SIMBiopsies : An Augmented Reality Training SIMulator for Needle Biopsies

INTRODUCTION Tissue biopsies, the procurement of small tissue samples, remain the gold standard for assessing the health of internal organs such as kidney, liver, and bone. Extracted tissues are used to find abnormal cells (e.g. cancers), investigate symptoms (e.g. ulcers, hepatitis, kidney disease or endometriosis), or inflammation [1]. The anatomical location of the tissue dictates the selection of the biopsy instrument (e.g., curette, punch, needle, endoscope), additional visualization modalities (e.g. CT/MR, ultrasound, fluoroscope), and in turn the biopsy setting (e.g. outpatient clinic, operating-room). Typically clinicians learn to do biopsies by observing more experienced personnel actually performing them. Then the student clinicians are ‘talked through it’ by the supervising clinician as the student performs the procedure on a live patient needing that particular biopsy. Students are shown the anatomical landmarks (visual clues) that guide placement of the biopsy, and urged to ‘feel’ the different tissues as the needle encounters or passes through them. In rare instances, novice clinicians are subjectively evaluated for such things as dexterity, speed, coordination and skill by an experienced examiner before the student is allowed to perform the procedure unassisted [2]. But, only few programs have such examinations because of, among other things, the time and expense involved and the unpredictable availability of subjects (patients) needing the particular biopsy. Consequently, training of these common, important, but potentially risky procedures are not uniformly taught or often repeated. The result is inadequately or poorly trained practitioners, who then go on to poorly train others [3]. In this work, we will focus on needle biopsies, wherein long needles are inserted through cutaneous tissues and body wall to obtain samples of lung, liver, prostate, kidney, etc. Several of the more intricate biopsies such as those of the kidney, prostate, or liver are difficult because of the risk of injury to other organs. These procedures require considerable training in both cognitive and sensorimotor skills to successfully execute. The sheer number of biopsies performed and the variety of clinical specialties performing these procedures argue compellingly for a more comprehensive, quantitative, computer-based training, testing, and certification regimen for resident training. Within the medical education community, there is also growing awareness of the need for a quantitative skillbased assessment of residents to justify credentialing and hence the need for such virtual trainers [4]. However, such biopsy simulators are still in their infancy due to limitations in visual and haptic technologies, lack of suitable assessment metrics, and, most importantly, lack of accreditation/validation/ certification of these methodologies. Our motivation to create and deploy such a simulator arises from several reasons, including: (i) inadequate conventional training due to various economic and logistical issues; (ii) marked differences in learning among trainees using current training techniques; (iii) evidence directly linking trainee improvement to duration, regularity, realism and diversity of training sessions; (iv) a growing need for procedural training that closely couples cognitive with sensorimotor training, and (v) improved understanding and technological progress in kinesthetic human-computer interactions that make feasible lowcost implementation of such a simulator. Thus, in this work we propose to develop and validate an Augmented Reality Training SIMulator for Needle Biopsies (ARSIMBiopsies) that replicates both the look (graphics) and feel (haptics) of an actual biopsy triaining.