Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{185712,
        author = {Siddharth Padwal and Vivek Sonone and Dr.S.S.Khatal and Omkar Bangar and Dr. Puja Gholap},
        title = {A Multimodal Medical AI Model for Clinical Imaging and Question Answering (Med Vision AI)},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {5},
        pages = {2762-2765},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=185712},
        abstract = {Artificial Intelligence (AI) is fundamentally transforming healthcare, particularly in the field of medical diagnostics that involves clinical imaging and natural language processing. A primary challenge with existing AI models is their unimodal focus, concentrating on either image analysis or text comprehension. This specialization limits their utility in clinical settings that require a comprehensive, context-aware understanding of patient data. This paper introduces "A Multimodal Medical AI Model for Clinical Imaging and Question Answering (QA)," a project that confronts this limitation by developing an integrated AI framework. This framework is engineered to simultaneously interpret and reason over both visual and textual medical data. Our methodology employs sophisticated transformer-based architectures to process medical imagery and associated clinical reports. These distinct data streams are integrated using a multimodal fusion technique that harmonizes the features from each source. The resulting unified representation allows the system to handle intricate medical questions formulated in natural language, thereby emulating the diagnostic thought process of a clinical expert. To validate its real-world applicability, the model undergoes training and validation using reputable public datasets, including MIMIC-CXR and MedQA. A user-friendly interface is also developed to allow healthcare professionals to upload medical images and submit diagnostic inquiries. By leveraging the synergy between medical imaging, NLP, and deep learning, this initiative aims to elevate diagnostic accuracy, lessen the workload on medical staff, and broaden access to advanced medical knowledge, especially in underserved regions. This holistic approach signifies a major leap forward in the creation of more intelligent and effective healthcare technologies.},
        keywords = {Clinical Decision Support, Deep Learning, Medical Imaging, Multimodal AI, Natural Language Processing, Transformer Models, Visual Question Answering (VQA).},
        month = {October},
        }