Virtual-reality Simulation-based Training in Ultrasound

Lars Konge,E. Albrecht-Beste,M. Nielsen

DOI: https://doi.org/10.1055/s-0034-1366191

2014-04-01

Abstract:The traditional apprenticeship model is being challenged in medical education. The “see-one, do-one, teach-one” approach does not live up to the demands regarding efficient training and patient safety. Ultrasound imaging is traditionally considered safe but the quality of the exam is very operator-dependent due to its dynamic nature 1. Highly skilled operators are essential as false negative findings could lead to inadequate investigation and treatment, and false positive findings might result in unnecessary interventions 2. This is true in particular because sonographic competence is a basic requirement for US as a dialog-based examination 3. Most trainees participate in defined ultrasound courses but specialized ultrasound examination expertise is mainly acquired in daily clinical practice 4 5. Fundamental problems are random occurrence of specific cases and also insufficient formalized supervision and feedback may occur due to scarce personnel resources 6. Ultrasound diagnostics quality assurance is a main concern for the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) 7 8. However, the current training conditions are likely to impair the quality 9. Virtual-reality (VR) simulation training has been mandatory in aviation for decades and is rapidly gaining importance in medical education. VR simulators have been developed for a multitude of different medical procedures, each consisting of a procedure-specific interface and a computer that generates images according to the movements of the user. A simulator has potential to deliver a safe, controlled and stress-free learning environment providing different performance scores as automatic feedback. Furthermore, standardized “patient”-cases allow for a systematic assessment of competence and simulation-based mastery learning 10. A systematic review and meta-analysis of more than 600 papers found large effects of VR simulation for outcomes of knowledge, skills, and behaviors, but very few of these papers related to ultrasound simulators 11. Several VR ultrasound simulators have become commercially available over the last years 12 13 14 15 16 17. A single system uses a haptic device to control the probe and the others use physical patient phantoms combined with electromagnetic tracking to track the position of the probe. The haptic device allows measuring the force that is applied but has a limited working range making probe movements less realistic. Electromagnetic tracking allow a greater range of motion on the patient phantom but the absence of haptics makes it possible to generate simulated images without actually touching the phantom. Furthermore, image deformation as a result of pressure applied to the “patient” is not simulated. The simulators use two different methods to generate images: interpolative or generative model-based. Interpolative simulation creates 2-dimensional ultrasound images from previously recorded 3-dimensional volumes acquired from real patients. It is relatively easy to acquire many different cases with very realistic images, but it is difficult to simulate view-dependent artifacts correctly. Generative model-based simulation uses manually generated computer models to create images, which makes detailed 4-dimensional simulation possible, e. g. echocardiography. However, the images are less realistic (“cartoon-like”) and building multiple cases is very time consuming 18. No simulator will ever be 100 % realistic and which simulator(s) to acquire depends largely on personal preferences and should be guided by thorough testing before purchase. Unfortunately, there is a paucity of high-quality scientific studies exploring the effects of the different VR ultrasound simulators. A recent systematic review only identified 14 articles and half of these were prospective non-controlled studies 19. Only five of 14 studies contained evaluation of a study population – two of these used randomization. Another serious limitation concerning the available literature is that the vast majority of studies explored training-related skills acquisition in the same simulated environment, i. e. training on a simulator improves performance on that simulator. Caution has to be used before generalizing these results: Whether the improvement can be directly transferred into daily clinical practice is doubtful. There is a huge need for high-quality psychometric studies gathering validity evidence concerning the different simulators and for randomized controlled trials (RCTs) exploring the effect of ultrasound simulator training. Performing RCTs in medical education poses several challenges 20. A comparison between intervention and no intervention is meaningless as any kind of training is likely to be better than no training at all. Due to the limited number of physicians it can be difficult to identify a sizable, homogenous group of equally experienced trainees waiting to learn ultrasound diagnostics. Furthermore, special attention must be paid to the outcome parameters used to examine the effect of different training regimens. Reliable and valid assessment instruments should be used 21 22 and blinded, video-based assessment should be preferred over direct observation to reduce bias 23. Overall, VR simulator training has a huge potential to improve training and education in ultrasound. Ultrasound simulators could be used in a highly standardized learning curriculum consisting of repetitive training on selected cases of increasing difficulty, including controlled exposure to rare cases. By providing automatic feedback the simulators can reduce the time that experts have to be present although they can never completely replace skilled supervisors. Finally, simulation-based tests including credible pass/fail standards may aid in the certification of future ultrasound operators 24. However, it is important to realize that the existing literature does not justify the huge organizational and financial investments that are necessary to make simulator training and certification a mandatory part of the education of future ultrasound operators. Implementation should be preceded by research focusing on the quality of simulator feedback, the reliability and validity of simulation-based tests, and the transfer of acquired skills into clinical practice.