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{209948,
        author = {DHANU SRI S and NITHYA R},
        title = {A Survey of AI-Based Radiology Report Generation and Clinical Consistency Verification},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {5},
        pages = {3933-3940},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=209948},
        abstract = {Radiology report generation (RRG) aims to translate medical images into fluent, clinically meaningful narrative reports, easing radiologist workload and shortening reporting delays. Progress from CNN–RNN captioning to memory-driven transformers, knowledge-enhanced networks and radiology-specific multimodal large language models has steadily improved lexical overlap with reference reports. However, generated reports still contain hallucinated findings, omitted abnormalities, wrong laterality or severity, and invented references to prior studies, errors that n-gram metrics barely register. This survey examines AI-driven radiology report generation (RRG) and explores techniques designed to improve the factual reliability of generated reports by evaluating their alignment with medical images and reference reports. We propose a taxonomy of generation paradigms and verification mechanisms, summarise public datasets (IU X-ray, MIMIC-CXR, CheXpert Plus, PadChest, CT-RATE) and evaluation metrics, and compare published results on MIMIC-CXR and IU X-ray. We close with open challenges, including dataset bias, evaluation validity, hallucination, longitudinal reasoning and clinical deployment, and outline research directions toward verifiable, radiologist-in-the-loop report generation.},
        keywords = {Radiology Report Generation, Clinical Consistency Verification, Hallucination, Vision–Language Models, Large Language Models, Chest X-ray, Evaluation Metrics.},
        month = {October},
        }