OmniMed-FL: A Robust Multimodal Federated Learning Framework for Clinical Diagnosis

• URL: http://arxiv.org/abs/2609.10364v1
• Date: Wed, 09 Sep 2026 15:56:26 GMT
• Title: OmniMed-FL: A Robust Multimodal Federated Learning Framework for Clinical Diagnosis
• Abstract summary: We present OmniMed-FL, a controlled systems study of multimodal learning for five-class clinical condition classification.<n>Our proxy corpus pairs 3,000 public chest radiographs with 3,000 class-FOLD synthetic notes, matched by class, not by patient.<n>As all notes are synthetic and pairing is not patient-level, these are descriptive proxy comparisons, not estimates of diagnostic performance or deployment readiness.
• Score: 17.411567668075577
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Simultaneous assessment of medical imaging and patient records is often required in clinical diagnosis. However, standard machine learning algorithms cannot analyze these data types together. Meanwhile, compliance with HIPAA and GDPR can constrain centralized aggregation of sensitive patient data. This leaves a crucial void of secure fusion of visual and textual context across distant networks. Thus, we present OmniMed-FL, a controlled systems study of multimodal federated learning for five-class clinical condition classification (Normal, Pneumonia, COVID-19, Pleural Effusion, Cardiomegaly). Our proxy corpus pairs 3,000 public chest radiographs with 3,000 class-conditioned synthetic notes, matched by class, not by patient. The framework benchmarks eight fusion strategies, three initializations, four missing-text imputation rules, and matched federated baselines under non-IID Dirichlet partitioning across 3 to 20 hospital clients. As all notes are synthetic and pairing is not patient-level, these are descriptive proxy comparisons, not estimates of diagnostic performance or deployment readiness. Within those limits with clients ($K=5$) and severe skew ($α=0.1$), local-only training achieves a macro-F1 score of 0.297, FedAvg achieves $0.662\pm0.074$, FedProx $0.737\pm0.085$, a matched FedMME-style one-shot ensemble $0.647\pm0.080$, and our SCAFFOLD-AdamW adaptation $0.070\pm0.015$, the 0.075 FedProx-FedAvg gap falling inside the wider of the two two-seed standard deviations. Over a $4\times3$ grid, label skew costs up to 0.27 F1 whereas a near-sevenfold client increase costs at most 0.10, while bidirectional volume grows linearly to 183.5 GiB at $K=20$. Multimodal fusion leads on both corpora, scoring 0.956 against 0.934 for text and 0.664 for images on the synthetic corpus and 0.906 against 0.880 and 0.737 on the radiograph corpus, for $2.3\times$ the model state of text alone.

Related papers

• Beyond Accuracy: Quantifying Pulmonary Attribution in Anatomy-Guided Chest X-Ray Classification Under Domain Shift [0.6213771671016098]
  Deep-learning models can achieve strong chest X-ray classification performance without establishing whether their predictions rely on pulmonary image content.<n>This study evaluates pulmonary attribution containment as an anatomy-related reliability property distinct from diagnostic performance.
  arXiv  Detail & Related papers  (2026-08-31T08:54:44Z)
• ConRub-Med: Reinforcement Learning with Consensus Rubrics for Open-Ended Medical Question Answering [33.27644708906731]
  Reinforcement learning with verifiable rewards has been especially effective in mathematics and coding.<n>We introduce ConRub-Med to preserve useful distinctions as rubric feedback moves from construction to policy optimization.<n>ConRub-Med ranks first on six of nine benchmarks and the highest medical and generalization averages.
  arXiv  Detail & Related papers  (2026-08-11T14:48:15Z)
• FedDP-PALD: A Privacy-Preserving Federated Latent Diffusion Framework with Prototype Aggregation for Medical Data Synthesis [0.6372261626436676]
  We propose FedDP-PALD, a privacy-preserving framework for medical data synthesis under formal privacy guarantees.<n>It processes chest X-ray images and electrocardiogram signals through gated multi-head attention with modality-availability masks.<n>It generates private synthetic representations that preserve useful decision performance while strongly resisting membership inference.
  arXiv  Detail & Related papers  (2026-07-13T21:04:37Z)
• FedCC: A Low-Resource Federated Adaptation of Foundation Models for Robust Corpus Callosum localization in Fetal Ultrasound Images [34.592227505836156]
  We propose FedCC, a framework for corpus callosum (CC) localization in fetal US images.<n>The framework integrates a frozen DINOv2 backbone with a lightweight YOLO-based detection head.<n>It substantially reduces both computational and communication overhead, making the framework suitable for low-resource environments.
  arXiv  Detail & Related papers  (2026-06-30T09:02:13Z)
• Cascade Classification of Dermoscopic Images of Skin Neoplasms with Controllable Sensitivity and External Clinical Validation [36.94429692322632]
  Four architectures were compared in three schemes: binary (malignant/benign), single-stage four-class (benign, MEL, SCC, BCC), and a two-stage cascade (binary triage, then three-class differentiation MEL/ SCC/BCC)<n>All models used ImageNet-pretrained weights and a single augmentation protocol on aggregated open ISIC Archive data.<n>On ISIC MILK10k, direct 11-class classification yields mean-class sensitivity 0.525.
  arXiv  Detail & Related papers  (2026-06-11T09:55:57Z)
• MedFL-Stress: A Systematic Robustness Evaluation of Federated Brain Tumor Segmentation under Cross-Hospital MRI Appearance Shift [0.0]
  Federated learning enables hospitals to collaboratively train segmentation models without sharing patient data.<n>Current evaluation protocols report only average performance across clients, masking failures at individual sites.<n>We introduce MedFL-Stress, a controlled stress-testing framework that exposes exactly this failure mode.
  arXiv  Detail & Related papers  (2026-05-09T16:04:54Z)
• VSF-Med:A Vulnerability Scoring Framework for Medical Vision-Language Models [6.390468088226493]
  We introduce VSF--Med, an end-to-end vulnerability-scoring framework for medical Vision Language Models (VLMs)<n>VSF--Med synthesizes over 30,000 adversarial variants from 5,000 radiology images and enables reproducible benchmarking of any medical VLM with a single command.<n>We show that Llama-3.2-11B-Vision-Instruct exhibits a peak vulnerability increase of $1.29sigma$ for persistence-of-attack-effects, while GPT-4o shows increases of $0.69sigma$ for that same vector and $0.28sigma$ for prompt-injection attacks.
  arXiv  Detail & Related papers  (2025-06-25T02:56:38Z)
• A Privacy-Preserving Federated Learning Framework for Generalizable CBCT to Synthetic CT Translation in Head and Neck [0.0]
  Cone-beam computed tomography (CBCT) has become a widely adopted modality for image-guided radiotherapy (IGRT)<n>We propose a cross-silo horizontal federated learning (FL) approach for CBCT-to-sCT synthesis in the head and neck region.
  arXiv  Detail & Related papers  (2025-06-10T10:10:56Z)
• A Comprehensive Framework for Automated Segmentation of Perivascular Spaces in Brain MRI with the nnU-Net [37.179674347248266]
  Enlargement of perivascular spaces (PVS) is common in neurodegenerative disorders.<n>There is a need for reliable PVS detection methods which are currently lacking.
  arXiv  Detail & Related papers  (2024-11-29T09:19:57Z)
• Deep Reinforcement Learning for Cost-Effective Medical Diagnosis [41.10546022107126]
  We use reinforcement learning to find a dynamic policy that selects lab test panels sequentially based on previous observations. We propose a Semi-Model-based Deep Diagnosis Policy Optimization framework that is compatible with end-to-end training and online learning. SM-DDPO is tested on diverse clinical tasks: ferritin abnormality detection, sepsis mortality prediction, and acute kidney injury diagnosis.
  arXiv  Detail & Related papers  (2023-02-20T19:47:25Z)
• Automatic sleep stage classification with deep residual networks in a mixed-cohort setting [63.52264764099532]
  We developed a novel deep neural network model to assess the generalizability of several large-scale cohorts. Overall classification accuracy improved with increasing fractions of training data.
  arXiv  Detail & Related papers  (2020-08-21T10:48:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.