098 Developing an Intelligent tutoring system for prostate cryosurgery

Abstract

The objective of this study is to develop a proof-of-concept for a cryosurgery Intelligent Tutoring System (ITS), aimed at shortening the clinician’s learning curve, by integrating computer-generated cryosurgery planning, abnormal growth of the prostate gland, and established criteria for clinical success. Cryosurgery is the destruction of undesired tissue by freezing. Minimally invasive cryosurgery is performed by strategically placing an array of cryoprobes within a pre-specified target region. In prostate cryosurgery, the target region for destruction can be a portion of or the entire gland. A key to cryosurgery success is optimal selection and layout of cryoprobes, which maximizes destruction to the target region while minimizing cryoinjury to surrounding healthy tissues. Currently, selection of cryoprobe type, layout, sequence of operation, and procedure duration is based on the cryosurgeon’s personal experience and accepted practices. Suboptimal cryosurgery protocols may leave untreated areas in the target region, lead to cryoinjury of healthy surrounding tissues, increase the duration of the surgical procedure, and increase the likelihood of post-cryosurgery complications—all of which affect the quality and cost of medical treatment. An ITS guides the student through a problem-solving process by providing instruction and tailored feedback promptly. The cryosurgery ITS presents problems, strategically broken down into smaller steps, that are consistent with clinical practice and pedagogical rational. The ITS proposed in the current study focuses on determining cryoprobe layout, given an overall number of cryoprobes and their operational parameters. Optimal (expert) planning is determined by previously developed planning algorithms, with the trainee attempting to generate a layout that (1) matches the expert layout, with a predefined tolerance, or (2) generates a defect value lower than the expert defect; a defect is defined as either external tissue to the target area that was simulated to be cryoinjured, or internal tissue that was simulated to be uninjured. Key to ITS success is a set of rules and constraints used to evaluate a student’s solution, such as: (1) minimum distance from the urethra; (2) minimum distance from the prostate surface; (3) active cooling surface of a probe included within the gland; and (4) minimum probe layout defect value. ITS effectiveness is measured by evaluating students’ learning gains with increased practice. This study demonstrates a proof-of-concept for a fully-functioning cryosurgery ITS. The cryosurgery ITS prototype is composed of four primary components: a domain model, a student model, a tutor, and a problem interface. In order to establish an expert solution and to evaluate the training progress, the domain model uses: cryoprobe planning algorithms, the simulator, and a set of cryosurgery rules and constraints. The student model uses various performance metrics to map the student’s current knowledge. The tutor provides instruction and returns feedback to the student. The ITS interface provides a realistic problem-solving environment. The ITS integrates previously developed building blocks: a cryosurgery simulator, planning algorithms, and visualization tools. The ITS prototype was developed using Matlab. Source of funding: This study was supported by Award Number R01CA134261 from the National Cancer Institute. Conflict of interest: None declared. rabin@cmu.edu