A Real-time augmented reality robot integrated with artificial intelligence for skin tumor surgery - experimental study and case series
Kai Huang,Jun Liao,Jishuai He,Sicen Lai,Yihao Peng,Qian Deng,Han Wang,Yuancheng Liu,Lanyuan Peng,Ziqi Bai,Nianzhou Yu,Yixin Li,Zixi Jiang,Juan Su,Jinmao Li,Yan Tang,Mingliang Chen,Lixia Lu,Xiang Chen,Jianhua Yao,Shuang Zhao
DOI: https://doi.org/10.1097/js9.0000000000001371
2024-03-28
International Journal of Surgery
Abstract:Background: Skin tumors affect many people worldwide, and surgery is the first treatment choice. Achieving precise preoperative planning and navigation of intraoperative sampling remains a problem and is excessively reliant on the experience of surgeons, especially for Mohs surgery for malignant tumors. Materials and Methods: To achieve precise preoperative planning and navigation of intraoperative sampling, we developed a real-time augmented reality (AR) surgical system integrated with artificial intelligence (AI) to enhance three functions: AI-assisted tumor boundary segmentation, surgical margin design, and navigation in intraoperative tissue sampling. Non-randomized controlled trials were conducted on manikin, tumor-simulated rabbits, and human volunteers in xxx Laboratory to evaluate the surgical system. Results: The results showed that the accuracy of the benign and malignant tumor segmentation were 0.9556 and 0.9548, respectively, and the average AR navigation mapping error was 0.644 mm. The proposed surgical system was applied in 106 skin tumor surgeries, including intraoperative navigation of sampling in 16 Mohs surgery cases. Surgeons who have used this system highly recognize it. Conclusions: The surgical system highlighted the potential to achieve accurate treatment of skin tumors and to fill the gap in global research on skin tumor surgery systems.
surgery