MedVisionLlama: Leveraging Pre-Trained Large Language Model Layers to Enhance Medical Image Segmentation
Bibliographic record
Abstract
Medical image segmentation plays a key role in healthcare, enabling accurate diagnosis and treatment planning. Vision Transformers (ViTs) show strong potential for segmentation tasks, but their dependence on large datasets limits practical usage in clinical settings. This study explores whether integrating pre-trained Large Language Models (LLMs) with ViT-based segmentation models can enhance feature refinement and improve performance in data-constrained environments. We introduce MedVisionLlama, which combines ViT encoders with pre-trained Llama weights and applies Low-Rank Adaptation (LoRA) for fine-tuning in 3D medical image segmentation. Evaluated on the Medical Segmentation Decathlon dataset, the model consistently outperformed a standard ViT, showing improved generalization across MRI and CT modalities. It maintained stable segmentation quality even with limited training data and across varied anatomical structures. Activation maps revealed sharper and more stable attention to relevant regions. Ablation studies confirmed that the performance gains stemmed from LLM-based feature refinement rather than increased model complexity. MedVisionLlama offers a scalable and data-efficient solution for medical image segmentation. Source code and implementation are available at: https://github.com/AS-Lab/Marthi-etal-2025-MedVisionLlama-Pre-Trained-LLM-Layers-to-Enhance-Medical-Image-Segmentation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.005 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".