基于病理大切片的胃癌T分期人工智能辅助诊断平台的建立
Establishment of Artificial Intelligence-Assisted Diagnosis Platform for T Staging Gastric Cancer Based on Pathological Large Section
DOI: 10.12677/ACM.2023.134801, PDF,   
作者: 张凯明, 刘 磊, 卢嘉琦：青岛大学医学部，山东 青岛；王东升*：青岛大学附属医院胃肠外科，山东 青岛
关键词: 人工智能；病理大切片；胃癌；T分期；卷积神经网络；Artificial Intelligence； Pathological Large Slice； Gastric Cancer； T Stage； Convolutional Neural Network
摘要: 目的：研究基于病理大切片的胃癌T分期人工智能辅助诊断平台建立的价值。当下全球范围内的胃癌总体发病率不断上升，对胃癌病理诊疗工作效率的要求提高。人工智能可以辅助临床医生进行诊疗，而使用病理大切片较使用常规病理小切片拼接观察更为准确。在本研究中，我们建立胃癌病理大切片数字图像深度学习模型，用以辅助病理医师进行胃癌T分期的判断。方法：收集青岛大学附属医院2019年1~12月行胃癌切除术并符合纳入标准的胃癌患者106例的病理标本制成病理大蜡块，经过切片、染色、扫描后共计得到1000张HE染色全玻片扫描图像，按照随机列表法进行分组，其中学习组700张，验证组300张，学习组数据用以建立人工智能模型，得出模型数据，对照验证组数据检测模型的性能。对比各组准确度、敏感度及精确度，统计平均诊断时间和不同分期的AUC (曲线下面积)及ACC (准确率)，应用AUC及ACC对人工智能与病理科高年资医师的诊断结果进行对比，分析其学习水平。结果：学习组的准确度、敏感度及精确度与验证组并无统计差异(P > 0.05)。学习组平均诊断时间更短于验证组(P < 0.05)。T1、T4的AUC、ACC均较T2、T3更低(P < 0.05)。结论：本研究的建立基于病理大切片的胃癌T分期人工智能辅助诊断平台，T分期自动识别准确率较高，且识别时间较短，有较高性能，可以辅助临床医生更加高效的诊疗，值得推广使用。
Abstract: Objective: To study the value of artificial intelligence-assisted diagnosis platform for gastric cancer T staging based on pathological large sections. At present, the overall incidence of gastric cancer worldwide is increasing, and the requirements for the efficiency of pathological diagnosis and treatment of gastric cancer are increasing. Artificial intelligence can assist clinicians in diagnosis and treatment, and the use of pathological large sections is more accurate than the use of conven-tional pathological small sections to splicing observation. In this study, we established a deep learning model of large digital images of gastric cancer pathology to assist pathologists in deter-mining the T stage of gastric cancer. Methods: A total of 106 pathological specimens of gastric cancer patients who underwent gastric cancer resection in the Affiliated Hospital of Qingdao Uni-versity from January to December 2019 and met the inclusion criteria were collected to make pathological wax blocks, and a total of 1000 HE stained whole slides were obtained after sectioning, staining and scanning, which were grouped according to the random list method, including 700 sheets in the learning group and 300 sheets in the verification group, and the data of the learning group were used to establish artificial intelligence models, obtain model data, and test the performance of the model against the data of the verification group. The accuracy, sensitivity and precision of each group were compared, the average diagnosis time and AUC (area under the curve) and ACC (accuracy) of different stages were counted, and AUC and ACC were used to compare the diagnostic results of artificial intelligence and senior doctors in the Department of Pathology, and their learning level was analyzed. Results: There was no statistical difference between the accuracy, sensitivity and precision of the learning group and the validation group (P > 0.05). The average time to diagnosis was shorter in the learning group than in the validation group (P < 0.05). The AUC and ACC of T1 and T4 are lower than those of T2 and T3 (P < 0.05). Conclusion: The artificial intelligence-assisted diagnosis platform for gastric cancer T staging based on pathological large sections established in this study has high accuracy of automatic T stage recognition, short recognition time, and high performance, which can assist clinicians in more efficient diagnosis and treatment, and is worthy of popularization.
文章引用：张凯明, 刘磊, 卢嘉琦, 王东升. 基于病理大切片的胃癌T分期人工智能辅助诊断平台的建立[J]. 临床医学进展, 2023, 13(4): 5672-5679. https://doi.org/10.12677/ACM.2023.134801

参考文献

[1] 王继仙, 桂坤, 陈炳宪, 等. 基于卷积神经网络的病理活检胃癌诊断模型[J]. 协和医学杂志, 2022, 13(4): 597-604.
[2] Ramana, K.S., Chowdappa, K.B., Obulesu, O., et al. (2022) Deep Convolution Neural Networks Learned Image Classification for Early Cancer Detection Using Lightweight. Soft Computing, 26, 5937-5943. [Google Scholar] [CrossRef] 
[3] 周意龙, 卫子然, 蔡清萍, 等. 基于卷积神经网络胃癌分割与T分期算法[J]. 中国医学物理学杂志, 2022, 39(2): 215-223.
[4] Murtaza, G., Shuib, L., Abdul Wahab, A.W., et al. (2020) Ensembled Deep Convolution Neural Network-Based Breast Cancer Classification with Misclassification Re-duction Algorithms. Multimedia Tools and Applications, 79, 18447- 18479. [Google Scholar] [CrossRef] 
[5] (2020) Convolutional Neural Network for Differentiating Gastric Cancer from Gastritis Using Magnified Endoscopy with Narrow Band Imaging. Digestive Diseases and Sciences, 65, 1355-1363. [Google Scholar] [CrossRef] [PubMed]
[6] 韩伟, 秦小金, 魏延, 等. 基于深度学习的智能辅助内镜诊断系统对上消化道早癌诊断价值[J]. 中华消化内镜杂志, 2021, 38(10): 828-835.
[7] Gadde, S., Charkravarthy, A.S.N., Satyanarayana, S., et al. (2022) Automatic Identification of Drug Sensitivity of Cancer Cell with Novel Re-gression-Based Ensemble Convolution Neural Network Model. Soft Computing, 26, 5399- 5408. [Google Scholar] [CrossRef] 
[8] Albalawi, U., Manimurugan, S. and Varatharajan, R. (2022) Classification of Breast Cancer Mammogram Images Using Convolution Neural Network. Concurrency and Computa-tion: Practice and Experience, 34, e5803. [Google Scholar] [CrossRef] 
[9] Lakshminarayanan, A.S., Radhakrishnan, S. and Pandiasankar, G.M. (2019) Diagnosis of Cancer Using Hybrid Clustering and Convolution Neural Network from Breast Thermal Image. Journal of Testing and Evaluation: A Multidisciplinary Forum for Applied Sciences and Engineering, 47, 1-13. [Google Scholar] [CrossRef] 
[10] Kausar, T., Wang, M.J., Idrees, M., et al. (2019) HWDCNN: Mul-ti-Class Recognition in Breast Histopathology with Haar Wavelet Decomposed Image Based Convolution Neural Network. Biocybernetics and Biomedical Engineering, 39, 967-982. [Google Scholar] [CrossRef] 
[11] Davoudi, K. and Thulasiraman, P. (2021) Evolving Convolutional Neural Network Parameters through the Genetic Algorithm for the Breast Cancer Classification Problem. Simulation: Journal of the Society for Computer Simulation, 97, 511-527. [Google Scholar] [CrossRef] [PubMed]
[12] 张训营, 张凯明, 张超, 等. 卷积神经网络在T3/4期胃癌影像学诊断中应用[J]. 青岛大学学报(医学版), 2021, 57(5): 731-735.
[13] Zeng, R.T., Zhang, X., Zheng, C.S., et al. (2021) Decoupling Convolution Network for Charac-terizing the Metastatic Lymph Nodes of breast Cancer Patients. Medical Physics, 48, 3679-3690. [Google Scholar] [CrossRef] [PubMed]
[14] Isunuri, B.V. and Kakarla, J. (2022) Three-Class Brain Tumor Classification from Magnetic Resonance Images Using Separable Convolution Based Neural Network. Concurrency and Computation: Practice and Experience, 34, e6541. [Google Scholar] [CrossRef] 
[15] 吴宏博, 姚幸雨, 曾丽莎, 等. 基于卷积神经网络的人工智能技术在早期胃癌识别中的应用[J]. 第三军医大学学报, 2021, 43(18): 1735-1742.
[16] Ahmad, B., Usama, M., Ahmad, T., et al. (2022) An Ensemble Model of Convolution and Recurrent Neural Network for Skin Disease Classification. Interna-tional Journal of Imaging Systems and Technology, 32, 218-229. [Google Scholar] [CrossRef] 
[17] 张育, 赵轶峰, 苏卓彬, 等. 基于卷积神经网络的胃癌癌前病变图像分类方法[J]. 中国医学物理学杂志, 2022, 39(2): 209-214.
[18] Yurttakal, A.H., Erbay, H., Ikizceli, T., et al. (2020) Detection of Breast Cancer via Deep Convolution Neural Networks Using MRI Images. Multimedia Tools and Appli-cations, 79, 15555-15573. [Google Scholar] [CrossRef] 

为你推荐