From Description to Reasoning: Multimodal VLMs in Smart Healthcare

This Research Topic aims to pioneer the next frontier of medical AI by focusing on the convergence of multimodal VLMs and causal reasoning. Our central objective is to shift the research paradigm from passive data description and pattern recognition to active causal inference for clinical decision support. We seek to answer fundamental questions: How can we leverage VLMs not just to generate descriptive reports, but to formulate and test causal hypotheses about disease progression? How can we disentangle causal factors from confounding variables within high-dimensional, multimodal data streams? This collection will assemble leading-edge research dedicated to building computational models that can reason about cause and effect in a clinical context. The ultimate goal is to foster the development of AI systems that do not merely assist in diagnosis but provide clinicians with deep, causal insights, thereby fundamentally enhancing the diagnostic process and mitigating the risk of misjudgment based on spurious correlations.



We invite submissions of high-quality Original Research, Review, and Perspective articles aimed at advancing the frontier of medical artificial intelligence, from foundational algorithms to clinical applications. We are particularly enthusiastic about pioneering work in multimodal learning, VLMs, and causal reasoning. At the same time, we also welcome foundational research dedicated to addressing the core challenge of building reliable, transparent, and equitable medical AI systems. Our scope of interest includes, but is not limited to, the following topics:

1. Large-scale pre-trained models (e.g., VLMs, Foundation Models) for healthcare applications.

2. Novel deep learning architectures for medical data, including but not limited to CNNs, Transformers, and Mamba networks.

3. Advanced fusion, alignment, and co-learning techniques for heterogeneous data (e.g., imaging, omics, EHR, signals).

4. Generative models for medical data synthesis, augmentation, domain adaptation, and simulation.

5. Self-supervised, semi-supervised, and reinforcement learning methods for representation learning from medical data.

6. Causal inference methods for discovering causal relationships from observational medical data.

7. Explainability (XAI) methods for complex medical AI systems.

8. Studies on model robustness, generalization, and out-of-distribution (OOD) detection.

9. Uncertainty quantification in AI-driven diagnosis and prognosis.

10. Ethical considerations in medical AI, including fairness, bias, equity, and transparency.

11. AI-powered applications for computer-aided diagnosis, prognosis prediction, treatment planning, and automated report generation.

12. Human-in-the-loop systems and human-AI collaboration in clinical workflows.

13. Efficient and scalable model deployment in real-world clinical settings (e.g., edge AI, model compression).

14. Federated learning, data privacy, and security in multimodal medical AI.

15. Development of novel benchmarks, comprehensive datasets, and evaluation metrics for medical AI.

16. Applications in diverse clinical domains such as radiology, pathology, oncology, cardiology, neurology, and public health.

Article types and fees

This Research Topic accepts the following article types, unless otherwise specified in the Research Topic description:

• Brief Research Report
• Case Report
• Data Report
• Editorial
• FAIR² Data
• Hypothesis and Theory
• Methods
• Mini Review
• Original Research
• ... View all formats

Articles that are accepted for publication by our external editors following rigorous peer review incur a publishing fee charged to Authors, institutions, or funders.

Article types

This Research Topic accepts the following article types, unless otherwise specified in the Research Topic description:

• Brief Research Report
• Case Report
• Data Report
• Editorial
• FAIR² Data
• Hypothesis and Theory
• Methods
• Mini Review
• Original Research
• Perspective
• Review
• Technology and Code

Keywords: Multimodal Learning, Vision-Language Models (VLMs), Causal Reasoning, Smart Healthcare, Clinical Decision Support, Explainable AI (XAI), Medical Imaging

Important note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.