LVPNet: A Latent-variable-based Prediction-driven End-to-end Framework for Lossless Compression of Medical Images

2025-06-22

AI Summary

Existing lossless medical image compression methods suffer from posterior collapse and inefficient utilization of latent variables due to sub-image segmentation, which enforces uniform distribution of latent information. To address this, we propose LVPNet: a prediction-driven, end-to-end latent variable compression framework that directly predicts pixel values and encodes their probability distributions via global latent modeling. We introduce two key innovations: (1) a Global Multi-Scale Perception Module (GMSM) to capture long-range dependencies in the latent space, and (2) a Quantization Compensation Module (QCM) that explicitly learns and reconstructs quantization errors. LVPNet integrates sub-image autoregression, multi-scale feature extraction, and error compensation. Evaluated on multiple authoritative medical imaging datasets, LVPNet achieves state-of-the-art performance—improving PSNR by up to 1.2 dB and average compression ratio by 8.3%, while maintaining real-time inference speed.

Technology Category

• Computer Vision: Large Vision Models
• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Autoregressive Initial Bits is a framework that integrates sub-image autoregression and latent variable modeling, demonstrating its advantages in lossless medical image compression. However, in existing methods, the image segmentation process leads to an even distribution of latent variable information across each sub-image, which in turn causes posterior collapse and inefficient utilization of latent variables. To deal with these issues, we propose a prediction-based end-to-end lossless medical image compression method named LVPNet, leveraging global latent variables to predict pixel values and encoding predicted probabilities for lossless compression. Specifically, we introduce the Global Multi-scale Sensing Module (GMSM), which extracts compact and informative latent representations from the entire image, effectively capturing spatial dependencies within the latent space. Furthermore, to mitigate the information loss introduced during quantization, we propose the Quantization Compensation Module (QCM), which learns the distribution of quantization errors and refines the quantized features to compensate for quantization loss. Extensive experiments on challenging benchmarks demonstrate that our method achieves superior compression efficiency compared to state-of-the-art lossless image compression approaches, while maintaining competitive inference speed. The code is at https://github.com/Anonymity00000/Anonymity-repository/.

Problem

• Improves lossless medical image compression efficiency
• Addresses posterior collapse in latent variable modeling
• Mitigates quantization loss in feature representation

Innovation

• Uses global latent variables for pixel prediction
• Introduces Global Multi-scale Sensing Module
• Proposes Quantization Compensation Module