An intelligent approach for automating robotic arm maneuvering in endometriosis surgery

Artificial Intelligence (AI) is transforming robotic surgery, yet autonomous maneuvering of robotic arms remains a critical challenge. This study presents an interpretable and reliable framework for enabling autonomous endometriosis surgeries, where robotic arms equipped with cameras and surgical tools (e.g., graspers, lasers) navigate human anatomical organs. A decision tree model is proposed that guides the robot’s actions at each surgical stage by evaluating its real-time status. To support this, ensemble deep learning models perform image classification and semantic segmentation, enhancing decision-making. These models assess image quality, detect obstructions from adhesions, identify anatomical targets such as the uterus and peritoneum, and determine the ovary’s proximity to the uterus. By integrating machine learning with surgical robotics, this approach enables the robot to navigate from the entry point of the camera and surgical instruments to the operative site, guiding the robotic arms to locate and examine organs. This integration enhances precision and autonomy in minimally invasive procedures, supporting assistant-based surgeries and serving as a core component of AI-driven automated gynecological surgeries.