AdaLoRA-QAT: Adaptive Low-Rank and Quantization-Aware Segmentation
2026-04-01 • Artificial Intelligence
Artificial IntelligenceComputer Vision and Pattern Recognition
AI summaryⓘ
The authors developed AdaLoRA-QAT, a method to make large chest X-ray analysis models smaller and easier to use in hospitals without losing accuracy. Their approach combines smart ways to reduce model size with careful training that keeps important details intact. Tests show that their smaller, faster model works just as well as the full-size one for identifying areas in chest X-rays. This method helps create practical medical imaging tools that require less computing power.
Chest X-ray (CXR)Image segmentationFoundation modelsLow-rank adaptationQuantization-aware trainingINT8 quantizationDice scoreModel compressionWilcoxon signed-rank testComputer-aided diagnosis
Authors
Prantik Deb, Srimanth Dhondy, N. Ramakrishna, Anu Kapoor, Raju S. Bapi, Tapabrata Chakraborti
Abstract
Chest X-ray (CXR) segmentation is an important step in computer-aided diagnosis, yet deploying large foundation models in clinical settings remains challenging due to computational constraints. We propose AdaLoRA-QAT, a two-stage fine-tuning framework that combines adaptive low-rank encoder adaptation with full quantization-aware training. Adaptive rank allocation improves parameter efficiency, while selective mixed-precision INT8 quantization preserves structural fidelity crucial for clinical reliability. Evaluated across large-scale CXR datasets, AdaLoRA-QAT achieves 95.6% Dice, matching full-precision SAM decoder fine-tuning while reducing trainable parameters by 16.6\times and yielding 2.24\times model compression. A Wilcoxon signed-rank test confirms that quantization does not significantly degrade segmentation accuracy. These results demonstrate that AdaLoRA-QAT effectively balances accuracy, efficiency, and structural trust-worthiness, enabling compact and deployable foundation models for medical image segmentation. Code and pretrained models are available at: https://prantik-pdeb.github.io/adaloraqat.github.io/