Enhanced Lung Cancer Detection Using Deep Learning Ensemble Approach

In emerging technologies, deep learning has become the solution for most real-world problems. This is used by integrating with computer science and predicting many diseases for people at an early stage. One of the major issues is the detection of Lung Cancer either Cancerous or Non-Cancerous. Many studies have proposed the solutions to detect the Lung Nodule. However, using an ensemble approach by taking CT(Computed Tomography) Scan images gives accurate results when compared to other methodologies. In this study, we proposed the methodology by combining three deep learning models instead of one. Thereby, the performance of three Convolutional Neural Network models is combined to get accurate results. The dataset used for this is LUNA16 from Grand Challenge which is available online. This dataset consists of CT scans and an annotations file which gives the information about each CT Scan. Our Ensemble 2D CNN model gave an effectiveness of 96% which is greater than the baseline methodology.