Background/Objectives: Chest radiography is the primary first-line imaging tool for diagnosing pneumothorax in pediatric emergency settings. However, interpretation under clinical pressures such as high patient volume may lead to delayed or missed diagnosis, particularly for subtle cases. This study aimed to evaluate the diagnostic performance of ChatGPT-5, a multimodal large language model, in detecting and localizing pneumothorax on pediatric chest radiographs using multiple prompting strategies. Methods: In this retrospective study, 380 pediatric chest radiographs (190 pneumothorax cases and 190 matched controls) from a tertiary hospital were interpreted using ChatGPT-5 with three prompting strategies: instructional, role-based, and clinical-context. Performance metrics, including accuracy, sensitivity, specificity, and conditional side accuracy, were evaluated against an expert-adjudicated reference standard. Results: ChatGPT-5 achieved an overall accuracy of 0.77–0.79 and consistently high specificity (0.96–0.98) across all prompts, with stable reproducibility. However, sensitivity was limited (0.57–0.61) and substantially lower for small pneumothoraces (American College of Chest Physicians [ACCP]: 0.18–0.22; British Thoracic Society [BTS]: 0.41–0.46) than for large pneumothoraces (ACCP: 0.75–0.79; BTS: 0.85–0.88). The conditional side accuracy exceeded 0.96 when pneumothorax was correctly detected. No significant differences were observed among prompting strategies. Conclusions: ChatGPT-5 showed consistent but limited diagnostic performance for pediatric pneumothorax. Although the high specificity and reproducible detection of larger pneumothoraces reflect favorable performance characteristics, the unacceptably low sensitivity for subtle pneumothoraces precludes it from independent clinical interpretation and underscores the necessity of oversight by emergency clinicians.
We report the case of a school-aged patient with attention-deficit/hyperactivity disorder who presented with a palpable epigastric mass. The initial abdominal ultrasonography indicated the presence of a heterogeneous space-occupying lesion in the upper abdomen. Subsequent computed tomography revealed a large intragastric mass with a mottled air-containing density, an imaging feature characteristic of a bezoar. Esophagogastroduodenoscopy confirmed a massive trichobezoar extending beyond the pylorus into the duodenum, consistent with Rapunzel syndrome. Although the endoscopic removal proved unsuccessful, a subsequent surgical extraction via laparotomy yielded a 22 cm trichobezoar. Further history revealed prior hair-picking behavior approximately one year earlier, with a localized bald patch noted by her parents. This case highlights the characteristic multimodal imaging findings of trichobezoars and serves to emphasize the diagnostic challenge posed by a limited clinical history in children with psychiatric comorbidities presenting with abdominal masses.
Background/Objectives: This study aimed to identify resting-state quantitative electroencephalography (QEEG) correlates of cognitive performance in children with attention-deficit/hyperactivity disorder (ADHD). Methods: ADHD diagnoses were established according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. Cognitive performance was evaluated using the Wechsler Intelligence Scale for Children, Fourth Edition, focusing on the Working Memory Index (WMI) and Processing Speed Index (PSI). Following at least 24 h medication washout, QEEG recordings were collected during both the eyes-open (EO) and eyes-closed (EC) resting states. The absolute spectral power across four frequency bands (delta, theta, alpha, and beta) was analyzed within seven predefined scalp regions. Associations between QEEG measures and cognitive indices were assessed using Spearman’s rank correlation with false discovery rate (FDR) correction. Results: A total of 45 children with ADHD were included: 29 with predominantly inattentive ADHD (ADHD-I) and 16 with combined ADHD (ADHD-C). No significant associations between the QEEG measures and cognitive performance were observed in the overall ADHD cohort. However, within the ADHD-I presentation, EO alpha power was positively correlated with WMI scores across bilateral frontal and central regions (Spearman’s ρ = 0.535–0.582, FDR-corrected q = 0.014–0.028). In contrast, no significant alpha–WMI associations were identified in the ADHD-C presentation. Furthermore, no significant correlations were observed during the EC condition, and no EEG-PSI associations survived the FDR in either presentation. Conclusions: Our findings demonstrate presentation-related differences in QEEG–cognitive relationships and suggest that EO frontocentral alpha activity may represent an electrophysiological correlate of working memory performance in pediatric ADHD-I, though limited by the lack of healthy controls.