Parameter-Efficient Subspace Decoupling ViT for Mitigating Multi-Task Negative Transfer in Histological Scoring
arXiv:2605.29852v1 Announce Type: cross Abstract: Histological scoring is essential for diagnosing Non-Alcoholic Fatty Liver Disease (NAFLD), yet its automation remains challenging due to the high annotation cost and negative transfer among the strongly correlated NAFLD Activity Score (NAS) indicators in multi-task learning. To address this issue, we propose a subspace-decoupled multi-task Vision Transformer (ViT) that integrates lightweight task-specific Adapters with orthogonality-based constraints. This design constructs independent feature subspaces for steatosis, ballooning, and inflammat
The increasing availability of large-scale medical imaging data and advances in Vision Transformer architectures are enabling more sophisticated AI applications in diagnostics.
This development addresses key challenges in automating medical diagnostics by improving the efficiency and accuracy of multi-task learning models, potentially reducing annotation costs and diagnostic variability.
The proposed method introduces a parameter-efficient way to mitigate negative transfer in multi-task learning for histological scoring, making AI-driven diagnoses more reliable and cost-effective.
- · Medical AI developers
- · Pathologists
- · Healthcare providers
- · Patients
- · Manual annotation services
Improved accuracy and efficiency in automated diagnostic scoring for conditions like NAFLD.
Reduced healthcare costs associated with manual histological analysis and faster diagnostic turnaround times.
Accelerated drug discovery and therapeutic development by providing more consistent and scalable diagnostic tools for clinical trials.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG