Large Language Models-Assisted Diagnosis of Catecholaminergic Polymorphic Ventricular Tachycardia in a Pediatric Cardiac Arrest Patient

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 10, 2025

Background: Catecholaminergic polymorphic ventricular tachycardia (CPVT), a rare hereditary ion channel disorder, is triggered by exercise or stress, causing PVT and sudden death. Diagnosis tough, especially with cardiac arrest as the initial symptom, guideline - recommended adrenaline may worsen it. Large language models offers new ways to identify it quickly. Case Presentation: A 7 year old boy had during rope skipping. After resuscitation, defibrillation, adrenaline, his circulation returned, but persisted. VA ECMO in our hospital couldn't control arrhythmia. Multidisciplinary discussion was inconclusive. ChatGPT DeepSeek suggested CPVT. stopping catecholamines using beta blockers, arrhythmias decreased. Gene testing confirmed an RYR2 gene mutation (c.6737C>T), diagnosing However, due long term late diagnosis, delayed ECMO, child developed severe complications. Despite successful weaning, parents gave up treatment, died. Conclusion: CPVT patients are critically ill hard diagnose. Early detection targeted treatment vital for prognosis. have value diagnosis should be combined clinical judgment further tests. For children unexplained arrest, assisted consultation can considered, clinicians better understand diseases like more timely accurate diagnoses. Clinical trial number No applicable.

Language: Английский

---

Grounding Large Language Model in Clinical Diagnostics

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 15, 2025

Large language models (LLMs) possess extensive medical knowledge and demonstrate impressive performance in answering diagnostic questions. However, responding to such questions differs significantly from actual clinical procedures. Real-world diagnostics involve a dynamic, iterative process that includes hypothesis refinement targeted data collection. This complex task is both challenging time-consuming, demanding significant portion of workload healthcare resources. Therefore, evaluating enhancing LLM real-world procedures crucial for deployment. In this study, framework was developed assess LLMs' capability complete encounters, including history, physical examination, tests diagnosis. A benchmark dataset 4,421 cases curated, covering rare common diseases across 32 specialties. Clinical evaluation methods were used comprehensively the GPT-4o-mini, GPT-4o, Claude-3-Haiku, Qwen2.5-72b, Qwen2.5-34b, Qwen2.5-14b Although these performed well questions, they consistently underperformed exhibited number errors. To address challenges, ClinDiag-GPT trained on over 8,000 cases. It emulates physicians' reasoning, collects information line with practice, recommends key definitive diagnoses. outperformed other LLMs accuracy procedural performance. We further compared alone, collaboration physicians, physicians alone. Collaboration between enhanced efficiency, demonstrating ClinDiag-GPT's potential as valuable assistant.

Language: Английский

---

A sepsis diagnosis method based on Chain-of-Thought reasoning using Large Language Models

Journal of Applied Biomedicine, Journal Year: 2025, Volume and Issue: 45(2), P. 269 - 277

Published: April 1, 2025

Language: Английский

---

Artificial intelligence in regional anesthesia

Current Opinion in Anaesthesiology, Journal Year: 2025, Volume and Issue: unknown

Published: April 21, 2025

Purpose of review Artificial intelligence (AI) is having an increasing impact on healthcare. In ultrasound-guided regional anesthesia (UGRA), commercially available devices exist that augment traditional grayscale ultrasound imaging by highlighting key sono-anatomical structures in real-time. We the latest evidence supporting this emerging technology and consider opportunities challenges to its widespread deployment. Recent findings The existing literature limited heterogenous, which impedes full appraisal systems, comparison between devices, informed adoption. AI-based promise improve clinical practice training UGRA, though their patient outcomes provision UGRA techniques unclear at early stage. Calls for standardization across both AI are increasing, with greater leadership required. Summary Emerging applications warrant further study due opaque fragmented base. Robust consistent evaluation reporting algorithm performance, a representative context, will expedite discovery appropriate deployment UGRA. A clinician-focused approach development, evaluation, implementation exciting branch has huge potential advance human art anesthesia.

Language: Английский

---