一项关于乳腺癌中FGFR2免疫组化自动数字图像分析的“真实案例”经验
A 'Real-Life' Experience on Automated Digital Image Analysis of FGFR2 Immunohistochemistry in Breast Cancer.
作者信息
Braun Marcin, Piasecka Dominika, Bobrowski Mateusz, Kordek Radzislaw, Sadej Rafal, Romanska Hanna M
机构信息
Department of Pathology, Chair of Oncology, Medical University of Lodz, 92-213 Lodz, Poland.
Department of Molecular Enzymology and Oncology, Intercollegiate Faculty of Biotechnology, University of Gdansk and Medical University of Gdansk, 80-211 Gdansk, Poland.
出版信息
Diagnostics (Basel). 2020 Dec 7;10(12):1060. doi: 10.3390/diagnostics10121060.
We present here an assessment of a 'real-life' value of automated machine learning algorithm (AI) for examination of immunohistochemistry for fibroblast growth factor receptor-2 (FGFR2) in breast cancer (BC). Expression of FGFR2 in BC ( = 315) measured using a certified 3DHistech CaseViewer/QuantCenter software 2.3.0. was compared to the manual pathologic assessment in digital slides (PA). Results revealed: (i) substantial interrater agreement between AI and PA for dichotomized evaluation (Cohen's kappa = 0.61); (ii) strong correlation between AI and PA H-scores (Spearman r = 0.85, < 0.001); (iii) a small constant error and a significant proportional error (Passing-Bablok regression y = 0.51 × X + 29.9, < 0.001); (iv) discrepancies in H-score in cases of extreme (strongest/weakest) or heterogeneous FGFR2 expression and poor tissue quality. The time of AI was significantly longer (568 h) than that of the pathologist (32 h). This study shows that the described commercial machine learning algorithm can reliably execute a routine pathologic assessment, however, in some instances, human expertise is essential.
我们在此展示了一种自动化机器学习算法（AI）在乳腺癌（BC）成纤维细胞生长因子受体2（FGFR2）免疫组化检测中的“实际应用”价值评估。使用经认证的3DHistech CaseViewer/QuantCenter软件2.3.0测量的315例BC中FGFR2的表达，与数字切片中的手动病理评估（PA）进行比较。结果显示：（i）AI和PA在二分评估方面具有高度的评分者间一致性（Cohen's kappa = 0.61）；（ii）AI和PA的H评分之间具有强相关性（Spearman r = 0.85，P < 0.001）；（iii）存在较小的恒定误差和显著的比例误差（Passing-Bablok回归y = 0.51 × X + 29.9，P < 0.001）；（iv）在FGFR2表达极端（最强/最弱）或异质性以及组织质量较差的情况下，H评分存在差异。AI的时间（568小时）明显长于病理学家的时间（32小时）。本研究表明，所描述的商业机器学习算法可以可靠地执行常规病理评估，然而，在某些情况下，人类专业知识至关重要。
Zotero 插件