Deep learning and large-scale image datasets in CAD systems have demonstrated rapid and enormous success in a variety of computer challenges. In this study, “ChestX-ray8” is presented, which consists of 108,948 frontal X-ray images of 32,717 patients with 14 disease labels. It is shown that these thoracic disorders can be noticed and spatially placed utilizing a unified weakly – supervised multi label image classification and disease localization framework. The approach taken is an automated approach for segmenting and 3D localizing multiple sclerosis (MS) lesions utilizing multi-modal brain magnetic resonance data provided. The method is based on a deep end-to-end convolutional neural network (CNN) for slice -based segmentation of 3D volumetric data. The architecture follows the original U-Net and enhanced variations. The proposed pipeline is tested on four different datasets: the 2008 MICCAI MS Lesion Segmentation Challenge dataset, that holds 20 figures for preparation and 24 representations for experiment; the Longitudinal Multiple Sclerosis Lesion Segmentation Challenge, that holds 40 concepts for preparation and 42 representations for experiment; and MSSEG 2016 challenge, which contains 15 images for training and 38 images for testing; and dataset provided by the NIH clinical centre containing more than 100,000 anonymized chest X-ray images.