The integration of deep learning into radiology has the potential to enhance diagnostic processes, yet its acceptance in clinical practice remains limited due to various challenges. This study aimed to develop and evaluate a fine-tuned large language model (LLM), based on Llama 3-8B, to automate the generation of accurate and concise conclusions in magnetic resonance imaging (MRI) and computed tomography (CT) radiology reports, thereby assisting radiologists and improving reporting efficiency. A dataset comprising 15,000 radiology reports was collected from the University of Medicine and Pharmacy of Craiova’s Imaging Center, covering a diverse range of MRI and CT examinations made by four experienced radiologists. The Llama 3-8B model was fine-tuned using transfer-learning techniques, incorporating parameter quantization to 4-bit precision and low-rank adaptation (LoRA) with a rank of 16 to optimize computational efficiency on consumer-grade GPUs. The model was trained over five epochs using an NVIDIA RTX 3090 GPU, with intermediary checkpoints saved for monitoring. Performance was evaluated quantitatively using Bidirectional Encoder Representations from Transformers Score (BERTScore), Recall-Oriented Understudy for Gisting Evaluation (ROUGE), Bilingual Evaluation Understudy (BLEU), and Metric for Evaluation of Translation with Explicit Ordering (METEOR) metrics on a held-out test set. Additionally, a qualitative assessment was conducted, involving 13 independent radiologists who participated in a Turing-like test and provided ratings for the AI-generated conclusions. The fine-tuned model demonstrated strong quantitative performance, achieving a BERTScore F1 of 0.8054, a ROUGE-1 F1 of 0.4998, a ROUGE-L F1 of 0.4628, and a METEOR score of 0.4282. In the human evaluation, the artificial intelligence (AI)-generated conclusions were preferred over human-written ones in approximately 21.8% of cases, indicating that the model’s outputs were competitive with those of experienced radiologists. The average rating of the AI-generated conclusions was 3.65 out of 5, reflecting a generally favorable assessment. Notably, the model maintained its consistency across various types of reports and demonstrated the ability to generalize to unseen data. The fine-tuned Llama 3-8B model effectively generates accurate and coherent conclusions for MRI and CT radiology reports. By automating the conclusion-writing process, this approach can assist radiologists in reducing their workload and enhancing report consistency, potentially addressing some barriers to the adoption of deep learning in clinical practice. The positive evaluations from independent radiologists underscore the model’s potential utility. While the model demonstrated strong performance, limitations such as dataset bias, limited sample diversity, a lack of clinical judgment, and the need for large computational resources require further refinement and real-world validation. Future work should explore the integration of such models into clinical workflows, address ethical and legal considerations, and extend this approach to generate complete radiology reports.

Currently, medical imaging has largely supplanted traditional methods in the realm of diagnosis and treatment planning. This shift is primarily attributable to the non-invasive nature, rapidity, and user-friendliness of medical-imaging techniques. The widespread adoption of medical imaging, however, has shifted the bottleneck to healthcare professionals who must analyze each case post-image acquisition. This process is characterized by its sluggishness and subjectivity, making it susceptible to errors. The anterior cruciate ligament (ACL), a frequently injured knee ligament, predominantly affects a youthful and sports-active demographic. ACL injuries often leave patients with substantial disabilities and alter knee mechanics. Since some of these cases necessitate surgery, it is crucial to accurately classify and detect ACL injury. This paper investigates the utilization of pre-trained convolutional neural networks featuring residual connections (ResNet) along with image-processing methods to identify ACL injury and differentiate between various tear levels. The ResNet employed in this study is not the standard ResNet but rather an adapted version capable of processing 3D volumes constructed from 2D image slices. Achieving a peak accuracy of 97.15% with a custom split, 96.32% through Monte-Carlo cross-validation, and 93.22% via five-fold cross-validation, our approach enhances the performance of three-class classifiers by over 7% in terms of raw accuracy. Moreover, we achieved an improvement of more than 1% across all types of evaluation. It is quite clear that the model’s output can effectively serve as an initial diagnostic baseline for radiologists with minimal effort and nearly instantaneous results. This advancement underscores the paper’s focus on harnessing deep learning for the nuanced detection and classification of ACL tears, demonstrating a significant leap toward automating and refining diagnostic accuracy in sports medicine and orthopedics.