Revolutionizing Lung Cancer Prognosis Through CNN-based Predictive Models
Ankita Sharma, Sonam Mittal
Abstract
Lung cancer is a matter of great concern, and unfortunately, it is the leading cause of mortality worldwide. Doctors are using many kinds of imaging scans like CT scans and surgical biopsies for the detection of lung cancer. In recent years, Deep Learning (DL) algorithms have produced encouraging results in automated prediction and classification tasks for medical imaging. Early prediction can resolve many problems, and the mortality rate can be lower after using various DL models. This research uses two models of DL, Sequential, and DenseNet for the prediction of lung cancer. This article uses chest X-ray scans for image preprocessing and feature selection before the model’s training, testing, and validation. After this process model is implemented with the different libraries and gives authenticated results. The DenseNet model gave the effective results than the sequential model with 95.86% accuracy.