查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Machine Learning is th e subject of a report. According to news reporting originating from Dublin, Irel and, by NewsRx correspondents, research stated, "Developing artificial intellige nce (AI) models for digital pathology requires large datasets from multiple sour ces. However, without careful implementation, AI models risk learning confoundin g site-specific features in datasets instead of clinically relevant information, leading to overestimated performance, poor generalizability to real-world data, and potential misdiagnosis." Our news editors obtained a quote from the research, "Whole-slide images (WSIs) from The Cancer Genome Atlas (TCGA) colon (COAD), and stomach adenocarcinoma dat asets were selected for inclusion in this study. Patch embeddings were obtained using three feature extraction models, followed by ComBat harmonization. Attenti on-based multiple instance learning models were trained to predict tissue-source site (TSS), as well as clinical and genetic attributes, using raw, Macenko norm alized, and Combatharmonized patch embeddings. TSS prediction achieved high acc uracy (AUROC > 0.95) with all three feature extraction m odels. ComBat harmonization significantly reduced the AUROC for TSS prediction, with mean AUROCs dropping to approximately 0.5 for most models, indicating succe ssful mitigation of batch effects (e.g., CCL-ResNet50 in TCGA-COAD: Pre-ComBat A UROC = 0.960, Post-ComBat AUROC = 0.506, 0.001). Clinical attributes associated with TSS, such as race and treatment response, showed decreased predictability p ost-harmonization. Notably, the prediction of genetic features like MSI status r emained robust after harmonization (e.g., MSI in TCGA-COAD: Pre-ComBat AUROC = 0 .667, Post- ComBat AUROC = 0.669, =0.952), indicating the preservation of true hi stological signals. ComBat harmonization of deep learning-derived histology feat ures effectively reduces the risk of AI models learning confounding features in WSIs, ensuring more reliable performance estimates."
NETL
NSTL