Lesion Labeling on Mammogram by Combining Object Detection with Structured Report Interpretation
Yi‐Chong Zeng · 2019 IEEE 1st Global Conference on Life Sciences and Technologies (LifeTech) · 2019
Data labeling plays an important role in many fields. Current approaches perform labeling of work by manual operation results in time consumption. For analysis of mammogram image, people need to mark region of interesting (ROI) of a lesion in advance, such as mass, calcification, architectural distortion, and asymmetry. In this paper, we propose a scheme to automatically label lesion on mammogram images, it is realized by combining object detection with structured report interpretation. Moreover, we develop a user interface to modify the ROIs of the labeled lesions. The experiment results show that the proposed scheme comes to similar performance to the compared approaches in lesion labeling.