Deep Learning-Based Detection Of Lung Nodules From Ct Scans

Abstract

Patients with lung cancer frequently have a bad prognosis if the disease is not detected early.  In order to visualize interior organs and tissues and identify illnesses, medical imaging is crucial.  X-rays, CT scans, and PET scans are commonly used in lung cancer screening; CT scans are preferred since they offer detailed information on lung cancers.  Medical imaging is a complicated and time-consuming method of detecting lung cancer. The increasing number of cancer patients, exacerbated by the COVID pandemic, has placed a significant strain on radiologists, who are already in short supply. An Indian report indicates that there is only one radiologist available for every one hundred thousand patients. Consequently, there is a pressing need for automated diagnostic techniques to aid physicians in promptly analysing patients and making swift treatment decisions. Semantic segmentation is crucial for evaluating the disease or identifying any abnormalities in bodily organs. Initially, traditional approaches relied on basic image processing or machine learning techniques, which necessitated extensive manual pre-processing. However, the advent of deep learning has provided a robust and automated solution. Recent advancements in deep learning are pivotal in enhancing the analysis of medical images, facilitating more efficient disease diagnosis, alleviating the workload on radiologists, and supporting physicians in making precise treatment choices.