Prognostic Pattern: A Hybrid Methodology For Predictivity in Medical Imaging Using VGG-16

At present, deep learning is being used to compute data related to medical images with the purpose of improving disease-related research. Deep learning based medical image analysis includes various important tasks such as medical image reformation, quality enhancement in different types of images, segmentation and classification of images (Computer-Aided Detection) CAD, registration to distinguish between normal cells and tumor cells. This article focuses on the hybrid approach for analyzing medical images that makes use of both clustering and classification via k-mean, VGG, RNN) techniques on different diseases datasets like cardio-vascular diseases, breast cancer and brain disease dataset. This paper utilizes predictive analytics in medical image data which is spread across a combination of clustering and classification. The objective of this study is to demonstrate that integrating these two methods will help improve accuracy as well as save time in diagnosing illnesses from image data by avoiding the process where labeling the data for feature detection takes place. Furthermore, this paper aims at presenting an overview of the methodology employed during investigation along with results emanating from this analysis highlighting how predictive analytics can be applied in healthcare.