Political Optimizer with Deep Learning Based Diagnosis for Intracranial Hemorrhage Detection

Intracranial haemorrhage (ICH) detection is a critical task in radiology and neurology, as timely recognition of haemorrhages in the brain can assist in rapid intervention and treatment. Several imaging modalities, including computed tomography (CT) and magnetic resonance imaging (MRI), are widely utilized to detect and classify ICH. Traditional methods for intracranial haemorrhage detection relied on manual inspection of CT images by radiologists. However, with advancements in machine learning (ML) and deep learning (DL) techniques, automated and computer-aided systems have been developed to assist radiologists in detecting and diagnosing ICH efficiently. DL models, particularly convolutional neural network (CNN), has shown promising results in ICH detection on CT images. With this motivation, this study focuses on the development of a Political Optimizer with Deep Learning based Intracranial Haemorrhage Diagnosis on Healthcare Management (PODL-ICHDHM) technique. The presented PODL-ICHDHM technique majorly concentrates on the recognition and classification of ICH on CT images. In this study, bilateral filtering (BF) is initially applied to pre-process the CT images. For feature extraction purposes, the Faster SqueezeNet approach is utilized in this study. At last, the PO algorithm with denoising autoencoder (DAE) model is utilized for the classification of ICH accurately. The experimental result analysis of the PODL-ICHDHM approach was validated on a benchmark dataset. The outcomes emphasized the improved performance of the PODL-ICHDHM algorithm over other recent approaches with a maximum detection accuracy of 98.43%.