A Novel Deep Learning Model Based Lung Cancer Detection of Histopathological Images

Lung cancer is still one of the world's top causes of death, and increasing survival rates require early identification. Biopsied tissue must be histopathologically analyzed for diagnosis; however, manual inspection can be laborious and error-prone, increasing the possibility of a mistake and inappropriate therapy. This study examines how well several deep learning models—EfficientNetB0, ResNet50, VGG19, VGG16, and InceptionV3—classify lung tissue into three categories: benign, malignant, and normal. InceptionV3 (81%), VGG19 (94%), VGG16 (90<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>), ResNet50 (95<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>), and the EfficientNetB0 model (98<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>) all had the highest accuracy. Through highly accurate cancer categorization automation, these models promise to improve diagnostic efficiency, reduce pathologists' workload, and minimize errors. Developing dependable computer-aided diagnostic systems is supported by the comparative assessment of various architectures, which guarantees optimal performance. To improve patient outcomes, this work offers insightful information for future developments in deep learning systems for lung cancer detection.