In vascular interventional radiology, procedures generally start with the Seldinger technique to access the vasculature, using a needle through which a guidewire is inserted, followed by navigation of catheters within the vessels. Visual and tactile skills are learnt in a patient apprenticeship which is expensive and risky for patients. We propose a training alternative through a new virtual simulator supporting the Seldinger technique: ImaGiNe (imaging guided interventional needle) Seldinger. It is composed of two workstations: (1) a simulated pulse is palpated, in an immersive environment, to guide needle puncture and (2) two haptic devices provide a novel interface where a needle can direct a guidewire and catheter within the vessel lumen, using virtual fluoroscopy. Different complexities are provided by 28 real patient datasets. The feel of the simulation is enhanced by replicating, with the haptics, real force and flexibility measurements. A preliminary validation study has demonstrated training effectiveness for skills transfer.

To validate a virtual reality simulator suitable for training Interventional Radiology (IR) skills. Task Analysis (TA) was employed to create a detailed description of a Liver Biopsy procedure, from which critical procedure steps (CPS) were identified. Engineers and computer scientists used the TA and CPS to build a simulator capable of measuring performance. Hierarchical and cognitive TA was carried out using interviews (n=12) with subject matter experts, and observation and recording of actual procedures (n=4). CPS were identified through interviews (n=12) and questionnaires (n=8). Engineers and computer scientists used the TA and CPS to build a training simulator capable of measuring performance. The discriminant validity of the simulator was investigated by measuring performance on a simulated Liver Biopsy procedure. The validation study was conducted at three UK clinical centres and participants were a convenience sample recruited on site with varying degrees of experience (consultants n=14, trainees n=26). The Liver Biopsy TA described174 procedural steps and 22 performance metrics were included in the simulator. Independent t-tests revealed significant differences between consultants (n=14) and trainees (n=26) on 4 performance metrics. ANOVA revealed significant differences between three groups with differing levels of experience in IR (< 1 year experience, 1-2 years, or 3+ years) on 7 performance metrics. Significant performance metrics were: no go area touched; targeting; length of session; probe usage time; total needle distance moved; number of skin contacts; total time in no go area. All 22 performance metrics followed the predicted pattern with level of performance consistently reflecting experience. The use of cognitive task analysis when developing training simulators allows the development of metrics that are a valid discriminator of skill between experts and novices. The indication is that simulation could be a useful training tool that can reflect performance levels on numerous procedural steps. Further validation work is needed to demonstrate transfer of training to the real world.