Assessment and Training System for Micromanipulation Tasks in Surgery
Nanyang Technological University
Office:
Tel:
Fax:
N3.1-B2c-20
(65) 6790 4911 (Office)
(65) 6791 8591 (International)
(65) 6792 4062 (Local)
Co-Principal Investigator: Associate Professor Teo Chee Leong
National University of Singapore
Office:
Tel:
Fax:
PROJECT DESCRIPTION:
Limitation on human accuracy, due in particular to physiological tremor, has long been a concern in microsurgery, required in eye, hand and neuro-surgery. This restricts the types of microsurgical procedures that are feasible, and makes it necessary to train apprentice surgeons and assess their performances.
This project will investigate the causes of non-voluntary deviations hindering the quality of interventions, and develop a virtual reality based training system for training microsurgeons to perform manipulation tasks under a microscope.
We will first develop a comprehensive Virtual Reality setup dedicated to micro-manipulation, including a novel contact free optical measurement system with microscopic precision, a dedicated haptic interfaces and a virtual dynamic environment based on fast algorithms.
We will then investigate the factors that may affect motor deviation, such as grip force, exercise, practice and age, systematically. Assessment will be based on the following accuracy primitives: i) stationary test, ii) tracing test (path dependent), and iii) tracking test (trajectory dependent).
The micromanipulation learning approach we will use is based on representative surgical tasks decomposed into simple dexterity primitives, learned using various multisensory cues selected based on neuroscience knowledge. In particular, we will investigate predictive cues providing compensation for the visual delay, as well as haptic cues known to produce accuracy in arm movement.
The objective of this project is to:
- To design and implement an optical sensing system to measure and assess micromanipulation tasks under a surgical microscope;
- To quantify the impact of different factors (e.g. strenuous exercise, practice, consumption of alcohol and caffeine) on physiological tremor and the quality of micromanipulation;
- To design and realize a virtual reality (VR) setup with a dedicated haptic interface for the training of manipulation under an optical microscope;
- To study micromanipulation learning based on dexterity primitives and multisensory cues, and to develop an efficient VR-based training system for microsurgery. The targeted applications are microsurgery and cell micromanipulation.
1. Overview
There are two teams working on this project. The team led by Co-Principal Investigator Assistant Professor Ang Wei Tech at NTU is responsible for objectives (1) & (2). The team led by Co-Principal Investigator Associate Professor Teo Chee Leong at NUS is responsible for objectives (3) & (4). External collaborator Dr Etienne Burdet of Imperial College, UK is actively involved with both teams to provide expertise in Human Computer Interaction. Medical collaborator Associate Professor Lim Thiam Chye of NUH is also actively involved in the project to provide expertise in microsurgery assessment and training.
- Micro Motion Sensing System
- Micromanipulation Assessment
- Virtual Reality Training System
- Study of Micromanipulation Learning
Micro Motion Sensing System
Limitation on human accuracy is well-known to hinder certain procedures in micromanipulation and microsurgical application. The most common source of such inaccuracy is called physiological tremor. A suitable tool that has high-accuracy is necessary to measure and record physiological tremor.
To achieve this, a novel contact-free optical-based sensing system with microscopic precision was developed. The system employs mini-size laser diode to act as a hand-held instrument and Position Sensitive Detector (PSD) placed on an anti-vibration table. The laser light is shinned from the laser diode onto the PSD surface which is faced upward. The PSD detects the centroid position of the laser light spot falling onto its surface. Thus, laser spot position is known directly from the PSD facing upward. The working region of the sensor is the area sensitive of region, 10 x 10 mm.
As far as the evaluation of the manipulation accuracy in x-y plane is concerned, Z axis movement detection is not necessary.
The software setup and graphical user interface (GUI) were created in LabView platform. DAQ card was used to capture voltage signals from PSD. Assessment of micromanipulation would be based on the following three accuracy primitives: i) stationary test, ii) tracing test (path dependent), and iii) tracking test (trajectory dependent). For tracing test, eight very small black markers were printed on tracing paper that would then be pasted on PSD sensing surface.
The inherent non-linear nature of the optical property of the PSD may lower the sensing accuracy, especially with the introduction of extra layer to print eight markers. To remove non-linear behavior of the PSD, feed forward neural network is applied. Initially, laser diode was attached to high-precision motorized stages. Afterwards, the laser diode was moved to a known position both in x-y axes. The two dimensional data was captured and then input into the MATLAB to perform neural network training.
Close-up picture of PSD sensor with laser diode shooting on it (left); and 16x magnification under the microscope, showing tracing paper pasted on PSD area containing eight black markers for tracing task (right).
Tasks performed under microscopic view (A, C) and screen view (B, D). There are three different accuracy primitives involved: pointing, tracing certain path, and tracking.
GRANT:
- S$839,760, Agency for Science, Technology & Research (A*STAR), BMRC Grant 07/1/22/19/538.
2008 – 2011.
PERSONNEL:
Ph.D. (UC Berkeley), B.Eng (National University of Singapore)
Director, NUS Overseas Colleges
M Math, M Phys, Ph.D. Robotics (ETH-Zurich)
Senior Research Fellow, National University of Singapore
MBBS, FRCSE, FRCS (Surgical Neurology)
Head, Section of Neuro Trauma & Neurovascular Surgery
Co-Director, Neurocritical Care Unit
Prinicipal Investigator, Acute Brain Injury Research Lab
MBBS, Master of Medicine (General Surgery) (National University of Singapore)
Adjunct Associate Professor, Department of Orthopaedic Surgery, National University of Singapore
MBBS (University of Malaya), FRCS (Royal College of Surgeons, Edinburgh), JCAST (Academy of Medicine, Singapore), AM (Malaya), FAMS (Plastic Surgery)
Sr Consultant, Division of Plastic & Reconstructive Surgery, Yong Loo Lin School of Medicine, NUS
Visiting Plastic Surgeon, Changi General Hospitaland Kandang Kerbau Women's & Children's Hospital
Convener & Instructor for course in Singapore, Basic Surgical Skill Course with the Royal College of Surgeons of Edinburgh
PUBLICATIONS:
Refereed Conference:
- E. L. M. Su, W. T. Latt, W. T. Ang, T. C. Lim, C. L. Teo, and E. Burdet, “Micromanipulation accuracy in pointing and tracing investigated with a contact-free measurement system,” in Proc. 31th Annual Intl. Conf. IEEE Engineering in Medicine and Biology Society, Minneapolis, Minnesota, USA, Sep. 2009, (in press).
- E. S. Ananda, W. T. Latt, C. Y. Shee, E. Burdet, T. C. Lim, C. L. Teo, and W. T. Ang, “Study of the effect of visual feedback and speed on the accuracy of micromanipulation tasks,” in Proc. 31th Annual Intl. Conf. IEEE Engineering in Medicine and Biology Society, Minneapolis, Minnesota, USA, Sep. 2009, (in press).
- W. T. Latt, E. S. Ananda, S. C. L. Ong, K. C. Veluvolu, C. Y. Shee, and W. T. Ang, “Design and implementation of a two degree-offreedom micromanipulation assessment system,” in Proc. 30th Annual Intl. Conf. IEEE Engineering in Medicine and Biology Society, Vancouver, BC, Canada, Aug. 2008, pp. 5640–5643.
Refereed Conference (Submitted):
- None.