Search papers, labs, and topics across Lattice.
This paper introduces a topology-aware query selection method for surgical instrument instance segmentation, addressing the challenge of maintaining accurate object identity and count amidst common prediction errors. By modeling candidates as a complete graph and employing relational candidate representations, the approach enhances segmentation fidelity and reduces failure rates in instance set predictions. The results show significant improvements in instance F1 scores and a decrease in positive-frame set-failure rates, demonstrating the effectiveness of relational reasoning in this context.
Instance segmentation accuracy improves by over 5% with a topology-aware approach that redefines query selection as a relational problem rather than a series of independent decisions.
Accurate foreground masks can still form an incorrect surgical-instrument instance set: duplicate, fragmented, merged, missed, or empty-frame predictions may preserve favorable pixel overlap while violating object identity and count. Final query selection is therefore a relational, variable-cardinality problem rather than a collection of independent candidate decisions. We evaluate topology-aware query selection, which represents the nonempty candidates of a fixed Mask2Former as a complete graph, learns relational candidate and pair representations, predicts set cardinality, and solves an exact structured subset problem. The formal comparison is the complete relational path versus a node-feature-matched path; it evaluates the combined effect of pairwise geometry, message passing, and the additional relational-path capacity, not an isolated component. On the sealed 22-case source test, all three discovery seeds supported instance-set performance improvement with segmentation fidelity and predefined technical-safety preservation: instance F1 increased by 0.0504--0.0612 and positive-frame set-failure rate decreased by 0.0848--0.1060. Direct ROBUST-MIPS transfer reproduced the complete result in all three seeds. Endoscapes supported only one of three seeds and therefore did not establish stable direct transfer. Taken together, the results support a bounded conclusion: the evaluated complete path improved coherent instance-set construction from fixed Mask2Former candidates in specified native-instance contracts, while stable cross-domain transfer and component-specific effects remain unestablished.
