Accuracy Evaluation for a Deep Learning Algorithm of Lesion Detection on Breast MR Images: A Preliminary Study
ECR24 / C-14978
Congress: ECR24
Poster Number: C-14978
Type: EPOS Radiologist (scientific)
DOI: 10.26044/ecr2024/C-14978
Authorblock: K. Geißler1, N. Uchiyama2, R. Grimm3, T. Suzuki4, U. Fischer5, D. H. M. Szolar6, H. Meine1, H. Von Busch3, R. M. Murakami4; 1Bremen/DE, 2TOKYO/JP, 3Erlangen/DE, 4Tokyo/JP, 5Goettingen/DE, 6Graz/AT
Disclosures:
Kai Geißler: Nothing to disclose
Nachiko Uchiyama: Nothing to disclose
Robert Grimm: Employee: Siemens Healthcare GmbH
Takashi Suzuki: Employee: Siemens Healthcare GmbH
Uwe Fischer: Nothing to disclose
Dieter H. M. Szolar: Nothing to disclose
Hans Meine: Nothing to disclose
Heinrich Von Busch: Employee: Siemens Healthcare GmbH
Ryusuke Medical Murakami: Nothing to disclose
Keywords: Artificial Intelligence, Breast, MR, Computer Applications-Detection, diagnosis, Segmentation, Cancer
Conclusion
The AI-based lesion detection shows high sensitivity with an acceptable level of false positives, potentially supporting the radiologist’s workflow. By adjusting the cut-off value, false positives can be reduced but sensitivity also decreases. Thus, different operating points may be appropriate depending on the expected prevalence of cancer in the studied cohort.
GALLERY
Fig 1: The research application used in this study
Fig 2: Proposed algorithm: The breast and parenchyma tissue are seg...
Table 3: Data used for training of lesion detection model. (* = Mean ...
Table 4: Lesion detection results on external test dataset
Fig 5: True positive lesions with automatic segmentation masks (gre...
Fig 6: False positive lesions with automatic segmentation masks
Fig 7: False negative lesions with manually annotated segmentation ...