Construction of automatic classification model for colorectal polyp diagnosis based on deep learning
Chen Jian1, Zhang Zihao2, Lu Yongda3, Xia Kaijian4, Wang Ganhong5, Liu Luojie1, Xu Xiaodan1
1Department of Gastroenterology, Changshu NO.1 People′s Hospital (Affiliated Changshu Hospital of Soochow University), Changshu 215500, China; 2Shanghai Hao Brothers Educational Technology Co., Ltd., Shanghai 200434, China; 3Department of Gastroenterology, the First Affiliated Hospital of Soochow University, Suzhou 215006, China; 4Changshu Key Laboratory of Medical Artificial Intelligence and Big Data, Changshu 215500, China; 5Department of Gastroenterology, Changshu Traditional Chinese Medicine Hospital (New District Hospital), Changshu 215500, China
Abstract:ObjectiveTo explore the construction of automatic classification model for the diagnosis of colorectal polyp based on deep learning. MethodsFrom January 2018 to January 2023, 957 colonoscopy images were collected at 3 endoscopy centers in Suzhou (537 at Changshu NO.1 People′s Hospital, 359 at Changshu Hospital of Traditional Chinese Medicine, and 61 at the First Affiliated Hospital of Soochow University), by using various image enhanced endoscopy (IEE) techniques. Based on pathological features, these images were classified into normal group, hyperplastic polyps group, and adenomatous polyps group. By using the DenseNet-121, EfficientNet, ResNet101, and ResNet50 convolutional neural network (CNN) frameworks, deep learning models were constructed and tested against the performance of endoscopists with varied experience levels in terms of accuracy, recall rate, precision, F1 score, and time taken to read images. ResultsEfficientNet outperformed the other models, with a 0.961 accuracy, 0.968 recall rate, 0.959 precision, and 0.962 F1 score. In terms of image reading time, all models significantly outperformed endoscopists, completing automatic diagnostic tasks in an average time of (4.08±0.63) seconds compared to the average time of (291.10±17.68) seconds required by endoscopists, showing a statistical difference (t=-36.22, P<0.01). The EfficientNet pretrained model, after transfer learning, was named "EffiPolypNet". It misclassified a few adenomatous polyps but achieved an accuracy of 0.90 and an AUC of 0.98. Visualization using t-distributed stochastic neighbor embedding (t-SNE) revealed semantic feature overlaps between adenomatous and hyperplastic polyps, which could account for the misclassifications. Gradient-weighted class activation mapping (Grad-CAM) and Shapley additive explanations (SHAP) elucidated the key image regions and relative importance of features in the model′s decision-making process. ConclusionThe EffiPolypNet model outperformed other IEE techniques in categorizing the nature of colorectal polyps, offering efficient and dependable support for optical diagnosis in colonoscopy.
陈健 张子豪 卢勇达 夏开建 王甘红 刘罗杰 徐晓丹. 基于深度学习构建结直肠息肉诊断
自动分类模型[J]. 中华诊断学电子杂志, 2024, 12(1): 9-17.
Chen Jian, Zhang Zihao, Lu Yongda, Xia Kaijian, Wang Ganhong, Liu Luojie, Xu Xiaodan. Construction of automatic classification model for colorectal polyp diagnosis based on deep learning. zhzdx, 2024, 12(1): 9-17.
[1]Morris VK,Kennedy EB,Baxter NN,et al.Treatment of metastatic colorectal cancer:ASCO guideline[J].J Clin Oncol,2023,41(3):678-700.DOI:10.1200/JCO.22.01690.
[2]He X,Hang D,Wu K,et al.Long-term risk of colorectal cancer after removal of conventional adenomas and serrated polyps[J].Gastroenterology,2020,158(4):852-861.e4.DOI:10.1053/j.gastro.2019.06.039.
[3]Shaukat A,Kaltenbach T,Dominitz JA,et al.Endoscopic recognition and management strategies for malignant colorectal polyps:recommendations of the US multi-society task force on colorectal cancer[J].Gastroenterology,2020,159(5):1916-1934.e2.DOI:10.1053/j.gastro.2020.08.050.
[4]Tate DJ, Desomer L, Heitman SJ, et al. Clinical implications of decision making in colorectal polypectomy:an international survey of Western endoscopists suggests priorities for change[J].Endosc Int Open,2020,8(3):E445-E455.DOI:10.1055/a-1079-4298.
[5]Byrne MF,Chapados N,Soudan F,et al.Real-time differentiation of adenomatous and hyperplastic diminutive colorectal polyps during analysis of unaltered videos of standard colonoscopy using a deep learning model[J].Gut,2019,68(1):94-100.DOI:10.1136/gutjnl-2017-314547.
[6]Djinbachian R,Marchand E,Pohl H,et al.Optical diagnosis of colorectal polyps:a randomized controlled trial comparing endoscopic image-enhancing modalities[J].Gastrointest Endosc,2021,93(3):712-719.e1.DOI:10.1016/j.gie.2020.11.023.
[7]Sundaram S,Giri S,Jearth V,et al.Image-enhanced endoscopy and endoscopic resection practices in the colon among endoscopists in India[J].Endosc Int Open,2022,10(9):E1181-E1187.DOI:10.1055/a-1914-6197.
[8]Bang CS,Lee JJ,Baik GH.Computer-aided diagnosis of diminutive colorectal polyps in endoscopic images:systematic review and meta-analysis of diagnostic test accuracy[J].J Med Internet Res,2021,23(8):e29682.DOI:10.2196/29682.
[9]Kikutsuji T,Mori Y,Okazaki KI,et al.Explaining reaction coordinates of alanine dipeptide isomerization obtained from deep neural networks using Explainable Artificial Intelligence (XAI)[J].J Chem Phys,2022,156(15):154108.DOI:10.1063/5.0087310.
[10]Zhang Y,Hong D,McClement D,et al.Grad-CAM helps interpret the deep learning models trained to classify multiple sclerosis types using clinical brain magnetic resonance imaging[J].J Neurosci Methods,2021(353):109098.DOI:10.1016/j.jneumeth.2021.109098.
[11]Tan M,Le QV.EfficientNet:rethinking model scaling for convolutional neural networks[J].In Proceedings of the 36th International Conference on Machine Learning,2019(97):6105-6114.DOI:10.48550/arXiv.1905.11946.
[12]Glissen Brown JR, Mansour NM, Wang P, et al. Deep learning computer-aided polyp detection reduces adenoma miss rate:a united states multi-center randomized tandem colonoscopy study (CADeT-CS Trial)[J].Clin Gastroenterol Hepatol,2022,20(7):1499-1507.e4.DOI:10.1016/j.cgh.2021.09.009.
[13]Rondonotti E,Hassan C,Tamanini G,et al.Artificial intelligence-assisted optical diagnosis for the resect-and-discard strategy in clinical practice:the Artificial intelligence BLI Characterization (ABC) study[J].Endoscopy,2023,55(1):14-22.DOI:10.1055/a-1852-0330.
[14]Patel K,Li K,Tao K,et al.A comparative study on polyp classification using convolutional neural networks[J].PLoS One,2020,15(7):e0236452.DOI:10.1371/journal.pone.0236452.
[15]Urban G,Tripathi P,Alkayali T,et al.Deep learning localizes and identifies polyps in real time with 96% accuracy in screening colonoscopy[J].Gastroenterology,2018,155(4):1069-1078.e8.DOI:10.1053/j.gastro.2018.06.037.