PREBA: Surgical Duration Prediction via PCA-Weighted Retrieval-Augmented LLMs and Bayesian Averaging Aggregation

2026-02-27

AI Summary

Existing approaches to surgical duration prediction either fail to model institution-specific clinical context in zero-shot settings—leading to unstable and clinically implausible large language model (LLM) outputs—or rely on supervised learning with extensive labeled data and high computational costs. To address this, we propose PREBA, a novel framework that integrates PCA-weighted retrieval and Bayesian averaging into the retrieval-augmented generation (RAG) paradigm. PREBA retrieves clinically similar historical cases using PCA-weighted similarity, constructs evidence-enhanced prompts, and fuses multi-round LLM predictions with population-level statistical priors via Bayesian averaging to produce calibrated duration estimates without any model training. Evaluated on two real-world clinical datasets, PREBA reduces mean absolute error by up to 40% compared to zero-shot inference and improves R² from −0.13 to 0.62, achieving accuracy, stability, and clinical plausibility on par with supervised models.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Accurate prediction of surgical duration is pivotal for hospital resource management. Although recent supervised learning approaches-from machine learning (ML) to fine-tuned large language models (LLMs)-have shown strong performance, they remain constrained by the need for high-quality labeled data and computationally intensive training. In contrast, zero-shot LLM inference offers a promising training-free alternative but it lacks grounding in institution-specific clinical context (e.g., local demographics and case-mix distributions), making its predictions clinically misaligned and prone to instability. To address these limitations, we present PREBA, a retrieval-augmented framework that integrates PCA-weighted retrieval and Bayesian averaging aggregation to ground LLM predictions in institution-specific clinical evidence and statistical priors. The core of PREBA is to construct an evidence-based prompt for the LLM, comprising (1) the most clinically similar historical surgical cases and (2) clinical statistical priors. To achieve this, PREBA first encodes heterogeneous clinical features into a unified representation space enabling systematic retrieval. It then performs PCA-weighted retrieval to identify clinically relevant historical cases, which form the evidence context supplied to the LLM. Finally, PREBA applies Bayesian averaging to fuse multi-round LLM predictions with population-level statistical priors, yielding calibrated and clinically plausible duration estimates. We evaluate PREBA on two real-world clinical datasets using three state-of-the-art LLMs, including Qwen3, DeepSeek-R1, and HuatuoGPT-o1. PREBA significantly improves performance-for instance, reducing MAE by up to 40% and raising R^2 from -0.13 to 0.62 over zero-shot inference-and it achieves accuracy competitive with supervised ML methods, demonstrating strong effectiveness and generalization.

Problem

• surgical duration prediction
• retrieval-augmented LLMs
• institution-specific clinical context
• zero-shot inference
• clinical alignment

Innovation

• Retrieval-Augmented LLMs
• PCA-Weighted Retrieval
• Bayesian Averaging
• Surgical Duration Prediction
• Zero-Shot Inference