Proceedings of the 2009 IEEE Systems and Information Engineering Design Symposium, University of Virginia, Charlottesville, VA, USA, April 24, 2009 FPM2HF.Health.2 A Virtual Reality Interface to Provide Point Interaction and Constriction to the Finger Kimberly A. Everett, Rachael E. Exon, Sylvia H. Rosales, and Gregory J. Gerling, Member, IEEE Abstract— Virtual reality (VR) simulation of tube thoracostomy may improve the procedural training for medical and nursing students. Current VR simulators, however, do not provide tactile feedback, which is essential for enabling certain tasks (e.g., surface palpation to identify rib location, blunt dissection for access to pleural space surrounding the lungs, and finger sweep to confirm location in the pleural space). This work develops a physical apparatus that provides users with point feedback at the fingertip when palpating an external surface and a sensation of constriction around the finger during insertion into a body. The physical apparatus is composed of two components that separately control the constriction on the tip and middle of the finger. Each constriction component is made of two nylon casings coated with a silicone-elastomer that enclose about the top and bottom of the finger. DC gearhead motors control the magnitude of pressure in proportion to feedback from force transducers embedded in the siliconeelastomer. The device is intended to communicate with a virtual environment (written in H3D). The apparatus augments traditional stick-based force feedback and should enhance the learning of tactile tasks in tube thoracostomy. T I. INTRODUCTION thoracostomy (chest tube insertion) is a medical procedure in which a tube is inserted into the pleural space to drain excess fluid, blood, or air, thereby relieving pressure from the lungs [1]. All medical and nursing students must learn to insert chest tubes. At present, the procedure has a complication rate of 30% [2]. As a supplement to current training materials, virtual reality simulation can provide an environment for repeated practice and for learning the overall cognitive steps. A preliminary computer simulation of chest tube insertion coupled with a haptic device has been developed [3], [4]. However, key tasks in the cognitive sequence cannot be performed at present because the simulator cannot adequately render tactile forces to the bare finger. The primary goal of this work is to develop the functionality to enable tasks involving tactile palpation. For instance, during chest tube insertion, the clinician must locate the correct location for the incision UBE Manuscript received April 6, 2009. K. A. Everett is a fourth year student in Systems and Information Engineering (SIE) and Biomedical Engineering at the University of Virginia, Charlottesville, 22904 (phone: 434-924-0533, e-mail: [email protected]). R. E. Exon is a fourth year student in SIE at the University of Virginia, Charlottesville, 22904 (e-mail: [email protected]). S. H. Rosales is a fourth year student in SIE at the University of Virginia, Charlottesville, 22904 (e-mail: [email protected]). G. J. Gerling is an assistant professor in SIE at the University of Virginia, Charlottesville, VA 22904 (e-mail: [email protected]). 1-4244-4532-5/©2009 IEEE by feeling ribs through the skin. Another tactile task occurs during blunt dissection, when the clinician places his index finger into the incision hole to confirm the trajectory of the path. Third, after the pleural space has been punctured, the clinician performs a ‘finger sweep’ by pushing the tip of his index finger into the pleural space and rotating his bent finger to ensure placement [1], [5], [6]. The virtual reality apparatus designed and built in this work seeks to enable the performance of these tactile steps. A. Prior Art in Virtual Reality Simulation in Medicine Virtual reality (VR) simulation has been employed to teach other types of surgery and is increasingly popular as a training tool. VR simulation can provide students with graduated levels of difficulty, train students for rare but dangerous events, and also provide objective feedback [7], [8]. VR training has been most frequently employed for laparoscopic surgeries, as they necessitate the use of videogame-like controls in the procedure itself. Preliminary studies have found that surgeons who practice laparoscopic cholecystectomy on VR devices make fewer mistakes in actual surgery [9]. VR tools have also been developed for training basic surgical skills, in addition to their usage in training bone surgery, cataract surgery, and cardiac catheterization, among others [10]. B. Prior Art in Bare Finger Interface Devices One mechanism developed to constrict the bare finger was developed by Inaba and Fujita. The user places his finger in between two contact plates that are connected by a loop of Velcro. A driving motor rotates a shaft, which tightens the contact belt. The two contact plates are pulled together, creating constriction [11]. Kramer developed an interface that transmits force, texture, temperature, and pressure to the entire arm. This invention includes a glove that can sense position and dynamically change the force exerted on each finger through a feedback control loop. The pressure applied to the hand is generated by a motor or pneumatic/hydraulic fluid [12]. While these types of devices constrict a user’s finger, none were built to mimic the sensation of being inside the human body nor are simple enough to incorporate with a SensAble OMNI or the H3D Software Virtual Environment. II. METHODS A motor-controlled constriction device with forcefeedback capabilities was designed and built to enable tactile and palpation tasks of chest tube insertion. The device is 203 composed of two motor-driven components that control constriction of the tip and middle of the finger to achieve different sensations depending on the task. The levels of constriction are achieved through feedback from the device. ensuring that there are no organs adherent, and that the tube can be inserted safely [1], [5], [6]. B. Requirements The device was built to meet a set of requirements specific to the three palpation tasks and to the overall usability of the device. 1) Task Requirements: Req. 1: The device shall provide point interaction on the user’s fingertip within the palpation task. Req. 2: The device shall constrict and release to reach the appropriate pressure on the finger, which is defined in each task, and do so within two seconds. Req. 3: The device shall allow for individual control of the two components that encase the finger. Fig. 1. The University of Virginia Chest Tube Insertion Simulator A. Task Analysis The physical VR interface will add tactile and palpation feedback to address three tasks that require interaction via the forefinger: surface palpation of the ribs before incision, blunt dissection, and finger sweep. A task analysis with physicians was done to better understand the goals and subtasks that underlie each task and the typical errors of novice learners. 1) Surface Palpation: To identify the correct site for incision, the physician must palpate the fifth intercostal space (the space between the fifth and sixth rib) [13]. Interviews with resident physicians identified that positional errors are commonly made by making too low an incision, resulting in a tube inserted in the abdomen [5]. During the surface palpation task, the clinician feels point interaction on his fingertip with a change in resistance depending on whether the finger is moved over a rib or muscle. There are no constrictive forces on the finger. 2) Blunt Dissection: After making an incision, the physician uses a Kelly clamp to bluntly dissect a tract toward the lungs. We observed that clinicians switch between using the Kelly clamps and using their index finger during the blunt dissection step and emphasized that students should do the same. The forefinger plays a large role in confirming the location of the tract. During blunt dissection the clinician feels the resistance of moving through, over, or around different types of tissue. When the finger enters the body the clinician feels constrictive forces on the tip of his finger, but not upon the middle of the finger. When the finger is completely inside the body, the clinician feels constrictive forces along the length of the finger. 3) Finger Sweep: After dissecting the tissue and muscle between the ribs, the physician punctures the pleural space. The physician then inserts his finger into the pleural cavity and performs a sweep of the space with his fingertip, 1-4244-4532-5/©2009 IEEE 2) General Requirements: Req. 4: Once inserted, one’s finger shall be aligned with and close to the stereo plug of the haptic device to minimize discrepancy between how the user perceives his finger in the physical and virtual worlds. Req. 5: The device shall accurately mimic the feeling of tissue on a user’s finger when interacting with the virtual body. Req. 6: Appropriate safety controls shall ensure that any constriction will stop before injuring a user’s finger. Req. 7: The placement of one’s finger in the device shall be intuitive and easy. Req. 8: The device shall be lightweight to allow the user to easily move one’s hand in space. Req. 9: The device shall fit in a 16 cm by 10 cm by 12 cm space in order to fit easily on the table. Req. 10: The device shall not impede the range of motion of the force feedback device, which allows a range of motion of a hand pivoting at the wrist. Req. 11: The device shall attach to a SensAble PHANTOM Omni haptic device. Req. 12: The device shall accommodate various finger lengths and diameters. 204 Fig. 2. Task Flow Diagram C. System Overview and Base Functionality The physical apparatus was designed to be mounted on the stereo plug of a SensAble PHANTOM Omni (Fig. 1). The apparatus provides users with both point feedback and constriction, and consists of two components that separately control the constriction on the tip and middle of the finger. Basically, the device is designed to react to the change in location in the virtual body. During the surface palpation task, both constriction components will tighten equally until there is very slight pressure on the user’s finger (Fig. 2). This will ensure contact so that point interaction feedback will be transferred. During blunt dissection, both motors will constrict to a higher level of pressure. As the user moves his finger into the pleural space, the fingertip component will relax, allowing the user to perform a sweep. D. Hardware Design 1) Finger Casings: The device’s main structure consists of two constricting components, one for the tip of the finger and one for the middle of the finger (Fig. 3 and Fig. 4, Label c). Both components have an upper and lower casing made of Nylon 6-6. The inner surfaces of both casings are coated in a solid silicone-elastomer (5005 BJB Enterprises with 50% C), with a thin layer on the flat interior surface and a thicker layer on the inside of the half-cylinders (Fig. 4, l) (Implementation Element 1). The silicone provides for the feeling of tissue during constriction of the finger. A force sensor (Flexiforce 0-1 lb, Tekscan, South Boston, MA) is embedded in the silicone of the lower casings (Fig. 3 and Fig. 4, d). Control of the casings is achieved by driving the lower casings, which glide upon aluminum rods, towards the fixed upper casings (Fig. 3 and Fig. 4, b). Various finger sizes are accommodated by the size of the cylinders cut in each casing and the adjustable distance between the top and bottom casings (Implementation Element 2). The stereo jack attaches the device to the existing PHANTOM Omni (Implementation Element 3), and does not impede the range of motion that already exists. Point-interaction was fulfilled by aligning the fingertip nylon casing with the stereo jack (Fig. 3, k), and by creating a silicone wall at the end of the fingertip casing (Implementation Element 4). 2) Drive Mechanism: Two DC gearhead motors are used to drive the lower casings towards the upper casings, one for each constriction component (Fig. 4, p). When constriction of a component is desired, the motor turns, which screws a bolt through the horizontal aluminum plate (Fig. 4, m). A flexible 24-inch socket extender (Implementation Element 5) is used to separate the heavy motors from the device. As the user’s finger is compressed, the force generated on the pressure sensor, embedded in the silicone-elastomer, controls the duration for which the motor turns (Fig. 3 and Fig. 4, d). Thus, the communication between the flexi-force sensors, STAMP code, and the motors (Implementation 1-4244-4532-5/©2009 IEEE 205 Fig. 3. Side-view of the device. (a) ¼” Aluminum nuts (b) Aluminum bolts threaded at each end (c) Nylon finger casings (d) Flexi-Force Sensor (e) Aluminum plates (f) Flexible socket extension (g) Steel bracket (h) Rivets (i) 3/16” Aluminum nuts (j) Nylon block (k) Radio jack Fig. 4. Straight-on view of the device. (a) ¼” Aluminum nuts (b) Aluminum bolts threaded at each end (c) Nylon finger casings (d) Flexi-Force Sensor (e) Aluminum plates (f) Flexible socket extension (h) Rivets (l) Silicon lining (m) Square head screw (n) Set screw (o) Nylon rod (p) Motor Fig. 5. Electronic Component Schematic for One Motor Table I. Results Matching Requirements to Implementation Elements 1-4244-4532-5/©2009 IEEE 206 Element 6) allows for the adjustable constriction of the finger based on the current pressure. This helps adjust for different finger sizes, as the running time for the motor may change with different users, but the resulting pressure is the same. Implementation Elements 2 and 6 also allow for the constriction of different parts of the finger depending on the task. E. Electrical Component Design A Basic STAMP microcontroller is used to control the two motors and to read the pressure exerted at the interior of the silicone casings. The analog signals from both force transducers are amplified by INA114 instrumentation amplifiers and converted from analog to digital with an ADC0831 (Fig. 5). The resulting signal is input into the microcontroller program. The two motors are wired in Hbridge circuits, which allow them to be turned in either direction. Each H-bridge is composed of two NPN transistors (TIP120) and two PNP transistors (TIP125). When one NPN and one PNP transistor are connected to opposite sides of the motor and turned on, the motor turns in one direction. The transistors are connected to the microcontroller output by opto-couplers, which protect the microcontroller from the rest of the circuit. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] III. RESULTS AND DISCUSSION The design of the VR finger interface is able to fulfill both the task and general requirements (Table I). The immediate next steps are to design a safety mechanism to protect against injury and to perform usability studies with potential users. The interface will then be integrated with the virtual environment. The development of point interaction and constriction feedback to the finger will allow the user to complete palpation tasks, thereby increasing the student’s knowledge and comfort with these tasks. The finger interface will join the Kelly clamp in what will one day be a series of appliances to augment the virtual sensations provided by the haptic device pen (Fig. 5). These appliances will increase the ability of the simulator to mimic the chest tube insertion tasks. When the simulation is completed, it is expected that medical students, nursing students, and graduate medical students will practice with the simulator, in addition to current training. The repetition allowed by the simulator will better enable them to learn the procedure and make fewer mistakes when inserting chest tubes in patients. [12] [13] ACKNOWLEDGMENTS The authors would like to thank Jacob Scanlon for his guidance. We would also like to thank Professor Reba Moyer Childress of the University of Virginia School of Nursing, and Dr. Marcus L. Martin of the University of Virginia School of Medicine for their clinical advice related to teaching and performing the tube thoracostomy procedure. 1-4244-4532-5/©2009 IEEE 207 E. Reichman , R. Simon. Emergency medicine procedures: Text and atlas. New York, McGraw-Hill Professional; 2003, pp. 226-232. C. Ball, J. Lord, K. Laupland , S. Gmora, R. Mulloy, A. Ng, et al. “Chest tube complications: How well are we training our residents?” Canadian Journal of Surgery. 2007; 50(6):450. B. Cline, A. Badejo, I. Rivest, J. Scanlon, W. Taylor, G. Gerling. “Human performance metrics for a virtual reality simulator to train chest tube insertion.” IEEE Systems and Information Design Symposium , University of Virginia. 2008; 168-73. N. Raja, J. Schleser, W. Norman, C. Myzie, G. Gerling, M. Martin.“Simulation framework for training chest tube insertion using virtual reality and force feedback.” IEEE systems, man and cybernetics conference, Montreal, Canada. 2007: 2261. S. Tropello. Emergency Medicine Resident at University of Virginia. Personal Interview. October 2008. M. Fitzgerald, C. Mackenzie, S. Marasco, R. Hoyle, T. Kossmann. “Pleural decompression and drainage during trauma reception and resuscitation.” Injury. 2009; 39(1):9-20. M. Agus, E. Gobbetti, G. Pintore. “Real-time cataract surgery simulation for training.” Eurographics Italian Chapter Conference. 2006. G. Megali, S. Sinigaglia, O. Tonet, P. Dario. “Modelling and evaluation of surgical performance using hidden markov models.” IEEE Transactions on Biomedical Engineering. 2006;53(20):1911. A. Gallagher, C. Cates. “Virtual reality training for the operating room and cardiac catheterization laboratory.” Lancet. 2004; 364: 1538-40. C. Sewell, D. Morris, N. Blevins, S. Dutta,S. Agrawal, F. Barbagli, et al. “Providing metrics and performance feedback in a surgical simulator.” Computer Aided Surgery. 2008;13(2):63-81. G. Inaba, K. Fujita. “A Pseudo-Force-Feedback Device by Fingertip Tightening for Multi-Finger Object Manipulation.” Proceedings of the Virtual Reality Society of Japan Annual Conference. 2006. 11: 2A3-5. J. Kramer. “Force feedback and texture simulating interface device.” U.S.Patent 5631861. 20 May 2007 S. Dev, B. Nascimiento, C. Simone, V. Chien. (2008). “Videos in clinical medicine: Chest-tube insertion.” New England Journal of Medicine. [Video]. 357(15). Available: http:content.nejm.org/cgi/video/357/15/e15/
© Copyright 2026 Paperzz