Cardiovascular and interventional radiology · 2026
Personalizing Bone Tumor Ablation: Decision Models from the University Hospital of Strasbourg.
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.
Cazzato Roberto Luigi, Severac Francois, Steffen Dominik A, Mandolini Matilde, Orkut Sinan, Gangi Afshin
Аннотация
PURPOSE: To describe the clinical decision-making underlying the selection of the percutaneous ablation modality for bone tumors.
MATERIALS AND METHODS: Two experienced interventional radiologists designed in consensus a model for ablation modality selection. Consecutive patients with primary and metastatic bone tumors were retrospectively identified. Demographic, clinical, and imaging data were collected for each tumor. The dataset was randomly divided into a training set (618/926 tumors; 66.7%) and a test set (308/926 tumors; 33.3%). A decision-tree model derived from the training set and the model proposed by the experts were both verified on the test set. Models' performance was assessed.
RESULTS: The decision-tree and the experts-based models achieved comparable accuracy (81.8% vs. 80.5%; p = 0.493). For cryoablation, the experts-based model demonstrated significantly higher sensitivity (94.0% vs. 87.5%, p = 0.008), whereas the decision-tree model showed significantly higher specificity (79.3% vs. 69.3%, p = 0.001). For radiofrequency ablation, the decision-tree model exhibited substantially higher sensitivity (35.8% vs. 7.5%, p < 0.001), while the experts-based model had higher specificity (98.4% vs. 93.7%, p = 0.004). Performance for interstitial laser ablation was comparable between the two models, with no significant differences in sensitivity (98.8% vs. 98.9%, p = 1) and specificity (94.1% vs. 95.0%, p = 0.500).
CONCLUSION: The experts-derived and the decision-tree models demonstrated comparable overall accuracy despite relying on the different decision-making. Prospective multicenter studies are required to determine which of these two approaches is more suited for implementation in clinical practice.
MATERIALS AND METHODS: Two experienced interventional radiologists designed in consensus a model for ablation modality selection. Consecutive patients with primary and metastatic bone tumors were retrospectively identified. Demographic, clinical, and imaging data were collected for each tumor. The dataset was randomly divided into a training set (618/926 tumors; 66.7%) and a test set (308/926 tumors; 33.3%). A decision-tree model derived from the training set and the model proposed by the experts were both verified on the test set. Models' performance was assessed.
RESULTS: The decision-tree and the experts-based models achieved comparable accuracy (81.8% vs. 80.5%; p = 0.493). For cryoablation, the experts-based model demonstrated significantly higher sensitivity (94.0% vs. 87.5%, p = 0.008), whereas the decision-tree model showed significantly higher specificity (79.3% vs. 69.3%, p = 0.001). For radiofrequency ablation, the decision-tree model exhibited substantially higher sensitivity (35.8% vs. 7.5%, p < 0.001), while the experts-based model had higher specificity (98.4% vs. 93.7%, p = 0.004). Performance for interstitial laser ablation was comparable between the two models, with no significant differences in sensitivity (98.8% vs. 98.9%, p = 1) and specificity (94.1% vs. 95.0%, p = 0.500).
CONCLUSION: The experts-derived and the decision-tree models demonstrated comparable overall accuracy despite relying on the different decision-making. Prospective multicenter studies are required to determine which of these two approaches is more suited for implementation in clinical practice.