Generalized approach for modeling minimally invasive surgery as a stochastic process using a discrete Markov model

Area of Science:

• Surgical Technology
• Medical Robotics
• Surgical Training

Background:

• Minimally invasive surgery (MIS) requires integrating visual data with surgical tool kinematics and dynamics.
• Objective performance metrics are crucial for surgical skill assessment.
• Existing methods lack comprehensive analysis of the multidimensional tasks in MIS.

Purpose of the Study:

• To introduce the Blue DRAGON system for synchronized kinematic and dynamic data acquisition in MIS.
• To develop an objective method for surgical performance assessment using finite state modeling.
• To define an objective learning curve for surgical training based on task decomposition.

Main Methods:

• The Blue DRAGON system recorded tool kinematics and dynamics synchronized with endoscopic video.
• A finite state model, specifically a Markov model (MM), was employed to represent surgical tasks.
• Thirty surgeons of varying expertise performed intracorporeal knot tying on a porcine model.

Main Results:

• The Markov model effectively decomposed the complex task of laparoscopic suturing.
• An objective learning curve was established by comparing the MMs of expert surgeons and residents.
• The derived objective learning curve correlated with subjective performance assessments.

Conclusions:

• The Markov model provides a powerful mathematical framework for analyzing MIS tasks.
• The Blue DRAGON system and MM-based analysis enable objective surgical performance evaluation.
• This methodology can enhance surgical robotics and virtual reality simulators by incorporating objective performance feedback.