Brain tumour detection & segmentation on head CT scans

Abstract

Medical detection and segmentation serve as pivotal components within clinical practices that facilitate healthcare professionals to provide high-quality patient care, such as establishing conclusive diagnoses to devising treatment plans. Consequently, the automation of these processes via machine and deep learning had a significant rise in such learning models, empowering the development and advancement of such technology in medical images to enhance clinical efficacy. These advancements have undergone extensive testing, primarily focused on evaluating their efficacy within the context of the abdominal cavity. The objective of this study is to assess these medical models and conduct a comparative analysis of them with their counterparts in their detection and segmentation prowess within the framework of the brain. These scans pose noteworthy challenges such as limited contrast resolution, neurological variability, and subtle pathologies, all set against the backdrop of the complex anatomy of the brain. These factors collectively contribute to a rigorous testing ground for assessing the efficacy of these models.