Diagnosis of lung and colon cancer based on clinical pathology images using convolutional neural network and CLAHE framework

Abstract

Cancer is a non-contagious disease that is the leading cause of death globally. The most common types of cancer with high mortality are lung and colon cancer. One of the efforts to reduce cases of death is early diagnosis followed by medical therapy. Tissue sampling and clinical pathological examination are the gold standard in cancer diagnosis. However, in some cases, pathological examination of tissue to the cell level requires high accuracy, depending on the contrast of the pathological image, and the experience of the clinician. Therefore, we need an image processing approach combined with artificial intelligence for automatic classification. In this study, a method is proposed for automatic classification of lung and colon cancer based on a deep learning approach. The object of the image that is classified is the histopathological image of normal tissue, benign cancer, and malignant cancer. Convolutional neural network (CNN) with VGG16 architecture and Contrast Limited Adaptive Histogram Equalization (CLAHE) were employed for demonstration of classification on 25000 histopathological images. The simulation results show that the proposed method is able to classify with a maximum accuracy of 98.96%. The system performance using CLAHE shows a higher detection accuracy than without using CLAHE and is consistent for all epoch scenarios. The comparative study shows that the proposed method outperforms some previous studies. With this proposed method, it is hoped that it can help clinicians in diagnosing cancer automatically, with low cost, high accuracy, and fast processing on large datasets.

Colon and lung cancer classification from multi-modal images using resilient and efficient neural network architectures

“… Sethy et al [ 12 ] combined the AlexNet architecture, wavelet transformations, and support vector machines to achieve an accuracy of 99.3 % and an impressive AUC of 0.99 for lung cancer classification on the LC25000 dataset. Hadiyoso et al [ 13 ] employed a CNN with the VGG16 architecture and CLAHE to achieve an accuracy of 98.96 % for lung cancer classification. Rajput and Subasi [ 11 ] used ResNet50 and achieved a remarkable accuracy of 99.8 % for colon cancer classification, although they did not report AUC metrics.…”

“…Hadiyoso et al [ 13 ] employed a CNN with the VGG16 architecture and CLAHE to achieve an accuracy of 98.96 % for lung cancer classification.…”

“…Hadiyoso et al [ 13 ] proposed a deep learning technique for automatically classifying colon and lung cancer. They classified 25,000 histopathology images using the VGG16 architecture and contrast-limited adaptive histogram equalization (CLAHE).…”

“…Sethy et al [ 12 ] combined AlexNet and SVM, achieving an impressive accuracy of 99.3 %. Methods encompassed a wide range, from hybrid networks to image enhancements using VGG16 and CLAHE, as seen in the work of Hadiyoso et al [ 13 ]. Other researchers, such as Iqbal et al [ 15 ], focused on achieving higher F1-scores, while others emphasized transfer learning, as exemplified by the studies conducted by Kumar et al [ 18 ] and Al Ghamdi et al [ 16 ].…”

“…Tested only on one dataset. Hadiyoso et al [ 13 ] VGG16 with CLAHE LC25000 Image enhancement 98.96 Wahid et al [ 14 ] Various CNNs LC25000 Multiple models Up to 99.87 Iqbal et al [ 15 ] ColonNet LC25000 High F1-score 96.31 Al Ghamdi et al [ 16 ] ShuffleNet, DCRNN LC25000 Transfer learning 98.99 Stephen and Sain [ 17 ] Neural Arch Search LC25000 Neural net search 93.91 Kumar et al [ 18 ] Transfer Learning LC25000 Feature extraction 98.60 Hage Chehade et al [ 19 ] Various ML models LC25000 Multiple models Up to 99 Mridha et al [ 20 ] DL Framework LC25000 5 types of cells 98.30 Mehmood et al [ 21 ] Modified AlexNet LC25000 Fine-tuned model 98.40 Provat et al [ 22 ] Custom CNN LC25000 Cyclic learning rates ...…”