Abstract:
BAA can diagnose endocrine and metabolic disorders in children. Deep learning-based bone age classification methods use global images or local information from bounding boxes or key points. Training with the global image underutilizes discriminative local information, while adding annotations is costly and subjective. This paper proposes an attention-guided method to automatically localize BAA discriminative regions without annotations. We first train a classification model to learn the attention maps of the discriminative regions, finding the hand, the most discriminative region (the carpal bones), and the next most discriminative region (the metacarpal bones). Following those attention maps, we crop the informative local regions from the original image and aggregate them for BAA. Instead of BAA as a general regression task, which is suboptimal due to the label ambiguity problem in the age label space, we propose using joint age distribution learning and expectation regression, which uses the ordinal relationship among hand images with different individual ages to estimate age more robustly. RSNA pediatric bone age data is extensively tested. Our method outperforms other deep learning-based methods without manual annotations.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
