A deep learning model trained for pneumothorax detection with a large dataset of chest radiographs generalized well to multiple external datasets that were independent of the training data source. Purpose To assess the generalizability of a deep learning pneumothorax detection model on datasets from multiple external institutions and examine patient and acquisition factors that might influen...

Artificial intelligence (AI) shows promise in streamlining MRI workflows by reducing radiologists’ workload and improving diagnostic accuracy. Despite MRI’s extensive clinical use, systematic evaluation of AI-driven productivity gains in MRI remains limited. This review addresses that gap by synthesizing evidence on how AI can shorten scanning and reading times, optimize worklist triage, and automate segmentation. On 15 November 2024, we searched PubMed, EMBASE, MEDLINE, Web of Science, Google Scholar, and Cochrane Library for English-language studies published between 2000 and 15 November 2024, focusing on AI applications in MRI. Additional searches of grey literature were conducted. After screening for relevance and full-text review, 66 studies met inclusion criteria. Extracted data included study design, AI techniques, and productivity-related outcomes such as time savings and diagnostic accuracy. The included studies were categorized into five themes: reducing scan times, automating segmentation, optimizing workflow, decreasing reading times, and general time-saving or workload reduction. Convolutional neural networks (CNNs), especially architectures like ResNet and U-Net, were commonly used for tasks ranging from segmentation to automated reporting. A few studies also explored machine learning-based automation software and, more recently, large language models. Although most demonstrated gains in efficiency and accuracy, limited external validation and dataset heterogeneity could reduce broader adoption. AI applications in MRI offer potential to enhance radiologist productivity, mainly through accelerated scans, automated segmentation, and streamlined workflows. Further research, including prospective validation and standardized metrics, is needed to enable safe, efficient, and equitable deployment of AI tools in clinical MRI practice.

Background: The introduction of lung cancer screening (LCS) programmes will lead to a surge in imaging volumes and place greater demands on radiologists to provide timely and accurate interpretation. This increased workload risks overburdening a limited radiologist workforce, delaying diagnosis, and worsening burnout. Advancements in artificial intelligence (AI) models offer the potential to detect and classify pulmonary nodules without a loss in diagnostic performance. Methods: A systematic review of AI performance in lung cancer detection on computed tomography (CT) scans was conducted. Multiple databases like Medline, Embase, PubMed, and Cochrane were searched within a 12-year range from 1 January 2010 to 21 December 2022. Results: Fourteen studies were selected for this systematic review, with seven in the detection subgroup and eight in the classification subgroup. Compared to radiologists’ performance in the respective articles, the AI models demonstrated a higher sensitivity (86.0–98.1% against 68–76%) but lower specificity (77.5–87% against 87–91.7%) for the detection of lung nodules. In classifying the malignancy of lung nodules, AI models generally showed a greater sensitivity (60.58–93.3% against 76.27–88.3%), specificity (64–95.93% against 61.67–84%), and accuracy (64.96–92.46% against 73.31–85.57%) over radiologists. Conclusion: AI models for the detection and classification of pulmonary lesions on CT have the potential to augment CT thorax interpretation while maintaining diagnostic accuracy and could potentially be harnessed to overcome challenges in the implementation of lung cancer screening programmes.

Diagnostic imaging, particularly MRI, plays a key role in the evaluation of many spine pathologies. Recent progress in artificial intelligence and its subset, machine learning, has led to many applications within spine MRI, which we sought to examine in this review. A literature search of the major databases (PubMed, MEDLINE, Web of Science, ClinicalTrials.gov) was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The search yielded 1226 results, of which 50 studies were selected for inclusion. Key data from these studies were extracted. Studies were categorized thematically into the following: Image Acquisition and Processing, Segmentation, Diagnosis and Treatment Planning, and Patient Selection and Prognostication. Gaps in the literature and the proposed areas of future research are discussed. Current research demonstrates the ability of artificial intelligence to improve various aspects of this field, from image acquisition to analysis and clinical care. We also acknowledge the limitations of current technology. Future work will require collaborative efforts in order to fully exploit new technologies while addressing the practical challenges of generalizability and implementation. In particular, the use of foundation models and large-language models in spine MRI is a promising area, warranting further research. Studies assessing model performance in real-world clinical settings will also help uncover unintended consequences and maximize the benefits for patient care.

(1) Background: Chest radiographs are the mainstay of initial radiological investigation in this COVID-19 pandemic. A reliable and readily deployable artificial intelligence (AI) algorithm that detects pneumonia in COVID-19 suspects can be useful for screening or triage in a hospital setting. This study has a few objectives: first, to develop a model that accurately detects pneumonia in COVID-19 suspects; second, to assess its performance in a real-world clinical setting; and third, by integrating the model with the daily clinical workflow, to measure its impact on report turn-around time. (2) Methods: The model was developed from the NIH Chest-14 open-source dataset and fine-tuned using an internal dataset comprising more than 4000 CXRs acquired in our institution. Input from two senior radiologists provided the reference standard. The model was integrated into daily clinical workflow, prioritising abnormal CXRs for expedited reporting. Area under the receiver operating characteristic curve (AUC), F1 score, sensitivity, and specificity were calculated to characterise diagnostic performance. The average time taken by radiologists in reporting the CXRs was compared against the mean baseline time taken prior to implementation of the AI model. (3) Results: 9431 unique CXRs were included in the datasets, of which 1232 were ground truth-labelled positive for pneumonia. On the “live” dataset, the model achieved an AUC of 0.95 (95% confidence interval (CI): 0.92, 0.96) corresponding to a specificity of 97% (95% CI: 0.97, 0.98) and sensitivity of 79% (95% CI: 0.72, 0.84). No statistically significant degradation of diagnostic performance was encountered during clinical deployment, and report turn-around time was reduced by 22%. (4) Conclusion: In real-world clinical deployment, our model expedites reporting of pneumonia in COVID-19 suspects while preserving diagnostic performance without significant model drift.