Scholarlink

Background/objective <p>Digitalization and artificial intelligence (AI) are increasingly being tested and integrated into the evaluation of histological sections. While the application in the field of tumor pathology is already well advanced, experience with the digital analysis of histological sections of the myocardium is limited. We present a&#xa0;project for AI-supported analysis of histological samples in myocardial infarction, where we tested whether a&#xa0;specifically trained algorithm could distinguish healthy myocardial sections from those affected by ischemia.</p> Material and methods <p>A total of 106 slides with HE-stained myocardial sections from 50&#xa0;deceased individuals were digitalized and annotated. A&#xa0;convolutional neural network (CNN) based on a&#xa0;ResNet-18 architecture was trained for AI-assisted classification of the digitalized myocardial sections. The data were divided and slide-wise stratified into a&#xa0;training, validation and test dataset and analyzed by the deep learning algorithm for image pattern recognition of the CNN.</p> Results <p>The algorithm developed was able to reliably distinguish healthy myocardium from pathological alterations in most cases during the test runs. Healthy myocardial tissue was detected with a&#xa0;precision of 81%, infarcted tissue with 78% and fibrosis with 85%.</p> Discussion <p>The insights gained in this project will be used to further develop the approach of AI-supported analysis of myocardial sections. With the expansion of the sample in terms of the number of cases and the pathologies present, future applications may arise in histological examinations of sudden cardiac death, especially in cases of underlying rare diseases (hereditary cardiomyopathies, etc.), where to date only limited micromorphological correlates could be identified using conventional histological analysis.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Maschinelles Lernen zur Identifikation pathologischer Myokardregionen: ein innovativer Ansatz für die digitale Pathologie

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月21日

查看全文
  • E. Corvest,
  • R. S. Mayer,
  • C. Kocuk,
  • V. Wilmes,
  • S. Gretser,
  • E. Gradhand,
  • P. J. Wild,
  • M. A. Verhoff,
  • N. Flinner,
  • S. Kauferstein

摘要

Background/objective

Digitalization and artificial intelligence (AI) are increasingly being tested and integrated into the evaluation of histological sections. While the application in the field of tumor pathology is already well advanced, experience with the digital analysis of histological sections of the myocardium is limited. We present a project for AI-supported analysis of histological samples in myocardial infarction, where we tested whether a specifically trained algorithm could distinguish healthy myocardial sections from those affected by ischemia.

Material and methods

A total of 106 slides with HE-stained myocardial sections from 50 deceased individuals were digitalized and annotated. A convolutional neural network (CNN) based on a ResNet-18 architecture was trained for AI-assisted classification of the digitalized myocardial sections. The data were divided and slide-wise stratified into a training, validation and test dataset and analyzed by the deep learning algorithm for image pattern recognition of the CNN.

Results

The algorithm developed was able to reliably distinguish healthy myocardium from pathological alterations in most cases during the test runs. Healthy myocardial tissue was detected with a precision of 81%, infarcted tissue with 78% and fibrosis with 85%.

Discussion

The insights gained in this project will be used to further develop the approach of AI-supported analysis of myocardial sections. With the expansion of the sample in terms of the number of cases and the pathologies present, future applications may arise in histological examinations of sudden cardiac death, especially in cases of underlying rare diseases (hereditary cardiomyopathies, etc.), where to date only limited micromorphological correlates could be identified using conventional histological analysis.