(1) Background: Coronavirus disease 2019 (COVID-19) is an infectious disease caused by SARS-CoV-2. Reverse transcription polymerase chain reaction (RT-PCR) remains the current gold standard for detecting SARS-CoV-2 infections in nasopharyngeal swabs. In Romania, the first reported patient to have contracted COVID-19 was officially declared on 26 February 2020. (2) Methods: This study proposes a federated learning approach with pre-trained deep learning models for COVID-19 detection. Three clients were locally deployed with their own dataset. The goal of the clients was to collaborate in order to obtain a global model without sharing samples from the dataset. The algorithm we developed was connected to our internal picture archiving and communication system and, after running backwards, it encountered chest CT changes suggestive for COVID-19 in a patient investigated in our medical imaging department on the 28 January 2020. (4) Conclusions: Based on our results, we recommend using an automated AI-assisted software in order to detect COVID-19 based on the lung imaging changes as an adjuvant diagnostic method to the current gold standard (RT-PCR) in order to greatly enhance the management of these patients and also limit the spread of the disease, not only to the general population but also to healthcare professionals.

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.

Ischemic stroke is a significant public health concern, with its incidence expected to double over the next 40 years, particularly among individuals over 75 years old. Previous studies, such as the DAWN trial, have highlighted the importance of correlating clinical severity with ischemic stroke volume to optimize patient management. Our study aimed to correlate the clinical severity of ischemic stroke, as assessed by the NIHSS score, with ischemic stroke volume measured using DWI, and short-term prognosis quantified by the mRS score at discharge. Conducted at the largest hospital in Gorj County from January 2023 to December 2023, this study enrolled 43 consecutive patients with acute ischemic stroke. In our patient cohort, we observed a strong positive correlation between NIHSS score and ischemic stroke volume (Spearman correlation coefficient = 0.982, p < 0.01), and a strong negative correlation between ASPECTS-DWI score and mRS score (Spearman correlation coefficient = −0.952, p < 0.01). Multiple linear regression analysis revealed a significant collective relationship between ASPECTS score, ischemic stroke volume, and NIHSS score (F(1, 41) = 600.28, p < 0.001, R² = 0.94, R²_adj = 0.93). These findings underscore the importance of DWI in assessing ischemic stroke severity and prognosis, warranting further investigation for its integration into clinical practice.

Lung cancer ranks as the second most prevalent cancer globally and is the primary contributor to neoplastic-related deaths. The approach to its treatment relies on both tumour staging and histological type determination. Data indicate that the prognosis of lung cancer is strongly linked to its clinical stage, underscoring the importance of early diagnosis in enhancing patient outcomes. Consequently, the choice of an appropriate diagnostic method holds significant importance in elevating both the early detection rate and prognosis of lung cancer. This paper aims to assess computer tomography features specific to the most common lung cancer types (adenocarcinoma, squamous cell carcinomas and small cell lung cancer). Data were collected retrospectively from CT scans of 58 patients pathologically diagnosed with lung cancer. The following CT features were evaluated and recorded for each case: location, margins, structure, lymph node involvement, cavitation, vascular bundle-thickening, bronchial obstruction, and pleural involvement. Squamous cell carcinoma (SQCC) and small cell lung cancer (SCLC) showed a higher incidence of central location, while adenocarcinoma (ADC) showed a significant predilection for a peripheral location. Internal cavitation was mostly observed in SQCC, and a solid structure was observed in almost all cases of ADC. These features can provide information about the prognosis of the patient, considering that NSCLCs are more frequent but tend to demonstrate positive results for targetable driver mutations, such as EGFR, thereby increasing the overall survival. In addition, SCLC presents with early distant spreads, which limits the opportunity to investigate the evolution of tumorigenesis and gene alterations at early stages but can have a rapidly positively response to chemotherapy. The location of the lung cancer exhibits distinct forecasts, with several studies suggesting that peripheral lung tumours offer a more favourable prognosis. Cavity formation appears correlate with a poorer prognosis. Histopathological analysis is the gold standard for diagnosing the type of lung cancer; however, using CT scanning for the purpose of a rough, but fast, preliminary diagnosis has the potential to shorten the waiting time for treatment by helping clinicians and patients to know more about the diagnosis and prognosis.

Prostate cancer is among the most frequently diagnosed cancers and a leading cause of cancer-related death in men. Currently, the most reliable and widely used imaging test for prostate cancer diagnosis is multiparametric pelvic magnetic resonance imaging (mpMRI). Modern biopsy techniques are based on the computerised merging of ultrasound and MRI images to provide better vision during the biopsy procedure (Fusion Biopsy). However, the method is expensive due to high equipment cost. Cognitive fusion of ultrasound and MRI images has recently emerged as a cheaper and easier alternative to computerised fusion. The aim of this prospective study is to perform an in-patient comparison of the systematic prostate biopsy procedure (SB) vs. cognitive fusion (CF) guided prostate biopsy method in terms of safety, ease of use, cancer detection rate and clinically significant cancer detection. We enrolled 103 patients with suspected prostate cancer that were biopsy naive, with PSA > 4 ng/dL and PIRADS score of 3, 4 or 5. All patients received a transperineal standard 12–18 cores systematic biopsy (SB) and a four-cores targeted cognitive fusion (CF) biopsy. Following the prostate biopsy, 68% of the patients were diagnosed with prostate cancer (70/103 patients). SB diagnosis rate was 62% while CF biopsy was slightly better with a 66% rate. There was a significant 20% increase in clinically significant prostate cancer detection rate for the CF compared to SB (p < 0.05) and a significant prostate cancer risk upgrade from the low to the intermediate risk category (13%, p = 0.041). Transperineal cognitive fusion targeted prostate biopsy is a straightforward biopsy method that is easy to perform and is a safe alternative to standard systematic biopsy with improved significant cancer detection accuracy. A combined targeted and systematic approach should be used for the best diagnostic results.