Surgical gesture classification using Dynamic Time Warping and affine velocity.

Area of Science:

• Surgical Technology
• Biomedical Engineering
• Motion Analysis

Background:

• Minimally Invasive Surgery (MIS) offers patient benefits but requires specialized skills, posing challenges for objective surgical gesture assessment.
• Kinematic and kinetic analysis of hand movements are valuable tools for evaluating basic surgical skills in MIS.
• Surgical motion can be modeled as a sequence of constant affine velocity movements.

Purpose of the Study:

• To present a novel method for classifying surgical gestures in MIS.
• To utilize affine velocity analysis of 3D motion combined with the Dynamic Time Warping algorithm.
• To reduce the dimensionality of 3D motion analysis for improved gesture classification.

Main Methods:

• Developed a novel method employing affine velocity analysis of 3D motion data.
• Implemented the Dynamic Time Warping algorithm for gesture classification.
• Correlated kinematic and geometric variables (curvature, torsion, Euclidean velocity) for affine velocity calculation.

Main Results:

• The proposed method simplifies 3D motion analysis by reducing dimensionality.
• Experimental validation using laparoscope instrument data demonstrated accurate gesture classification.
• The method showed advantages over conventional Multidimensional Dynamic Time Warping for MIS gesture classification.

Conclusions:

• The novel affine velocity analysis combined with Dynamic Time Warping offers an accurate, simple, and understandable approach to surgical gesture classification in MIS.
• This method provides a valuable tool for the objective assessment of surgical skills in Minimally Invasive Surgery.
• The technique has the potential to enhance surgical training and performance evaluation.