Computer-Assisted Medical Training and Skill Assessment (Perk Tutor)



The past decade has witnessed a dramatic shift in medical education. Competency-based Medical Education (CBME), with a focus on outcomes, has been embraced at all levels of training and endorsed by the Royal College of Physicians and Surgeons of Canada, the Association of Faculties of Medicine of Canada and locally at Queen's University by the Undergraduate and Postgraduate Medical Curriculum Committees. At the same time there has been an exponential growth in the use of simulation in medical education which is being driven by a number of factors, not least of which is recognition that the time honored apprenticeship model of learning exclusively in the real clinical setting is no longer valid. Queen’s University finds itself at the forefront of both CBME and Medical Simulation. The Dean of the Faculty of Health Sciences, Dr. Richard Reznick, is recognized internationally as a pioneer in both CBME and medical simulation and he is committed to ensuring that Queen’s emerges as a leader in simulation-based training. Queen’s is the first to implement CBME across all of its specialty programs in 2017.

There is a growing recognition that medical education needs to move from a model of chance clinical encounters over a defined training period to a more structured model that defines observable abilities required by graduates, and then supports acquisition of these abilities through carefully planned and targeted educational programs. CBME represents a paradigmatic shift in medical pedagogy. Learners advance through stages of their training as they demonstrate competency, as opposed to spending a pre-determined length of time in each phase, thus ensuring that competent physicians move through their training in a quicker and more efficient manner. Competency needs to be assured by rigorous tracking and assessment. Traditionally, medical trainees practice on real patients under the supervision of senior physicians. The CBME model implies a volume of clinical encounters, practice and ongoing assessment that simply cannot be accomplished in the traditional clinical setting alone. In addition to real clinical exposure a competency-based model requires simulated encounters to ensure achievement of desired outcomes.

The Faculty of Health Sciences has established an 8,000 square foot state-of-the-art Clinical Simulation Centre in our new medical building. This center represents the cornerstone of our undergraduate and postgraduate CBME programs, specifically as they apply to technical skills, surgical skills and resuscitation medicine.  Over the past 10 or so years, faculty at Queen’s University have developed and implemented what is arguably the most comprehensive and sophisticated simulation-based curriculum in Canada. Our approximately 1,000 health care trainees from the schools of Medicine, Nursing and Rehabilitation Therapy all receive formal training in the Clinical Simulation Centre, often in an inter-professional setting. Simulation-based training begins in the first year of undergraduate training for all our students, and continues throughout. We are striving for a model where all procedures are practiced to the point of proficiency, prior to any real clinical encounter.



As part of the overall commitment to CBME and simulation-based training the School of Medicine and the School of Computing has been partnering over the past several years in the development of a comprehensive Medical Education Informatics program to achieve computerized objective assessment of medical skill and assessment with minimizing the dependence of human trainers and proctors and assessors. Simulated environments have been proven, through medical education research, to be superior for training in the early stages of learning. Many current training simulators merely replace the patient with a mannequin, oftentimes with a very sophisticated and expensive one, but they uniformly fail to provide quantitative assessment of competence and they require tutors and proctors to be present.   

At Queen’s, we pursue broad research into novel medical training systems that will allow CBME-compliant learning and certification without tutors and proctors. We are developing systems where arrays of miniature and non-contact sensors and instruments acquire a real-time flow of digital data allowing us to compute objective metrics of competence and provide immediate quantitative feedback to the learner. Then, during patient encounters, trainee performance is measured again using similar sensory instruments without interfering with patient care. We are blending physical reality with digital, virtual or holographic experiences during training, and gradually withdrawing these visual aids as the trainee progresses toward clinical competence.

Digitization of training and evaluation will ultimately enable confidential data management with the ability for data mining across trainee databases; a capability ultimately required for developing CBME-compliant national medical training curricula at all levels of medical education.



Needle placement is a universal manual skill in procedures such as injections, catheter placement or tissue biopsy. Needle placement training at Queen’s requires over 20,000 practices a year. In the era of skill-intensive medicine and at the arrival of CBME, the millennia-old apprenticeship model is impossible to support with tutoring and proctoring manpower.

We have been developing Perk Tutor, a highly customizable training platform that provides flexible software modules for trainee performance metrics and feedback in needle placement procedures. The task trainer models provide objective measures of tool and hand motion that instructors can use to track learning and assess competence.  Presently supported procedures include lumbar puncture, facet joint injection, prostate biopsy, central venous catheter placement, percutaneous nephrostomy, FAST (see images below.)  Procedures are practiced on modular task trainers, either developed in house based on imaging scans of actual patients or purchased, such as embodiments of the Blue PhantomTM in Figure 4.

Initial results with the Perk Tutor training platform suggest that trainees are more accurate and efficient in their learning when able to visualize the needle path rather than inserting in the traditional blind fashion.

All of these training models have been developed in a true multi-disciplinary fashion with input from a team that includes computer scientists and clinicians from the specialties of Critical Care Medicine, Anesthesiology, Radiology, Surgery and Urology. This broad input ensures that models are pedagogically sound and address the specific challenges faced by students in learning the procedure. Additionally, the clinicians involved in development are all leaders in curriculum development and delivery in the School of Medicine, and so are ideally situated to implement the models at the point in training where they are likely to have the greatest impact on student learning.



We are presently rolling out the Perk Tutor platform to training solutions for applications outside the percutaneous domain are underway, including colonoscopy, FAST, brachytherapy and neurosurgery.

We have began developing a computer-assisted tutoring system for soft tissue resection open surgery.

Other major new initiatives include quantitative measurement of individual and team performance in emergency situations and tutor-less and proctor-less training for open surgery skills such as soft tissue resection.

Importantly, during patient encounters, we will measure trainee performance using similar sensory instruments (hand/tool/eye tracking and pupillometry) without interfering with patient care, to validate how skill and competence acquired in simulation carries forward to patient encounters. A tutoring scheme arching from task trainer (phantoms) to patient care will be tested in epidural anesthesia training in summer 2017, for which research ethics approval has been already received.

Apart from our own preliminary work, we did not find any literature reference to data management associated with simulation-based medical education, an absolutely critical additive to CBME-compliant training. We propose a comprehensive trainee data manager system, with storage, query and statistical processing of trainee data, allowing medical education researchers to perform multivariate data analysis on outcomes and the efficacy of training schemes towards developing CBME programs. The data manager system must interface with the Queen’s online learning management system, where documentation, enrollment, grades, certificates, tracking and reporting are administered securely.

To date our team has had considerable success with grant applications and peer-reviewed publications. It is apparent, however, that success with agencies like NSERC and CIHR is more and more being tied to clear evidence of partnerships that will result in knowledge translation. In particular, granting agencies want to see a concrete plan that demonstrates how the knowledge generated in research will be used to improve the health of Canadians. With that in mind we are looking for academic and industry partners to join our efforts at the technology development level, at the testing level and finally in producing models that can be made available to teachers and learners in Faculties of Health Science both nationally and internationally; ultimately to achieve tutor/proctor-less training solutions and associated curricula.