Feature Mixing Approach for Detecting Intraoperative Adverse Events in Laparoscopic Roux-en-Y Gastric Bypass Surgery

2025-04-23

AI Summary

Addressing the challenging detection and severity quantification of rare but high-risk intraoperative adverse events—such as hemorrhage and thermal injury—in laparoscopic Roux-en-Y gastric bypass (LRYGB), this paper proposes an end-to-end AI framework tailored for extreme class imbalance. Our method innovatively integrates Beta-distribution-driven feature mixing and continuous severity modeling, coupled with intermediate embedding regularization and generative feature alignment, enabling joint event classification and fine-grained 0–5 severity regression within a unified Transformer-based architecture. Evaluated on our newly curated MultiBypass140 dataset, the model achieves weighted F1 = 0.76, recall = 0.81, positive predictive value = 0.73, and negative predictive value = 0.84—demonstrating substantially improved robustness in detecting rare events and enhanced clinical interpretability.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Natural Language Processing: Safety and Robustness
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Intraoperative adverse events (IAEs), such as bleeding or thermal injury, can lead to severe postoperative complications if undetected. However, their rarity results in highly imbalanced datasets, posing challenges for AI-based detection and severity quantification. We propose BetaMixer, a novel deep learning model that addresses these challenges through a Beta distribution-based mixing approach, converting discrete IAE severity scores into continuous values for precise severity regression (0-5 scale). BetaMixer employs Beta distribution-based sampling to enhance underrepresented classes and regularizes intermediate embeddings to maintain a structured feature space. A generative approach aligns the feature space with sampled IAE severity, enabling robust classification and severity regression via a transformer. Evaluated on the MultiBypass140 dataset, which we extended with IAE labels, BetaMixer achieves a weighted F1 score of 0.76, recall of 0.81, PPV of 0.73, and NPV of 0.84, demonstrating strong performance on imbalanced data. By integrating Beta distribution-based sampling, feature mixing, and generative modeling, BetaMixer offers a robust solution for IAE detection and quantification in clinical settings.

Problem

• Detecting intraoperative adverse events in laparoscopic surgery
• Addressing imbalanced datasets for AI-based IAE detection
• Quantifying IAE severity via continuous regression modeling

Innovation

• Beta distribution-based mixing for severity regression
• Regularized embeddings for structured feature space
• Generative approach for robust classification