Abstract: A robust segmentation technique based on multi-layer classification technique to identify the lesion boundary is described. The inventors have discovered a technique based on training several classifiers such that to classify each pixel as lesion versus normal Each classifier is trained on a specific range of image resolutions. Then, for a new test image, the trained classifiers are applied on the image. Then by fusing the prediction results in pixel level a probability map is generated. In the next step, a thresholding method is applied to convert the probability map to a binary mask, which determines a mole border.

Abstract: Training a machine to provide specialized health care apparatus may include receiving text describing a user's health condition via a user interface. Text may be converted into corresponding medical terms. A database may be searched for a list of health care providers treating health conditions associated with the medical terms. A machine learning model may be built that may include user preference for a predefined set of features associated with the user's health condition and health care provider preference for the predefined set of features in treating the user's health condition. The machine learning model may predict one or more of the health care providers that provide treatment for the user's health condition that matches the user's preference. The machine learning model may be retrained based on one or more of feedback from the user, the health care providers, and updated traits of the users and the health care providers.

Abstract: A method for a cloud-based feedback-driven image training and recognition includes receiving a set of expert annotations of a plurality of training images of a predetermined subject matter, wherein the expert annotations include a clinical diagnosis for each image or region of interest in an image, training one or more classification models from the set of expert annotations, testing the one or more classification models on a plurality of test images that are different from the training images, wherein each classification model yields a clinical diagnosis for each image and a confidence score for that diagnosis, and receiving expert classification result feedback regarding the clinical diagnosis for each image and a confidence score yielded by each classification model.