Deep Ensemble 3D CNN Framework for Lung Cancer Detection using CT-Images

Deep learning has become useful for medical research in recent years. The use of computer science to detect and predict human diseases has been extensive. This study uses Deep Learning’s Convolutional-Neural-Network to identify a potentially malignant lung carcinoma (LC) from many CT scans. This research uses ensemble methods to detect LC. Instead of using a single Deep Learning model, we combined CNNs for improved performance and prediction. This study uses the public LUNA 16 Grand challenge dataset. The dataset’s foundation is an annotated CT scan, which illuminates its contents. Deep learning (DL) uses artificial neural networks because they work like human brain neurons. To train the DL model, several CT scans are collected. CNNs learn to identify cancerous and healthy pictures from the dataset. Our Deep Ensemble 3D CNN is trained utilizing validation and testing datasets. Deep Ensemble 3D CNNs have three CNNs with different kernels, layers, and pooling algorithms. Our Deep Ensemble 3D CNN outperformed the baseline with 95% accuracy.