TY - GEN
T1 - The Case of Missed Cancers
T2 - 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2020
AU - Chorev, Michal
AU - Shoshan, Yoel
AU - Spiro, Adam
AU - Naor, Shaked
AU - Hazan, Alon
AU - Barros, Vesna
AU - Weinstein, Iuliana
AU - Herzel, Esma
AU - Shalev, Varda
AU - Guindy, Michal
AU - Rosen-Zvi, Michal
N1 - Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020/1/1
Y1 - 2020/1/1
N2 - We investigate the potential contribution of an AI system as a safety net application for radiologists in breast cancer screening. As a safety net, the AI alerts on cases suspected to be malignant which the radiologist did not recommend for a recall. We analyzed held-out data of 2,638 exams enriched with 90 missed cancers. In screening mammography settings, we show that a system alerting on 11 out of every 1,000 cases, could detect up to 10.7% of the radiologists’ missed cancers. Thus, significantly increasing radiologist’s sensitivity to 80.3%, while only slightly decreasing their specificity to 95.3%. Importantly, the safety net demonstrated a significant contribution to their performance even when radiologists utilized both mammography and ultrasound images. In those settings, it would have alerted 8.5 times per 1,000 cases, and detected 11.7% of the radiologists’ missed cancers. In an analysis of the missed cancers by an expert, we found that most of the cancers detected by the AI were visible post-hoc. Finally, we performed a reader study with five radiologists over 120 exams, 10 of which were originally missed cancers. The AI safety net was able to assist 3 out of the 5 radiologists in detecting missed cancers without raising any false alerts.
AB - We investigate the potential contribution of an AI system as a safety net application for radiologists in breast cancer screening. As a safety net, the AI alerts on cases suspected to be malignant which the radiologist did not recommend for a recall. We analyzed held-out data of 2,638 exams enriched with 90 missed cancers. In screening mammography settings, we show that a system alerting on 11 out of every 1,000 cases, could detect up to 10.7% of the radiologists’ missed cancers. Thus, significantly increasing radiologist’s sensitivity to 80.3%, while only slightly decreasing their specificity to 95.3%. Importantly, the safety net demonstrated a significant contribution to their performance even when radiologists utilized both mammography and ultrasound images. In those settings, it would have alerted 8.5 times per 1,000 cases, and detected 11.7% of the radiologists’ missed cancers. In an analysis of the missed cancers by an expert, we found that most of the cancers detected by the AI were visible post-hoc. Finally, we performed a reader study with five radiologists over 120 exams, 10 of which were originally missed cancers. The AI safety net was able to assist 3 out of the 5 radiologists in detecting missed cancers without raising any false alerts.
KW - Breast imaging
KW - Computer-aided diagnosis
KW - Deep learning
UR - http://www.scopus.com/inward/record.url?scp=85092794141&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-59725-2_22
DO - 10.1007/978-3-030-59725-2_22
M3 - Conference contribution
AN - SCOPUS:85092794141
SN - 9783030597245
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 220
EP - 229
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2020 - 23rd International Conference, Proceedings
A2 - Martel, Anne L.
A2 - Abolmaesumi, Purang
A2 - Stoyanov, Danail
A2 - Mateus, Diana
A2 - Zuluaga, Maria A.
A2 - Zhou, S. Kevin
A2 - Racoceanu, Daniel
A2 - Joskowicz, Leo
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 4 October 2020 through 8 October 2020
ER -
