A Comprehensive AI-Based Approach in Classifying Breast Lesions: Focusing on Improving Pathologists' Accuracy and Efficiency.

IF 2.9 3区医学 Q2 ONCOLOGY

Clinical breast cancer Pub Date : 2025-03-26 DOI:10.1016/j.clbc.2025.03.016

Maryam Tahir, Yan Hu, Himani Kumar, Nada Shaker, David Kellough, Shaya Goodman, Manuela Vecsler, Giovanni Lujan, Wendy L Frankel, Anil V Parwani, Zaibo Li

{"title":"A Comprehensive AI-Based Approach in Classifying Breast Lesions: Focusing on Improving Pathologists' Accuracy and Efficiency.","authors":"Maryam Tahir, Yan Hu, Himani Kumar, Nada Shaker, David Kellough, Shaya Goodman, Manuela Vecsler, Giovanni Lujan, Wendy L Frankel, Anil V Parwani, Zaibo Li","doi":"10.1016/j.clbc.2025.03.016","DOIUrl":null,"url":null,"abstract":"Background: Accurate classification of breast lesions is essential for effective clinical decision-making and patient management. In this study, we evaluated an artificial intelligence (AI) solution to classify whole slide images (WSIs) of breast lesions.Methods: We analyzed a cohort of 104 breast cases, including 20 invasive carcinomas, 4 microinvasive carcinomas, 15 ductal Carcinoma in situ (DCIS) cases, and 65 lobular neoplasia/benign cases. The AI's performance was compared with the ground truth established by breast pathologists.Results: For invasive carcinoma, it achieved an area under the curve (AUC) of 0.976, with sensitivity and specificity of 91.7% (84.4%, 95.4%) and 95% (88.0% 97.3%) respectively. For DCIS, the AUC was 0.976, with sensitivity and specificity of 93.3% and 96.6%. For lobular neoplasm, it achieved an AUC of 0.953, sensitivity of 94.1%, and specificity of 95.8%. The AI also performed well in detecting microcalcifications, with an AUC of 0.925 and sensitivity of 95%. Pathologists' diagnostic accuracy improved from 97.1% to 100% with AI support (303 vs. 312 accurate case reads per arm). Additionally, the AI use significantly enhanced the pathologists' efficiency, reducing their review time by an average of 16.5% across the 3 pathologists and led to a 33% reduction in immunohistochemistry usage.Conclusion: This study highlights the potential of AI in breast lesion classification, demonstrating high sensitivity, specificity, and efficiency, and supports its integration into routine pathology practice.","PeriodicalId":10197,"journal":{"name":"Clinical breast cancer","volume":" ","pages":""},"PeriodicalIF":2.9000,"publicationDate":"2025-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Clinical breast cancer","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1016/j.clbc.2025.03.016","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ONCOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Accurate classification of breast lesions is essential for effective clinical decision-making and patient management. In this study, we evaluated an artificial intelligence (AI) solution to classify whole slide images (WSIs) of breast lesions.

Methods: We analyzed a cohort of 104 breast cases, including 20 invasive carcinomas, 4 microinvasive carcinomas, 15 ductal Carcinoma in situ (DCIS) cases, and 65 lobular neoplasia/benign cases. The AI's performance was compared with the ground truth established by breast pathologists.

Results: For invasive carcinoma, it achieved an area under the curve (AUC) of 0.976, with sensitivity and specificity of 91.7% (84.4%, 95.4%) and 95% (88.0% 97.3%) respectively. For DCIS, the AUC was 0.976, with sensitivity and specificity of 93.3% and 96.6%. For lobular neoplasm, it achieved an AUC of 0.953, sensitivity of 94.1%, and specificity of 95.8%. The AI also performed well in detecting microcalcifications, with an AUC of 0.925 and sensitivity of 95%. Pathologists' diagnostic accuracy improved from 97.1% to 100% with AI support (303 vs. 312 accurate case reads per arm). Additionally, the AI use significantly enhanced the pathologists' efficiency, reducing their review time by an average of 16.5% across the 3 pathologists and led to a 33% reduction in immunohistochemistry usage.

Conclusion: This study highlights the potential of AI in breast lesion classification, demonstrating high sensitivity, specificity, and efficiency, and supports its integration into routine pathology practice.

查看原文本刊更多论文

一种基于人工智能的乳腺病变综合分类方法：注重提高病理学家的准确性和效率。

背景：乳腺病变的准确分类对有效的临床决策和患者管理至关重要。在这项研究中，我们评估了一种人工智能（AI）解决方案，用于对乳房病变的全幻灯片图像（wsi）进行分类。方法：我们对104例乳腺病例进行队列分析，其中浸润性癌20例，微浸润性癌4例，导管原位癌15例，小叶肿瘤/良性65例。人工智能的表现与乳房病理学家确定的基本事实进行了比较。结果：浸润性癌曲线下面积（AUC）为0.976，敏感性91.7%(84.4%)，特异性95.4%(88.0%)，特异性97.3%（97.3%）。DCIS的AUC为0.976，敏感性93.3%，特异性96.6%。对于小叶肿瘤，AUC为0.953，敏感性为94.1%，特异性为95.8%。人工智能在检测微钙化方面也表现良好，AUC为0.925，灵敏度为95%。在人工智能的支持下，病理学家的诊断准确率从97.1%提高到100%（每组303对312个准确病例）。此外，人工智能的使用显著提高了病理学家的效率，在3名病理学家中，他们的复习时间平均减少了16.5%，免疫组织化学的使用减少了33%。结论：本研究突出了人工智能在乳腺病变分类中的潜力，具有较高的敏感性、特异性和效率，支持其融入常规病理实践。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Clinical breast cancer 医学-肿瘤学

CiteScore

5.40

自引率

3.20%

发文量

174

审稿时长

48 days

期刊介绍： Clinical Breast Cancer is a peer-reviewed bimonthly journal that publishes original articles describing various aspects of clinical and translational research of breast cancer. Clinical Breast Cancer is devoted to articles on detection, diagnosis, prevention, and treatment of breast cancer. The main emphasis is on recent scientific developments in all areas related to breast cancer. Specific areas of interest include clinical research reports from various therapeutic modalities, cancer genetics, drug sensitivity and resistance, novel imaging, tumor genomics, biomarkers, and chemoprevention strategies.