International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 187 - Number 43 |
Year of Publication: 2025 |
Authors: A.N. Ramya Shree, Nithya N., Lavanya Kamaraju, Sahana M.B., Hrithwika, Supriya |
![]() |
A.N. Ramya Shree, Nithya N., Lavanya Kamaraju, Sahana M.B., Hrithwika, Supriya . MedXGen: LLM leveraged Framework for Automated Clinical Coherent Medical Report Generation. International Journal of Computer Applications. 187, 43 ( Sep 2025), 23-28. DOI=10.5120/ijca2025925746
Artificial intelligence-based automatic medical report creation has accelerated significantly since the introduction of cross-modal learning, sophisticated transformer structures, and knowledge-enhanced pretraining methods. The detector attention modules, adapter tuned vision language models, and graph-guided hybrid strategies are integrated to propose framework for automated medical report generation. Utilizing topic wise separable retrieval, hierarchical cross-modal alignment, and phrase-level augmentation, the proposed MedXGen confront semantic inconsistency, hallucination and redundancy. Memory-guided transformers and semi- supervised learning is used to enhance interpretability and adaptability. The suggested framework offers a practical implementation of clinical diagnostic support systems and it supports medical language creation and visual perception.