Next-Generation Tactile Sensing and Machine Learning Integration for Robot-Assisted Minimally Invasive Surgery

Area of Science:

• Robotics
• Surgical Technology
• Biomedical Engineering

Background:

• Tactile feedback is essential in robot-assisted minimally invasive surgery (RAMIS) for surgeons to accurately palpate subsurface structures.
• Current RAMIS systems often lack adequate haptic feedback, limiting the surgeon's ability to discern tissue properties like softness and texture.

Purpose of the Study:

• To develop and evaluate a new system for generating tactile sensations in RAMIS.
• To enable accurate detection of tissue deformation and texture during surgical procedures.

Main Methods:

• Data acquisition using micro-electromechanical systems (MEMS) and force-sensitive resistor (FSR) sensors on a da Vinci Surgical System grasper.
• Digital signal processing for feature extraction from sensor data.
• Training and testing of machine learning algorithms (Reflex Fuzzy Min-Max Neural Network and Time Series Classification - Learning Shapelets) for tissue classification.
• Implementation of a visual-tactile display and wearable device for surgeon feedback.

Main Results:

• The proposed system successfully extracts relevant features from sensor data.
• Machine learning algorithms accurately classify physiological structures based on softness and roughness.
• The feedback system effectively mimics palpation sensations for surgeons.

Conclusions:

• The developed tactile feedback system enhances sensory information in RAMIS.
• This technology has the potential to improve surgical outcomes by providing crucial haptic data.
• Further integration into surgical platforms can advance the capabilities of robotic surgery.