JMIR Medical Informatics, Год журнала: 2024, Номер unknown
Опубликована: Окт. 22, 2024
Язык: Английский
Cureus, Год журнала: 2024, Номер unknown
Опубликована: Май 8, 2024
Background Google Gemini (Google, Mountain View, CA) represents the latest advances in realm of artificial intelligence (AI) and has garnered attention due to its capabilities similar increasingly popular ChatGPT (OpenAI, San Francisco, CA). Accurate dissemination information on common conditions such as hypertension is critical for patient comprehension management. Despite ubiquity AI, comparisons between remain unexplored. Methods were asked 52 questions derived from American College Cardiology's (ACC) frequently hypertension, following a specified prompt. Prompts included: no prompting (Form 1), patient-friendly 2), physician-level 3), statistics/references 4). Responses scored incorrect, partially correct, or correct. Flesch-Kincaid (FK) grade level word count recorded. Results Across all forms, scoring frequencies follows: 23 (5.5%) 162 (38.9%) 231 (55.5%) showed higher rates correct answers than (p = 0.0346). Physician-level prompts resulted across both platforms < 0.001). FK 0.033) physician-friendly prompting. exhibited significantly mean 0.001); however, had forms > Conclusion To our knowledge, this study first compare cardiology-related responses Gemini, two most AI chatbots. The was collegiate level, which above average National Institutes Health (NIH) recommendations, but par with online medical information. Both chatbots responded high degree accuracy, inaccuracies being rare. Therefore, it reasonable that cardiologists suggest either chatbot source supplementary education.
Язык: Английский
Процитировано
17
Journal of Clinical Hypertension, Год журнала: 2024, Номер 26(5), С. 588 - 593
Опубликована: Апрель 22, 2024
Язык: Английский
Процитировано
9
Journal of Clinical Hypertension, Год журнала: 2025, Номер 27(1)
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
0
Digital Health, Год журнала: 2025, Номер 11
Опубликована: Янв. 1, 2025
Background: The use of artificial intelligence (AI) for interpreting ambulatory blood pressure monitoring (ABPM) data is gaining traction in clinical practice. Evaluating the accuracy AI models, like ChatGPT 4.0, settings can inform their integration into healthcare processes. However, limited research has been conducted to validate performance such models against expert interpretations real-world scenarios. Methods: A total 53 ABPM records from Mayo Clinic, Minnesota, were analyzed. 4.0's compared with consensus results two experienced nephrologists, based on American College Cardiology/American Heart Association (ACC/AHA) guidelines. study assessed ChatGPT's and reliability over rounds testing, a three-month interval between rounds. Results: achieved an 87% identifying hypertension, 89% nocturnal 81% dipping, 94% abnormal heart rate. correctly identified all conditions 60% records. percentage agreement first second round analysis was 85% There no significant difference (all p > 0.05). Kappa statistic 0.63 0.66 0.76 0.70 Conclusions: 4.0 demonstrates potential as reliable tool 24-h data, achieving substantial nephrologists. These findings underscore hypertension management workflows, while highlighting need further validation larger, diverse cohorts.
Язык: Английский
Процитировано
0
Journal of Clinical Hypertension, Год журнала: 2024, Номер 26(7), С. 872 - 873
Опубликована: Июнь 14, 2024
Язык: Английский
Процитировано
1
Atención Primaria, Год журнала: 2024, Номер 56, С. 103123 - 103123
Опубликована: Ноя. 1, 2024
The recommendations of the semFYC's Program for Preventive Activities and Health Promotion (PAPPS) prevention vascular diseases (VD) are presented. New in this edition new sections such as obesity, chronic kidney disease metabolic hepatic steatosis, well a 'Don't Do' section different pathologies treated. have been updated: epidemiological review, where current morbidity mortality CVD Spain its evolution main risk factors described; (VR) calculation CV risk; arterial hypertension, dyslipidemia diabetes mellitus, describing method their diagnosis, therapeutic objectives lifestyle measures pharmacological treatment; indications antiplatelet therapy, screening atrial fibrillation, management conditions. quality testing strength recommendation included recommendations.
Язык: Английский
Процитировано
1
Renal Failure, Год журнала: 2024, Номер 46(2)
Опубликована: Сен. 11, 2024
ChatGPT, a state-of-the-art large language model, has shown potential in analyzing images and providing accurate information. This study aimed to explore ChatGPT-4 as tool for identifying commonly prescribed nephrology medications across different versions testing dates.
Язык: Английский
Процитировано
0
Journal of Nephrology, Год журнала: 2024, Номер unknown
Опубликована: Дек. 2, 2024
Язык: Английский
Процитировано
0
Bioengineering, Год журнала: 2024, Номер 12(1), С. 17 - 17
Опубликована: Дек. 28, 2024
Current clinical care relies heavily on complex, rule-based systems for tasks like diagnosis and treatment. However, these can be cumbersome require constant updates. This study explores the potential of large language model (LLM), LLaMA 2, to address limitations. We tested 2′s performance in interpreting complex process models, such as Mayo Clinic Care Pathway Models (CPMs), providing accurate recommendations. LLM was trained encoded pathways versions using DOT language, embedding them with SentenceTransformer, then presented hypothetical patient cases. compared token-level accuracy between output ground truth by measuring both node edge accuracy. 2 accurately retrieved diagnosis, suggested further evaluation, delivered appropriate management steps, all based pathways. The average across different 0.91 (SD ± 0.045), while 0.92 0.122). highlights LLMs healthcare information retrieval, especially when relevant data are provided. Future research should focus improving models’ interpretability their integration into existing workflows.
Язык: Английский
Процитировано
0
JMIR Medical Informatics, Год журнала: 2024, Номер unknown
Опубликована: Окт. 22, 2024
Язык: Английский
Процитировано
0