Journal of surgical education · 2026
ADAMS: An Artificial Intelligence- Assisted Scoring System for Microanastomosis Training.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Hung Janet, Shakeel Muhammad Adnan, Hsiao Shih-Chi, Wang Yu-Chi, Catarzi Lisa, Kuo Yur-Ren
Аннотация
BACKGROUND: Artificial intelligence (AI) has transformed many fields of medicine but remains underutilized in microsurgical training. Existing microanastomosis assessment tools rely on fixed expert-defined criteria and faculty supervision, limiting scalability and objectivity. To address these limitations, the Artificial Intelligence-Developed and Assisted Microanastomosis Scoring System (ADAMS) was created to provide an AI-assisted, reproducible framework for competency assessment using static images of synthetic training models.
MATERIALS AND METHODS: ADAMS was developed through an iterative process using OpenAI's ChatGPT-5. The system assesses six domains-suture placement, bite symmetry, edge apposition, knot quality, knot tightness, and knot tails-each scored from 1 to 5, for a maximum of 30 points. Twenty surgical trainees performed end-to-end microanastomoses on synthetic latex simulation models using 9-0 sutures under surgical loupes. Standardized images were analysed by ADAMS, which automatically generated total scores and individualized feedback. Scores were categorized into five performance levels: High Proficiency (24-30), Competent (20-23), Developing Competence (18-19), Needs Improvement (10-17), and Novice (<10). AI-derived scores were compared with expert assessment to evaluate validity.
RESULTS: Twenty trainees were evaluated using ADAMS. Scores ranged from 6 to 20 (median 13), with most performances classified as Needs Improvement (n = 12, 60%) or Novice (n = 4, 20%). Three trainees (15%) were Developing Competence, one (5%) was Competent, and none reached High Proficiency. ADAMS demonstrated good agreement with expert assessment, with a Pearson correlation coefficient of r = 0.81 (p < 0.001), a mean score difference of 0.4 points, and a mean absolute difference of 2.2 points. The system also generated immediate case-specific recommendations, including optimizing suture spacing, improving bite depth, and refining knot tension.
CONCLUSIONS: This preliminary report of ADAMS scoring system represents a novel AI-assisted scoring framework for microanastomosis training using synthetic models. It standardizes assessment, provides immediate individualized feedback, objectively differentiates trainee performance, and demonstrates good agreement with expert assessment.
MATERIALS AND METHODS: ADAMS was developed through an iterative process using OpenAI's ChatGPT-5. The system assesses six domains-suture placement, bite symmetry, edge apposition, knot quality, knot tightness, and knot tails-each scored from 1 to 5, for a maximum of 30 points. Twenty surgical trainees performed end-to-end microanastomoses on synthetic latex simulation models using 9-0 sutures under surgical loupes. Standardized images were analysed by ADAMS, which automatically generated total scores and individualized feedback. Scores were categorized into five performance levels: High Proficiency (24-30), Competent (20-23), Developing Competence (18-19), Needs Improvement (10-17), and Novice (<10). AI-derived scores were compared with expert assessment to evaluate validity.
RESULTS: Twenty trainees were evaluated using ADAMS. Scores ranged from 6 to 20 (median 13), with most performances classified as Needs Improvement (n = 12, 60%) or Novice (n = 4, 20%). Three trainees (15%) were Developing Competence, one (5%) was Competent, and none reached High Proficiency. ADAMS demonstrated good agreement with expert assessment, with a Pearson correlation coefficient of r = 0.81 (p < 0.001), a mean score difference of 0.4 points, and a mean absolute difference of 2.2 points. The system also generated immediate case-specific recommendations, including optimizing suture spacing, improving bite depth, and refining knot tension.
CONCLUSIONS: This preliminary report of ADAMS scoring system represents a novel AI-assisted scoring framework for microanastomosis training using synthetic models. It standardizes assessment, provides immediate individualized feedback, objectively differentiates trainee performance, and demonstrates good agreement with expert assessment.