In Plastic and reconstructive surgery ; h5-index 62.0
BACKGROUND : Postoperative free flap monitoring is a critical part of reconstructive microsurgery. However, postoperative clinical assessments rely heavily on specialty-trained staff. Therefore, in regions with limited specialists, the feasibility of performing microsurgery is restricted. This study aims to apply artificial intelligence in postoperative free flap monitoring and validates the ability of machine learning in predicting and differentiating types of postoperative free flap circulation.
METHODS : Postoperative data from 176 patients who received free flap surgery were prospectively collected, including free flap photos and clinical evaluation parameters. The flap circulation outcome variables are normal, arterial insufficiency, and venous insufficiency. The Synthetic Minority Oversampling Technique plus Tomek Links (SMOTE-Tomek) was applied for data balance. Data were divided into 80%: 20% for model training and validation. SHapley Additive exPlanations was used for prediction interpretations of the model.
RESULTS : A total of 805 data were collected, 555 (69%) were normal flaps, 97 (12%) were flaps with arterial insufficiency, and 153 (19%) were flaps with venous insufficiency. The most effective prediction model was developed based on Random Forest with an accuracy of 98.4%. Temperature and color differences between the flap and the surrounding skin were the most significant contributing factors to predict a vascular compromised flap.
CONCLUSION : This study demonstrated the reliability of our machine learning model in differentiating various types of postoperative flap circulation. This novel technique may reduce the burden of free flap monitoring and encourage the broader use of reconstructive microsurgeries in regions with limited staff specialists.
Huang Ren-Wen, Tsai Tzong-Yueh, Hsieh Yun-Huan, Hsu Chung-Chen, Chen Shih-Heng, Lee Che-Hsiung, Lin Yu-Te, Kao Huang-Kai, Lin Cheng-Hung
2023-Feb-15
