prescriptions in producing similar occlusions. It also suggests that scanning models for computer analysis may be a practical and precise way to measure similar rotations in untreated normal and treated occlusions.
Three professors: a crucial coincidence. Interview by Brian Lee.
Explore the source record for details and available documents.
The 'functional matrix' hypothesis--current concepts and conflicts.
Explore the source record for details and available documents.
Exaggeration of defective platelet function by methyldopa in hypothyroidism.
Platelet function was analyzed in two hypothyroid patients who were taking methyldopa. Results from these two patients are compared to those obtained in seven others with hypothyroidism on no medications. Bleeding times in the two drug patients were 33 and 26 minutes compared to normal values (4.0 to 8.0 minutes) in the other hypothyroid individuals. In addition, platelet aggregation responses to epinephrine, collagen, and ristocetin were abnormal in these two patients; these were normal in the others with hypothyroidism. Repeat studies were done in the first drug patient when she was euthyroid and still receiving methyldopa. Her bleeding time decreased to 11 minutes, platelet retention increased from 31% to 75%, and abnormal aggregation responses corrected to normal. Factor VIII coagulant activity, VIII antigenic activity, VIII ristocetin cofactor activity, and two-dimensional immunoelectrophoresis of VIII antigen were normal. These findings suggest that methyldopa can potentiate the mild platelet function defect seen in patients with hypothyroidism.