Pith. sign in

REVIEW 5 cited by

PeFoMed: Parameter Efficient Fine-tuning of Multimodal Large Language Models for Medical Imaging

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2401.02797 v3 pith:T2Q4MNJ4 submitted 2024-01-05 cs.CL cs.AI

PeFoMed: Parameter Efficient Fine-tuning of Multimodal Large Language Models for Medical Imaging

classification cs.CL cs.AI
keywords modelsmedicaltaskslanguagefine-tuninggpt-4largemllms
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Multimodal large language models (MLLMs) represent an evolutionary expansion in the capabilities of traditional large language models, enabling them to tackle challenges that surpass the scope of purely text-based applications. It leverages the knowledge previously encoded within these language models, thereby enhancing their applicability and functionality in the reign of multimodal contexts. Recent works investigate the adaptation of MLLMs as a universal solution to address medical multi-modal problems as a generative task. In this paper, we propose a parameter efficient framework for fine-tuning MLLMs, specifically validated on medical visual question answering (Med-VQA) and medical report generation (MRG) tasks, using public benchmark datasets. We also introduce an evaluation metric using the 5-point Likert scale and its weighted average value to measure the quality of the generated reports for MRG tasks, where the scale ratings are labelled by both humans manually and the GPT-4 model. We further assess the consistency of performance metrics across traditional measures, GPT-4, and human ratings for both VQA and MRG tasks. The results indicate that semantic similarity assessments using GPT-4 align closely with human annotators and provide greater stability, yet they reveal a discrepancy when compared to conventional lexical similarity measurements. This questions the reliability of lexical similarity metrics for evaluating the performance of generative models in Med-VQA and report generation tasks. Besides, our fine-tuned model significantly outperforms GPT-4v. This indicates that without additional fine-tuning, multi-modal models like GPT-4v do not perform effectively on medical imaging tasks. The code will be available here: https://github.com/jinlHe/PeFoMed.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. ErrorRadar: Benchmarking Complex Mathematical Reasoning of Multimodal Large Language Models Via Error Detection

    cs.CL 2024-10 unverdicted novelty 8.0

    ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.

  2. EchoVQA: Enabling Conversational Assistance for Point-of-Care Cardiac Ultrasound

    cs.CV 2026-05 unverdicted novelty 7.0

    EchoVQA is the first large-scale VQA dataset for echocardiography spanning high- and low-quality images across views, with acquisition guidance questions, paired with a low-parameter multimodal prompt model that repor...

  3. Retina-RAG: Retrieval-Augmented Vision-Language Modeling for Joint Retinal Diagnosis and Clinical Report Generation

    cs.CV 2026-05 unverdicted novelty 4.0

    Retina-RAG combines a DR classifier, LoRA-tuned Qwen2.5-VL, and RAG to jointly grade retinopathy, detect macular edema, and generate reports, reaching F1 0.731/0.948 and ROUGE-L 0.429 on a retinal dataset while runnin...

  4. Retina-RAG: Retrieval-Augmented Vision-Language Modeling for Joint Retinal Diagnosis and Clinical Report Generation

    cs.CV 2026-05 unverdicted novelty 4.0

    Retina-RAG combines a retinal classifier, LoRA-tuned Qwen2.5-VL, and RAG to jointly grade DR, detect ME, and generate reports, reaching F1 scores of 0.731 and 0.948 while exceeding baselines on ROUGE-L and SBERT metrics.

  5. Noise-Aware Visual Representation Learning for Medical Visual Question Answering

    cs.CV 2026-06 unverdicted novelty 3.0

    Denoising autoencoder pretraining on corrupted visual embeddings yields more robust Med-VQA performance on SLAKE and PathVQA while using LoRA for efficient LLM adaptation.