Lung Cancer Diagnosis Using Artificial Intelligence: A Radiomics-Based Approach

Abstract

Cancer is one of the most serious health threats we face today, with lung cancer standing out as the globe. If it's not
caught early, the chances of survival for lung cancer patients can be quite low. Medical imaging spotting diseases by
allowing us to look inside organs and tissues. When it comes to lung cancer screening but CT scans are often the goto
choose because they provide detailed insights into tumors within the lungs. However, identifying lung cancer
through medical images can be a tough and time-consuming task. Radiologists are under immense pressure, especially
with the rising number of cancer cases and the impact of the COVID pandemic. To make matters worse, there aren’t
enough radiologists to meet the demand—an Indian report highlights that there’s only one radiologist for every
hundred thousand patients. This is why we need automatic diagnosis techniques to help doctors analyze patient data
quickly and make timely treatment decisions. We’ve tested our proposed deep learning algorithms on both publicly
available datasets and those from local Indian hospitals. Additionally, we developed the Indian lung nodule dataset
for nodule segmentation and the lung tumor dataset to predict how patients will respond to treatment, all in
collaboration with medical professionals.