Big Data Analytics with Deep Learning based Intracranial Haemorrhage Diagnosis and Classification Model

Abstract

Owing to the rapid growth of medical imaging technologies, medical image analysis is entered the period of big data for proper diagnosis of diseases. Intracranial hemorrhage (ICH) means a pathological disorder that requires quick decision making and diagnosis. Computed tomography (CT) can be accurate and extremely dependable diagnosis method for detecting hemorrhages. Automatic recognition of ICH from CT scans by using a computer-aided diagnosis (CAD) method was helpful in detecting and classifying various grades of ICH. Due to the recent development of deep learning (DL) methods in image processing applications, numerous medical imaging approaches use it. This article introduces innovative Big Data Analytics with Deep Learning based Automated Brain Intracranial Haemorrhage Diagnosis and Classification Model (BDDL-IHDCM). The presented BDDL-IHDCM model examines the computed tomography (CT) scans for detecting and classifying ICH. To accomplish this, the presented BDDL-IHDCM model applies bilateral filtering (BF) technique for noise elimination procedure. The BDDL-IHDCM method employs neural architectural search network (NASNet) feature extractor with Bayesian optimization (BO) based hyperparameter tuning. For ICH detection and classification, a grey wolf optimizer (GWO) with auto encoder (AE) model is utilized in this study. To handle big data, Hadoop MapReduce will be used. To exhibit the significant performance of the presented BDDL-IHDCM method, a comprehensive simulation analysis is made by making use of benchmark dataset. The experimental outcomes indicate the betterment of the presented BDDL-IHDCM method to other DL models with maximum accuracy of 96.22%.