Confidence Calibration of a Medical Imaging Classification System That is Robust to Label Noise.

A classification model is calibrated if its predicted probabilities of outcomes reflect their accuracy. Calibrating neural networks is critical in medical analysis applications where clinical decisions rely upon the predicted probabilities. Most calibration procedures, such as temperature scaling, operate as a post processing step by using holdout validation data. In practice, it is difficult to collect medical image data with correct labels due to the complexity of the medical data and the considerable variability across experts. This study presents a network calibration procedure that is robust to label noise. We draw on the fact that the confusion matrix of the noisy labels can be expressed as the matrix product between the confusion matrix of the clean labels and the label noises. The method is based on estimating the noise level as part of a noise-robust training method. The noise level is then used to estimate the network accuracy required by the calibration procedure. We show that despite the unreliable labels, we can still achieve calibration results that are on a par with the results of a calibration procedure using data with reliable labels.

ADHD and dysgraphia: underlying mechanisms.

Multiple complaints in the domain of writing are common among children with Attention Deficit Hyperactivity Disorder (ADHD). In this work we sought to characterize the writing disorder by studying dysgraphia in twenty 6th grade boys with ADHD and normal reading skills matched to 20 healthy boys who served as a comparison group. Dysgraphia, defined as deficits in spelling and handwriting, was assessed according to neuropsychological explanatory processes within 3 primary domains: linguistic processing, motor programming and motor kinematics. Children with ADHD made significantly more spelling errors, but showed a unique pattern introducing letter insertions, substitutions, transpositions and omissions. This error type, also known as graphemic buffer errors, can be explained by impaired attention aspects needed for motor planning. Kinematic manifestations of writing deficits were fast, inaccurate and an inefficient written product accompanied by higher levels of axial pen pressure. These results suggest that the spelling errors and writing deficits seen in children with ADHD and normal reading skills stem primarily from non-linguistic deficits, while linguistic factors play a secondary role. Recommendations for remediation include educational interventions, use of word processing and judicious use of psychostimulants.