Lung Cancer Detection Using Image Processing

Conditions known as lung diseases have an impact on the respiratory system and lungs. It is commonly acknowledged that lung cancer is among the top five global causes of death in people. Early diagnosis can increase the chances that a person will survive. Lung cancer patients have a 14% survival rate, but if the disease is caught early, the rate jumps to 49%. Even though computed tomography (CT) is significantly more reliable than X-rays, numerous imaging modalities must still be used to get a complete diagnosis. This article discusses the creation and evaluation of a deep neural network for the detection of lung cancer in CT scans. To determine if a lung image was cancerous or benign, a densely coupled convolutional neural network, or DenseNet, and an adaptive boosting strategy were used. Over 1,000 images of the lung make up the dataset, with 85% being used for training and 15% being used for testing and classification. The trial’s outcomes showed that the suggested approach was successful in achieving an accuracy of 90%.