基于人工智能预测肿瘤分子特征，从形态学到分子病理的综述

作者

  • 李莉珍 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 吴迪 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 龙紫灵 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 杨成 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 赵毅 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 邹盈盈 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 陆林 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者
  • 许强 遵义医科大学第二附属医院泌尿外科，中国 贵州 遵义 563000 作者
  • 梁娜 遵义医科大学基础医学院组胚教研室，中国 贵州 遵义 563000 作者

关键词:

数字病理, 分子特征预测, 微卫星不稳定（MSI）, 肿瘤突变负荷（TMB）, 精准肿瘤学

摘要

数字病理与深度学习技术的快速发展正在重塑传统肿瘤病理诊断的格局。HE染色切片作为肿瘤组织学诊断与分级的黄金标准，其形态学信息虽能反映肿瘤结构，却难以直接揭示分子层面的异常。然而，近年研究证实，肿瘤的分子改变（如微卫星不稳定、驱动基因突变及肿瘤突变负荷等）可通过影响细胞形态和微环境，在HE切片中留下可被深度学习模型捕捉的细微特征。这一发现催生了“由形态推断分子”的新范式，使人工智能能够从全切片图像中自动提取并预测分子特征，从而减少对额外免疫组化、PCR或高通量测序的依赖。本文系统综述了该领域的发展历程、核心算法与主要应用场景，分析了当前面临的临床验证、数据标准化及模型可解释性等挑战，并展望了其在精准肿瘤学中的广阔应用前景。该范式有望推动病理诊断从形态学向分子病理的深度融合，实现更高效、经济的肿瘤分子分型。

参考

[1] 张百红,岳红云.人工智能和肿瘤研究[J].现代肿瘤医学,2022,30(10):1887-1889.

[2] Kourounis G, Elmahmudi AA, Thomson B, et al. Computer image analysis with artificial intelligence: a practical introduction to convolutional neural networks for medical professionals[J]. Postgrad Med J, 2023, 99(1178): 1287-1294.

[3] 吴宇芝,程彩霞.人工智能在肿瘤分子分型中的应用[J].临床与病理杂志,2024,44(3):419-425.

[4] 王李杰. 人工智能在肿瘤中的应用[J]. 癌症进展,2021,19(21):2184-2184.

[5] Li X, Yang X, Yang X, et al. Machine Learning-Based Pathomics Model to Predict the Prognosis in Clear Cell Renal Cell Carcinoma[J]. Technol Cancer Res Treat, 2024, 23: 15330338241307686-15330338241307701.

[6] Pfahl EL, Pracha NS, Emlemdi MH, et al. Artificial Intelligence in the Diagnosis and Prognostic Stratification of Hepatocellular Carcinoma: Current Evidence, Clinical Applications, and Future Perspectives[J]. Biomedicines, 2026, 14(3): 505-526.

[7] Van Dieren L, Amar JZ, Geurs N, et al. Unveiling the power of convolutional neural networks in melanoma diagnosis[J]. Eur J Dermatol, 2023, 33(5): 495-505.

[8] Alkhathlan L, Saudagar AKJ. Predicting and Classifying Breast Cancer Using Machine Learning[J]. J Comput Biol, 2022, 29(6): 497-514.

[9] Alsallal M, Habeeb MS, Vaghela K, et al. Artificial intelligence in gastric cancer: a systematic review of machine learning and deep learning applications[J]. Abdom Radiol (NY), 2026, 51(4): 1694-1710.

[10] Gupta I, Singh S, Gupta S, et al. Classification of Brain Tumours in MRI Images using a Convolutional Neural Network[J]. Curr Med Imaging, 2024, 20: e270323214998-e270323215006.

[11] Ozcelik N, Kıvrak M, Kotan A, et al. Lung cancer detection based on computed tomography image using convolutional neural networks[J]. Technol Health Care, 2024, 32(3): 1795-1805.

[12] Nawaz K, Zanib A, Shabir I, et al. Skin cancer detection using dermoscopic images with convolutional neural network[J]. Sci Rep, 2025, 15(1): 7252-7282.

[13] Zhang P, Yang Z. Convolutional neural networks: applications, challenges and future prospects in brain tumor research[J]. Front Neurol, 2026, 17: 1759459-1759464.

[14] Alghamdi M, Alsamri J, Almustafa KM, et al. Computer vision assisted deep transfer learning model for accurate grading of renal cell carcinoma from kidney histopathology images[J]. Sci Rep, 2025, 15(1): 36005-36024.

[15] Pan Y, He L, Chen W, et al. The current state of artificial intelligence in endoscopic diagnosis of early esophageal squamous cell carcinoma[J]. Front Oncol, 2023, 13: 1198941-1198952.

[16] Kadir AKMS, Asaduzzaman M, Kundu J, et al. Integrating Genetic Insights and Artificial Intelligence for Enhanced Oral and Maxillofacial Cancer Care[J]. Methods Mol Biol, 2025, 2952: 107-124.

[17] 高留刚,张恒,张宇涛,等.人工智能在肿瘤影像学中的应用进展[J].中国肿瘤外科杂志,2026,18(1):22-30.

[18] Yahaya BS, Osman ND, Karim NKA, et al. Radiomics and deep learning characterisation of liver malignancies in CT images-A systematic review[J]. Comput Biol Med, 2025, 194: 110491-110505.

[19] Wang J, Gao G, Tian C, et al. Next-generation sequencing based deep learning model for prediction of HER2 status and response to HER2-targeted neoadjuvant chemotherapy[J]. J Cancer Res Clin Oncol, 2025, 151(2): 72-82.

[20] Poalelungi DG, Musat CL, Fulga A, et al. Advancing Patient Care: How Artificial Intelligence Is Transforming Healthcare[J]. J Pers Med, 2023, 13(8): 1214-1226.

[21] Schulz S, Jesinghaus M, Foersch S. [Artificial intelligence in diagnostics-a pathology perspective] [J]. Pathologie (Heidelb), 2026, 47(1): 42-47.

[22] Akbarnejad A, Ray N, Barnes PJ, et al. Toward Accurate Deep Learning-Based Prediction of Ki67, ER, PR, and HER2 Status From H&E-Stained Breast Cancer Images[J]. Appl Immunohistochem Mol Morphol, 2025, 33(3): 131-141.

[23] Xu Q, Xu QQ, Shi N, et al. A multitask classification framework based on vision transformer for predicting molecular expressions of glioma[J]. Eur J Radiol, 2022, 157: 110560-110567.

[24] Kai C, Tamori H, Ohtsuka T, et al. Classifying the molecular subtype of breast cancer using vision transformer and convolutional neural network features[J]. Breast Cancer Res Treat, 2025, 210(3): 771-782.

[25] Veljanoski D, Golbaf A, Chawda P, et al. Artificial intelligence pipeline predicts the integrated molecular-morphologic risk score of meningiomas from routine preoperative MRI[J]. Neurooncol Adv, 2026, 8(1): vdag017-vdag027.

[26] Bi Q, Ai C, Meng Q, et al. A multi-modal model integrating MRI habitat and clinicopathology to predict platinum sensitivity in patients with high-grade serous ovarian cancer: a diagnostic study[J]. Int J Surg, 2025, 111(7): 4222-4233.

[27] Li W, Li Y, Gao S, et al. Integrating lipid metabolite analysis with MRI-based transformer and radiomics for early and late stage prediction of oral squamous cell carcinoma[J]. BMC Cancer, 2024, 24(1): 795-806.

[28] Balezo G, Trullo R, Pla Planas A, et al. MIPHEI-ViT: Multiplex immunofluorescence prediction from H&E images using ViT foundation models[J]. Comput Biol Med, 2026, 206: 111564-111577.

[29] Yin Y, Sun Z, Deng X, et al. Integration of deep learning and radiomic features from multiplex immunohistochemistry images for reproducible Multi-Outcome prediction in a Multi-Center study of colorectal cancer[J]. Int J Med Inform, 2026, 214: 106436-106453.

[30] 黄丹,王奕,游庆华,等.基于注意力机制网络的多实例学习框架实现慢性胃炎多项病理指标的自动识别[J].中华病理学杂志,2021,50(10):1116-1121.

[31] Murchan P, Ó Broin P, Baird AM, et al. Deep feature batch correction using ComBat for machine learning applications in computational pathology[J]. J Pathol Inform, 2024, 15: 100396-100407.

[32] Hezi H, Gelber M, Balabanov A, et al. CIMIL-CRC: A clinically-informed multiple instance learning framework for patient-level colorectal cancer molecular subtypes classification from H&E stained images[J]. Comput Methods Programs Biomed, 2025, 259: 108513-108522.

[33] Xu H, Wang M, Shi D, et al. When multiple instance learning meets foundation models: Advancing histological whole slide image analysis[J]. Med Image Anal, 2025, 101: 103456-103476.

[34] 钟海勤,赵程,雷柏英,等.基于聚类的多实例学习全视野数字切片分类[J].中国生物医学工程学报,2024,43(6):652-661.

[35] Lingasamy P, Ostrowska-Leśko M, Tsakalis P, et al. Deep learning-based prediction of gene expression from histopathology identifies NR5A1 as a candidate biomarker and druggable target in high-grade serous ovarian carcinoma[J]. J Ovarian Res, 2026, 19(1): 210-227.

[36] Nap R, Aburidi M, Marcia R. Contrastive Pre-Training and Multiple Instance Learning for Predicting Tumor Microsatellite Instability[J]. Annu Int Conf IEEE Eng Med Biol Soc, 2024: 1-7.

[37] Pan L, Luo J, Nie C, et al. A Deep Learning Framework Integrating Tumor Microenvironmental Features Accurately Predicts Multiple Driver Gene Mutations in Lung Cancer Pathology Images[J]. Cancer Res, 2026, 86(5): 1319-1337.

[38] 伍俊毅,潘志方.基于双尺度深度学习的肝细胞癌常规病理图像基因突变预测[J].肝胆胰外科杂志,2026,38(5):338-345.

[39] Hu Y, Sirinukunwattana K, Li B, et al. Self-interactive learning: Fusion and evolution of multi-scale histomorphology features for molecular traits prediction in computational pathology[J]. Med Image Anal, 2025, 101: 103437-103453.

[40] Ghaffari Laleh N, Muti HS, Loeffler CML, et al. Benchmarking weakly-supervised deep learning pipelines for whole slide classification in computational pathology[J]. Med Image Anal, 2022, 79: 102474-102490.

[41] Miao J, Liu Q, Zhao J, et al. Multimodal artificial intelligence predicts PIK3CA mutation in breast cancer from digital pathology and clinical a multicenter study[J]. Cancer Biol Med, 2026, 23(3): 430-450.

[42] Li J, Zhang Y, Shu W, et al. M4: Multi-proxy multi-gate mixture of experts network for multiple instance learning in histopathology image analysis[J]. Med Image Anal, 2025, 103: 103561-103572.

[43] Zheng Q, Wang X, Yang R, et al. Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self-supervised deep learning[J]. Cancer Med, 2024, 13(16): e70112-e70123.

[44] 刘阳,蹇顺海.人工智能在结直肠癌分子病理诊断中的研究进展[J].诊断病理学杂志,2026,33(5):752-757,762.

[45] Da Z, Yang H, Zhaxi B, et al. Multiple instance learning-based prediction of programmed death-ligand 1 (PD-L1) expression from hematoxylin and eosin (H&E)-stained histopathological images in breast cancer[J]. PeerJ, 2025, 13: e19201-e19224.

[46] Jiao F, Shang Z, Lu H, et al. A weakly supervised deep learning framework for automated PD-L1 expression analysis in lung cancer[J]. Front Immunol, 2025, 16: 1540087-1540097.

[47] Xia P, Chen D, An H, et al. Learnable prototype-guided multiple instance learning for detecting tertiary lymphoid structures in multi-cancer whole-slide pathological images[J]. Med Image Anal, 2025, 104: 103652-103678.

[48] Zhou Y, Shu X, Wang F, et al. Prediction of neoadjuvant therapy response in breast cancer based on interpretable artificial intelligence[J]. Int J Surg, 2026, 112(1): 1066-1080.

[49] Zeng Q, Klein C, Caruso S, et al. Artificial intelligence predicts immune and inflammatory gene signatures directly from hepatocellular carcinoma histology[J]. J Hepatol, 2022, 77(1): 116-127.

[50] Kawka M, Dawidziuk A, Jiao LR, et al. Artificial intelligence in the detection, characterisation and prediction of hepatocellular carcinoma: a narrative review[J]. Transl Gastroenterol Hepatol, 2022, 7: 41-55.

[51] Hosseini SMM, Hannetel P, Di Cataldo S, et al. Uncertainty-aware AI for tumor subtyping with histology and immunohistochemistry: A multi-center study in Renal Cell Carcinoma[J]. Comput Methods Programs Biomed, 2026, 282: 109373-109387.

[52] Calderaro J, Seraphin TP, Luedde T, et al. Artificial intelligence for the prevention and clinical management of hepatocellular carcinoma[J]. J Hepatol, 2022, 76(6): 1348-1361.

[53] Sinha T, Khan A, Awan M, et al. Artificial Intelligence and Machine Learning in Predicting the Response to Immunotherapy in Non-small Cell Lung Carcinoma: A Systematic Review[J]. Cureus, 2024, 16(5): e61220-e61229.

[54] Li K, Qian Z, Mao J, et al. Applications of artificial intelligence in postoperative surveillance and management of esophageal squamous cell carcinoma[J]. Front Artif Intell, 2026, 9: 1806691-1806703.

[55] Kumar N, Kumari S, Sultana S, et al. Early detection in oral cancer: Are we prepared for artificial intelligence-driven precision medicine[J]? Ann Med Surg (Lond), 2026, 88(1): 1046-1047.

[56] Shaktah LA, Carrero ZI, Hewitt KJ, et al. Application of artificial intelligence and digital tools in cancer pathology[J]. Lancet Digit Health, 2025, 7(10): 100933-100939.

[57] Prelaj A, Miskovic V, Zanitti M, et al. Artificial intelligence for predictive biomarker discovery in immuno-oncology: a systematic review[J]. Ann Oncol, 2024, 35(1): 29-65.

[58] Zhang X, Yang L, Liu C, et al. An Artificial Intelligence Pipeline for Hepatocellular Carcinoma: From Data to Treatment Recommendations[J]. Int J Gen Med, 18: 3581-3595.

[59] Baughan N, Douglas L, Giger ML. Past, Present, and Future of Machine Learning and Artificial Intelligence for Breast Cancer Screening[J]. J Breast Imaging, 2022, 4(5): 451-459.

[60] Mansur A, Vrionis A, Charles JP, et al. The Role of Artificial Intelligence in the Detection and Implementation of Biomarkers for Hepatocellular Carcinoma: Outlook and Opportunities[J]. Cancers (Basel), 2023, 15(11): 2928-2944.

下載

已出版

2026-08-12

期次

Volume 1, Issue 1, 2026

栏目

文章