Multi Cancer Prediction using Deep Learning and CNN Algorithm
Siddharth Patel, Zayd Hassan, S. Iniyan, Usha Desai · 2024
Deep learning approach is used to predict the chance of acquiring several types of malignancies, including breast, brain, lung, colon, oral, kidney, and cervical cancer. To find patterns and risk factors linked to each kind of cancer, deep learning algorithms, such as convolutional neural networks and recurrent neural networks, are trained on enormous databases of patient data, including genetic markers, lifestyle variables, and medical history. Deep learning algorithms can accurately forecast an individual’s risk of acquiring cervical, lung, colon, oral, kidney, breast, or brain cancer by evaluating these intricate datasets. Results indicate varied performance across cancer types, with cervical, breast, lung, and colon cancer models exhibiting high accuracy (>98%) and lower validation losses, suggesting strong generalization capabilities. Conversely, kidney and oral cancer models displayed signs of overfitting, emphasizing the challenges of model generalization. These findings highlights the DL’s potential in enhancing cancer detection, early diagnosis, and personalized treatment planning.