首页|University of Sydney Reports Findings in Artificial Intelligence (RadiomicAnaly sis of Cohort-Specific Diagnostic Errors in ReadingDense Mammograms Using Artif icial Intelligence)
University of Sydney Reports Findings in Artificial Intelligence (RadiomicAnaly sis of Cohort-Specific Diagnostic Errors in ReadingDense Mammograms Using Artif icial Intelligence)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – New research on Artificial Intelligenc e is the subject of a report. According to news reportingoriginating in Sydney, Australia, by NewsRx journalists, research stated, “This study aims to investigate radiologists’ interpretation errors when reading dense screening mammograms using a radiomics-basedartificial intelligence approach. Thirty-six radiologist s from China and Australia read 60 dense mammograms.”The news reporters obtained a quote from the research from the University of Syd ney, “For eachcohort, we identified normal areas that looked suspicious of canc er and the malignant areas containingcancers. Then radiomic features were extra cted from these identified areas and random forest models weretrained to recogn ize the areas that were most frequently linked to diagnostic errors within each cohort.The performance of the model and discriminatory power of significant rad iomic features were assessed. Wefound that in the Chinese cohort, the AUC value s for predicting false positives were 0.864 (CC) and 0.829(MLO), while in the A ustralian cohort, they were 0.652 (CC) and 0.747 (MLO). For false negatives, theAUC values in the Chinese cohort were 0.677 (CC) and 0.673 (MLO), and in the Au stralian cohort, theywere 0.600 (CC) and 0.505 (MLO). In both cohorts, regions with higher Gabor and maximum responsefilter outputs were more prone to false p ositives, while areas with significant intensity changes and coarsetextures wer e more likely to yield false negatives. This cohort-based pipeline proves effect ive in identifyingcommon errors for specific reader cohorts based on image-deri ved radiomic features.”
SydneyAustraliaAustralia and New Zea landArtificial IntelligenceBreast Cancer ScreeningCancerDiagnostics and ScreeningEmerging TechnologiesHealth andMedicineMachine LearningMammogr amMammographyOncologyRisk and PreventionWomen’sHealth
NETL
NSTL
