An Efficient Anchor-Free Universal Lesion Detection In CT-Scans

Existing universal lesion detection (ULD) methods utilize computation intensive anchor based architectures which rely on predefined anchor boxes, often resulting in an unsatisfactory detection performance, especially in small and mid sized lesions. Further, these heuristically determined fixed anchor sizes and ratios do not generalize well to different datasets.This motivates us to propose a robust one-stage anchor-free lesion detection network which can perform well across varying lesions sizes by exploiting the fact that the box predictions can be sorted for relevance based on their center rather than their overlap with the object. Furthermore, we demonstrate that the ULD can be improved by explicitly providing them the domain-specific information in the form of multi-intensity images generated using multiple HU windows, followed by their feature-fusion using self-attention and initialization of the backbone using weights learned via self-supervision over CT-scans. We obtain comparable results to the state-of-the-art methods achieving an overall sensitivity of 86.05 on the DeepLesion dataset which comprises of approximately 32K CT scans with lesions annotated across various body organs.