Biomedizinische Technik. Biomedical engineering · 2025
Prediction of remaining surgery duration based on machine learning methods and laparoscopic annotation data.
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.
Kostopoulos Spiros, Cavouras Dionisis, Glotsos Dimitris, Loukas Constantinos
Аннотация
OBJECTIVES: The operating room is a fast-paced and demanding environment. Among the various factors involved in its optimization, predicting surgery duration is critical for scheduling and resource organization, ultimately resulting in improved quality of surgical care.
METHODS: We design, implement and evaluate a semi-automated machine learning method that takes as input the current phase and tools employed and provides prediction of the Remain Surgery Duration (RSD) in laparoscopic cholecystectomy operations. We use the annotated information of tools and phases provided in the publicly available dataset Cholec80. The method is based on a Random Forest regression model that considers two data streams: the surgical phase and the type of tools employed, at each time-frame of the operation. The data were split into Training-, Validation- and Test-sets. The Mean Absolute Error (MAE) was used as the performance metric for the various models examined.
RESULTS: Our approach managed to achieve a MAE=5.89 min across the overall duration of the surgeries in the test-set and MAE=4.61 min at 20 min before the end of the operation.
CONCLUSIONS: The employment of two separate regression models switched at a specific elapsed time threshold provides significant improvement in RSD prediction compared to other methods that process the video from the endoscope.
METHODS: We design, implement and evaluate a semi-automated machine learning method that takes as input the current phase and tools employed and provides prediction of the Remain Surgery Duration (RSD) in laparoscopic cholecystectomy operations. We use the annotated information of tools and phases provided in the publicly available dataset Cholec80. The method is based on a Random Forest regression model that considers two data streams: the surgical phase and the type of tools employed, at each time-frame of the operation. The data were split into Training-, Validation- and Test-sets. The Mean Absolute Error (MAE) was used as the performance metric for the various models examined.
RESULTS: Our approach managed to achieve a MAE=5.89 min across the overall duration of the surgeries in the test-set and MAE=4.61 min at 20 min before the end of the operation.
CONCLUSIONS: The employment of two separate regression models switched at a specific elapsed time threshold provides significant improvement in RSD prediction compared to other methods that process the video from the endoscope.