Background: Data extraction for systematic reviews is highly resource-intensive. This study evaluated four frontier large language models (LLMs) on complex structured metadata extraction from specialized neuroimaging artificial intelligence (AI) literature to determine their performance in automated evidence synthesis. Methods: We compared Google Gemini 3 Pro Preview, Anthropic Claude Opus 4.5, Perplexity Sonar Pro, and OpenAI GPT 5.2. Using a standardized prompt, each model extracted 22 variables from 91 peer-reviewed neuroimaging AI articles. The variables were stratified into low-, medium-, and high-complexity tiers. The performance was measured via the exact-match accuracy against a consensus-based expert ground truth. Results: The overall exact-match accuracy was moderate. Gemini 3 Pro Preview achieved the highest overall rate (56.4%), followed by Sonar Pro (52.1%), Claude Opus 4.5 (51.3%), and GPT 5.2 (46.5%). Gemini significantly outperformed all other models (p < 0.001). The performance declined dramatically as the variable complexity increased. Across models, the accuracy was 88.9–92.9% for low-complexity categorical fields, 47.0–63.3% for medium-complexity text extraction, and 2.7–15.5% for high-complexity variables requiring clinical judgment or multi-section synthesis. The most common type of error was misclassification. All four models scored 0% on the main performance metric, but this reflected a representational mismatch with the ground truth rather than extraction failure, indicating that the exact-match accuracy underestimates the true semantic performance. Conclusions: Frontier LLMs can effectively automate the retrieval of simple categorical data, but have serious difficulties with methodological variables that are complex. Although extraction can be fully automated for low-complexity fields, human review remains essential for context-dependent variables that require clinical judgment.

Background: The reliable integration of large language models (LLMs) into neuroimaging data extraction workflows remains unresolved. Prior benchmarking shows that exact-match accuracy underestimates LLM extraction performance, but whether inter-model consensus and variable complexity can guide automation remains unclear. We evaluated whether inter-model consensus can serve as a confidence signal for human–artificial intelligence (AI) extraction and can guide complexity-stratified workflow triage. Methods: Four frontier LLMs were queried via OpenRouter with an identical zero-shot structured prompt to extract 22 predefined variables from 91 peer-reviewed neuroimaging AI articles, yielding 2002 article–variable items per model. Variables were stratified a priori into low- (n = 7), medium- (n = 8), and high-complexity (n = 7). Performance was compared with an expert reference using exact-match and semantic-equivalence accuracy. Item-level consensus and five triage strategies characterized the efficiency–accuracy trade-off. Results: Semantic-equivalence accuracy converged to 80.5–83.4% across models despite approximately ten percentage-point exact-match differences. Unanimous 4/4 consensus occurred in 45.6% (910/1994) of items, with exact-match accuracy of 85.8%, rising to 95.3% after semantic normalization; however, 14.2% still failed to match the reference. Reliability was complexity-dependent: 96.6% for low-complexity variables, 73.2% for medium-complexity variables, and 38.1% for high-complexity variables. A hybrid strategy auto-accepting 4/4 items and routing 3/4 items to rapid verification reduced estimated review effort by approximately 59%. Conclusions: Inter-model consensus is useful, but it is incomplete and depends on variable complexity. We show that LLM-assisted extraction in neuroimaging AI is a complexity-stratified workflow design problem: low-complexity neuroimaging variables may be selectively automated, while medium-complexity variables require rapid verification, and high-complexity methodological variables should remain human-led.

Background/Objectives: Medical Malpractice Stress Syndrome (MMSS) describes the psychological and behavioral consequences of malpractice-related concerns and may contribute to defensive medicine practices. Emergency medicine physicians are particularly vulnerable to medicolegal risk because of the high-acuity and time-sensitive nature of their work. This study aimed to evaluate MMSS awareness, malpractice-related stress, and defensive medicine practices among emergency medicine specialists in Türkiye. Methods: A national cross-sectional web-based survey was conducted among emergency medicine specialists practicing in Türkiye between 1 April and 1 June 2026 and was distributed through professional emergency medicine networks and congress-related channels; because the survey was disseminated through multiple open professional platforms, an exact response rate could not be determined. The questionnaire assessed demographic characteristics, MMSS awareness, malpractice-related experiences and anxiety, and defensive medicine behaviors. Defensive medicine practices were evaluated using the validated Turkish version of the Defensive Medicine Attitude Scale. Associations between malpractice-related anxiety and defensive medicine behaviors were analyzed using Spearman correlation analysis. Results: A total of 128 emergency medicine specialists completed the survey. Malpractice-related anxiety was highly prevalent and frequently influenced professional attitudes and clinical decision-making. Defensive medicine practices were common, with assurance-type behaviors being substantially more frequent than avoidance-type behaviors. Malpractice-related anxiety demonstrated a significant positive correlation with assurance defensive medicine behaviors (r = 0.383, p < 0.001) but not with avoidance behaviors (r = 0.139, p = 0.118). Assurance and avoidance defensive medicine scores were positively correlated (r = 0.575, p < 0.001). Conclusions: MMSS and malpractice-related anxiety are highly prevalent among emergency medicine physicians in Türkiye and are associated with widespread defensive medicine practices. Malpractice-related anxiety appears to primarily promote assurance-type defensive medicine behaviors rather than avoidance behaviors. These findings suggest that medicolegal stress may influence clinical decision-making and contribute to increased healthcare utilization in emergency medicine.

This study aims to compare the results of cases with a history of assault presented to our emergency department over a period of three years with literature data and to share our experiences with colleagues. Demographic characteristics, trauma findings, clinical conditions, surgical interventions, mortality, and outcomes of patients with a history of assault to the emergency department between 2020 and 2022 were retrospectively examined. The results were analyzed using the Number Cruncher Statistical System 2007 Statistical Software package program (Utah, USA). A total of 2900 patients, including 1850 males (63.79%) and 1050 females (36.21%), with an average age of 35.53±12.46 years, were included in the study. Most admissions (31.28%) occurred during the summer months and outside working hours. Males inflicted all injuries. The most common injuries occurred in the head-neck region (63.31%) and upper extremities (47.97%) in both sexes. Facial injuries were significantly more common in males (p=<0.05), whereas spinal and pelvic injuries were common in females (p<0.05). Tenderness, edema, hematoma, abrasion, ecchymosis, and soft tissue laceration were the most frequently observed findings. Female patients were more likely to be discharged home (p=0.0001), whereas male patients had a higher hospitalization rate (p=0.0001). Life-threatening ailments were discovered in 23 patients (0.79%). Surgical intervention was performed in 50 patients (1.72%). No patient experienced fatal outcomes. All assaults were committed by males. Males are more frequently and severely assaulted than females. The victims were mainly from young people, and assaults predominantly occurred during the summer months and outside working hours. Head-neck and extremity injuries were the most common. There were no fatal outcomes. Nine of the 10 patients were discharged to home from the emergency department.