Neural Finite-State Machines for Surgical Phase Recognition

TL;DR

The paper introduces Neural Finite-State Machines (NFSM), a novel neural approach that enforces temporal coherence in surgical phase recognition, significantly improving accuracy and robustness in analyzing surgical videos.

Contribution

It presents NFSM, a plug-and-play module combining finite-state principles with neural networks, enhancing sequential modeling in surgical phase recognition without altering existing architectures.

Findings

Achieved state-of-the-art results on multiple benchmarks.

Improved video-level accuracy by 0.9 points on BernBypass70.

Enhanced phase-level precision, recall, F1-score, and mAP by over 3 points.

Abstract

Surgical phase recognition (SPR) is crucial for applications in workflow optimization, performance evaluation, and real-time intervention guidance. However, current deep learning models often struggle with fragmented predictions, failing to capture the sequential nature of surgical workflows. We propose the Neural Finite-State Machine (NFSM), a novel approach that enforces temporal coherence by integrating classical state-transition priors with modern neural networks. NFSM leverages learnable global state embeddings as unique phase identifiers and dynamic transition tables to model phase-to-phase progressions. Additionally, a future phase forecasting mechanism employs repeated frame padding to anticipate upcoming transitions. Implemented as a plug-and-play module, NFSM can be integrated into existing SPR pipelines without changing their core architectures. We demonstrate state-of-the-art performance across multiple benchmarks, including a significant improvement on the BernBypass70 dataset - raising video-level accuracy by 0.9 points and phase-level precision, recall, F1-score, and mAP by 3.8, 3.1, 3.3, and 4.1, respectively. Ablation studies confirm each component's effectiveness and the module's adaptability to various architectures. By unifying finite-state principles with deep learning, NFSM offers a robust path toward consistent, long-term surgical video analysis.