隨著數(shù)字組織圖像分析和全切片圖像（WSI）在毒性病理學(xué)診斷和同行評議的應(yīng)用日益廣泛，傳統(tǒng)病理學(xué)逐漸向數(shù)字病理學(xué)過渡。近10年來，人工智能（AI）及機(jī)器學(xué)習(xí)（ML）的快速發(fā)展促進(jìn)了組織病理學(xué)診斷模式的轉(zhuǎn)變，數(shù)字組織圖像分析和AI已成為新藥研發(fā)、藥物非臨床安全性評價(jià)中毒性病理學(xué)診斷和病理學(xué)同行評議中不可或缺的工具和技術(shù)手段。概述數(shù)字組織圖像分析在毒性病理學(xué)中應(yīng)用和挑戰(zhàn)、AI的發(fā)展進(jìn)程及在毒性病理學(xué)中應(yīng)用和挑戰(zhàn)，以期為我國藥物非臨床安全性評價(jià)毒理學(xué)試驗(yàn)組織圖像分析和AI在毒性病理學(xué)中的應(yīng)用提供一定參考。

With the increasing application of digital tissue image analysis and whole slide image (WSI) in the diagnosis and peer review of toxicologic pathology, traditional pathology gradually transits to digital pathology. In recent ten years, the rapid development of artificial intelligence (AI) and machine learning (ML) has promoted the transformation of histopathological diagnosis mode. Digital tissue image analysis and AI have become indispensable tools and technical means for the research and development of new drugs, the toxicologic pathology diagnosis during non-clinical safety evaluation of drugs and the pathology peer review. This paper summarizes the application and challenge of digital tissue image analysis in toxicologic pathology, the development process of AI and its application and challenge in toxicologic pathology, in order to provide some reference for the application of tissue image analysis and AI in toxicology experiments of non-clinical safety evaluation of drugs in China.

R965.3