Early Detection of Lung Cancer Using Deep Convolution Neural Networks

The study concentrated on the identification of lung cancer, a disease that affects around 225,000 individuals yearly and requires treatment that costs close to $12 billion in total. Early cancer identification is, therefore, crucial to offer patients the best chance of recovery and survival. This study aims to evaluate the data (CT scan slices) through various pre-processing techniques and then analyze the data using machine learning algorithms, mainly 3D Convolutional Neural Networks, to train and test the model and generate an accurate cancer diagnosis tool. This will be very beneficial in identifying and eliminating cancer cells early on. To forecast if a patient would be diagnosed with lung cancer, this study attempts to create an automated method. The most accurate findings come from experiments that reached 73% accuracy after a comparatively short model learning period.