Sun Young Yi

New colonoscopy simulator with improved haptic fidelity

Colonoscopy is a safe and effective procedure to diagnose and treat the large bowel with the help of the flexible endoscope. This paper presents a new colonoscopy training simulator to help trainees practice and acquire the necessary skills and experiences with no risk to the patients and possibly less cost. The simulator includes a specialized haptic interface to transfer force feedback through a long and flexible tube, and graphics algorithms to display the virtual colon realistically while managing the large number of polygons. A new 2-d.o.f. haptic device with folding guides is developed to transmit large decoupled forces of the colonoscopy simulation to the user. The physicians apply a jiggling motion to the colonoscopy tube to advance the scope. This jiggling is an important skill of colonoscopy and is incorporated for the first time by using the new sensor mechanism. A colonoscope handle that shares the look, feel and functions with an actual colonoscope is developed with the necessary electronics inside. The simulator contains controllers to compensate for the inertia and friction effects, and is evaluated by physicians. New graphics algorithms including polygon reduction, navigation and collision detection are developed to compute the deformation and the corresponding reflective force in real-time.

Haptic Interface of the KAIST-Ewha Colonoscopy Simulator II

This paper presents an improved haptic interface for the Korea Advanced Institute of Science and Technology Ewha Colonoscopy Simulator II. The haptic interface enables the distal portion of the colonoscope to be freely bent while guaranteeing sufficient workspace and reflective forces for colonoscopy simulation. Its force-torque sensor measures the profiles of the user. Manipulation of the colonoscope tip is monitored by four deflection sensors and triggers computations to render accurate graphic images corresponding to the rotation of the angle knob. Tack sensors are attached to the valve-actuation buttons of the colonoscope to simulate air injection or suction as well as the corresponding deformation of the colon. A survey study for face validation was conducted, and the result shows that the developed haptic interface provides realistic haptic feedback for colonoscopy simulations.

Model of frictional contact with soft tissue for colonoscopy simulator

This paper presents a model of frictional contact with soft tissue for the colonoscopy simulator with haptic fidelity. The colonoscopy simulator allows the user to navigate the three dimensional virtual colon constructed from CT (computer tomography) data of patients. The frictional contact model is proposed to integrate the therapeutic procedures such as polyp removal. The proposed model uses the finite element method based on the linear static elasticity to simulate deformation and reflective force including friction effects. The variation of the reflective force is computed according to the contact state, stick or slip state, during sliding contact motions. Simulation results show responses following the Coulombs friction law.

Improved Haptic Interface for Colonoscopy Simulation

This paper presents an improved haptic interface of the KAIST-Ewha colonoscopy simulator II. The haptic interface enables the distal portion of the colonoscope to be freely bent while guaranteeing enough workspace and reflective forces for colonoscopy simulation. Its force-torque sensor measures profiles of the user. Manipulation of the colonoscope tip is monitored by four deflection sensors, and triggers computation to render accurate graphic images corresponding to the angle knob rotation. Tack switches are attached on the valve-actuation buttons of the colonoscope to simulate air-injection or suction, and the corresponding deformation of the colon.

Design of a Haptic Interface for a Gastrointestinal Endoscopy Simulation

Abstract This paper presents a new design and analysis of a haptic interface for a gastrointestinal endoscopy simulation. The gastrointestinal endoscopy is a procedure in which the digestive tract and organs of a patient are diagnosed and treated using a long and flexible endoscope. The developed haptic interface incorporates two degrees of freedom (DOF), each of which is necessary to describe the movements of an endoscope during the actual endoscopy procedures. The haptic interface has a translational motion mechanism to implement the insertion movement of the endoscope, and a rotational motion mechanism to implement the rotational movement of the endoscope. The endoscope included in the haptic interface is supported by a folding guide to prevent the endoscope from buckling. Force feedback in each direction is provided by wire-driven mechanisms. The developed haptic interface has a workspace, sensitivity, and maximum attainable force and torque enough to simulate the endoscopy procedures such as colonoscopy, upper GI (gastrointestinal) endoscopy, and endoscopic retrograde cholangiopancreatography (ERCP). The developed haptic interface is applied to implementation of a colonoscopy simulation. Performance including force bandwidth is evaluated through experiments and simulation.

Su1246 Temporal Trends and Natural History of Eosinophils-Related Gastrointestinal Disease in Asia

BACKGROUND: Colitis describes inflammation in the colon and is classified into defined diseases: inflammatory bowel disease (IBD), microscopic colitis, iatrogenic colitis, ischemic colitis, and infectious colitis. Accurate classification helps guide management, but this is often difficult. Nonspecific colitis, or also known as atypical colitis, describes the cases that do not show characteristic features of a specific colitis. The clinical significance of nonspecific colitis is unclear as there are limited studies. Notteghem et al. (1993) showed that out of 104 patients with nonspecific colitis, 52.3% had another episode within 3 years and 54% were diagnosed with ulcerative colitis, 33% with Crohns, and 13% remained unclassified, suggesting that nonspecific colitis could potentially be undiagnosed IBD. PURPOSE: The objective of this study is to understand their clinical course by examining subsequent colonoscopies and determining if any presenting symptom can predict subsequent diagnosis. METHODS: Patients who had colonoscopies from 1/2004-12/2006 with biopsies showing nonspecific colitis were enrolled in this retrospective observational study. The inclusion criteria are biopsies from colonoscopies showing nonspecific colitis, follow up with a physician for at least one year, and no previous diagnosis of a specific colitis based on history or biopsy. A chart review of the electronic record was done from the time of colonoscopy until current time or the last note. RESULTS: 101 patients were included in the study with a mean follow-up period of 5.09 years. The most common indications for colonoscopy were diarrhea (44.6%), abdominal pain (29.7%), blood in stool (36.6%), and anemia (7.9%). 50.5% of patients had at least one subsequent colonoscopy with 68.6% of them with normal biopsies, 33.3% with nonspecific biopsies, and 15.7% with biopsies consistent with IBD. No presenting symptom was predictive of subsequent biopsies with IBD, such as abdominal pain (OR, 2.4, 95% CI, 0.57-10.5, p = 0.23) or blood in stool (OR, 3.2, 95% CI, 0.7114.2, p = 0.13). 19 out of 101 patients with nonspecific colitis biopsies were treated for IBD given symptoms (18.8%), and 6 out of the 19 patients had subsequent biopsies showing IBD (31.6%). CONCLUSIONS: 50.5% of patients who have biopsies with nonspecific colitis had a subsequent colonoscopy, and 68.6% of subsequent colonoscopies were normal while 15.7% had IBD. Therefore, although most commonly nonspecific colitis is self-limited colitis, there is a portion that are undiagnosed IBD. In fact, 18.8% of patients with nonspecific colitis were empirically treated for IBD given symptoms, and 31.6% of these patients showed IBD on subsequent colonoscopies.

S1909 Gender Specific Risk Factors for Reflux Esophagitis in Asia: Role of Age, Body Mass Index, Metabolic Syndrome and Menopause

Introduction: Nonacid reflux has been shown to be able in inducing reflux symptoms in gastroesophageal reflux disease (GERD) patients. Non-erosive reflux disease (NERD) and erosive esophagitis (EE) are the most frequent phenotypic presentations of GERD. Aim: To evaluate symptom association probability (SAP) in patients with NERD and EE. Methods: We evaluated 44 patients with EE (26M, median age 48; range 23-73 yrs), 173 with NERD (67M, median age 49; range 18-78 yrs) and 48 controls (22M, median age 44; range 2277) using multichannel intraluminal impedance and pH while off-PPI therapy. All patients had typical reflux symptoms. We measured distal esophageal acid exposure time (AET), number of impedance-detected reflux episodes (acid, nonacid) and symptom association probability (SAP; positive if ≥95%). Results: Symptom-reflux association with SAP in EE and NERD patients is shown in the Figure below. Out of 35 (75%) EE patients with an abnormal AET, 33 (75%) had a positive SAP. Twenty-seven (61%) patients had a positive SAP for acid reflux only, 2 (5%) for nonacid reflux only and 4 (9%) for both. Among the 8 (18%) patients with a normal AET and a positive SAP, 3 (7%) had a positive SAP for nonacid reflux only, 2 (5%) for acid reflux only and 3 (7%) for both. Out of 72 (42%) NERD patients with an abnormal AET, 64 (37%) had positive SAP. Fifty-two (31%) patients had a positive SAP for acid reflux only, 8 (5%) for nonacid reflux only and 6 (3%) for both. Among the 101 (58%) patients with a normal AET, 59 (34%) had a positive SAP. Twentytwo (13%) patients had a positive SAP for nonacid reflux only, 23 (13%) for acid reflux only and 14 (8%) for both. Conclusions: Nonacid reflux clearly contributes to provoke reflux symptoms in both NERD and EE patients. The increased percentage of NERD patients whose symptoms are related to nonacid reflux could be due to the higher sensitivity of these patients to stimuli different from acid.

W1837 Stress and Fatigue in Reflux Esophagitis

Background Typical symptoms of gastroesophageal reflux disease (GERD) are heartburn and regurgitation. Extraesophageal (EE) symptoms such as cough and throat clearing are less commonly associated with GERD but represent a common presenting symptom for reflux testing. Proximal extent of refluxate into the larynx or even the lungs is proposed as a potential mechanism for EE symptoms. Aim We tested the hypothesis that EE symptoms are more frequently associated with proximal esophageal reflux than typical symptoms. Methods A blinded assessment was rigorously applied. Themultichannel intraluminal impedance-pH (MII-pH) tracings analyzed for this study were edited by one of the investigators so that none of the symptom markers were visible and the reflux associated symptoms reviewed were simply marked as symptom 1-5 for each patient. 815 consecutive MII-pH reports were reviewed to identify 40 patients with at least 5 symptom events preceded by a reflux episode, thus a total of 50 reflux related events in each of 4 symptom groups: cough, throat clearing, heartburn, and regurgitation. A second investigator blindly analyzed all 200 episodes by reporting the extent of the refluxate at 7, 9, 15, and 17 cm above the LES. Results The percentage of symptom related reflux extending proximally to 17 cm above the LES is similar among all four symptom types (Figure 1). At least 50% of all symptoms were associated with proximal reflux to 17 cm above the LES, with the numerically highest being regurgitation (60%). There is no significant difference between any of the values. Conslusion This study confirms the frequent association of proximal reflux with GERDrelated symptoms previously identified. However, as opposed to common beliefs, our data indicate that EE GERD-related symptoms are not more frequently associated with proximal reflux than typical GERD symptoms. Both EE and esophageal symptoms are associated with reflux to 17cm at least 50% of the time.

Colonoscopy Simulator with Enhanced Haptic Realism and Visual Feedback

This paper presents the KAIST-Ewha colonoscopy simulator II with increased haptic fidelity as well as visual feedback. The colonoscopy simulator includes a specially developed haptic interface and optimized colon models constructed from CT data of actual patients. The colonoscopy simulator provides with enhanced haptic realism, which is made possible by employing a haptic device with decoupled translational and roll force-feedback mechanisms. Manipulation of the colonoscope tip is monitored by four deflection sensors, and triggers computation to render accurate graphic images corresponding to the knob rotation. A photo-sensors unit is developed to simulate jiggling motion which is one of the important colonoscopy skills to fold and straighten the colon. Friction in the haptic device is compensated by using force-torque sensors, and the force-torque profile during simulation is also logged by these sensors. The colonoscopy simulator employs a centerline-based parametric modeling of the colon for fast and accurate collision detection and haptic response between the colon and the colonoscope. The developed Colon Modeler provides with patient-customized colon models, and various training scenarios including rare cases.

Clinical Evaluation of a Colonoscopy Simulator with Improved Haptics

This paper presents a clinical evaluation of a newly-developed colonoscopy training simulator that includes a specialized haptic interface to transfer force-feedback through a long and flexible colonoscope tube. The colonoscopy simulator has a 2-DOF haptic device with folding guides, which can transfer large decoupled forces of the colonoscopy simulation to the user. Centerline-based parametric colon models are developed to compute collision detection and response in real-time. The clinical study included 22 subjects consisting of colonoscopy experts (n=3), fellows (n=4), residents (n=6), and medical school students (n=9). The study shows that the developed colonoscopy simulator can effectively improve the skills of the residents and medical school students, although no statistically significant training-effects can be established for the fellows