Detection and Classification of Ovarian Carcinoma Using Deep Learning Techniques

Abstract

Cancer is a disease that originates when cells in the human body divide at a newlinefaster pace than usual. These anomalous cells combine to develop into a lump or tumour. newlineCancerous tumours infect the neighboring tissue structures and can escalate to other newlineregions of the body, forming new tumours. Such tumours are known as malignant newlinetumours. Malignancy instils a widespread fear among the human population due to its newlinelethal nature. Among women, the most reported malignancy is ovarian cancer. Ovarian newlinecancer is a malignant growth that develops in the ovaries and spreads uncontrollably. The newlinehigh death rate of ovarian cancer is attributed to a lack of early-stage diagnostic newlinetechniques, as the disease is less curable at advanced stages. The global prognosis of newlineovarian cancer patients tends to be low due to poor clinical outcomes, mostly due to a newlinelack of adequate screening procedures at the early stage. In this regard, an expedited, newlineenergy-efficient, and automated bio-medical system is required to facilitate early newlinedetection of ovarian cancer. newlineAI is gaining traction in smart healthcare as it could be used as an early newlinescreening test for a wide range of diseases, including cancer. With the advancement of newlinedeep learning, AI has gained appeal in precise and accurate disease detection due to its newlinecapacity to mimic the human brain. Deep learning based computer-aided-diagnostic newlinesystems are the current cutting-edge approaches developed to aid radiologists in image newlineanalysis and as a means of double-checking. This work is one such approach to detect newlineand classify ovarian tumours from computed tomography images at an early stage using newlinea combination of efficient deep learning models newline