VR Lab Home
VR Lab Home See what's new Research See journal papers & proceedings Our virtual tools Meet virtual people Visit useful links Questions or comments?

Development of a Method for Surface and Subsurface Modeling Using Force and Position Sensors

Goal
The goal of this research is to develop a method that acquires force and position data from an object using sensors, and to then approximate the surface of this object using NURBS.



Abstract
Subsurface modeling of deformable objects, such as soft tissues and organs, involves the use of non-destructive methods of determining the properties of an object encased by a material. Some of the properties that can be determined from subsurface modeling include: shape, hardness, texture and possibly material. The ability to determine these properties is based on the accuracy of the method used and the properties of the surface encasing the object. As computers become more powerful and are able to produce even more realistic graphics, it will be possible to store and recreate precise duplicates of the original for later analysis. This paper will present a method of approximately modeling both the surface and an object below the surface of the skin by a method of palpation using both a position and force sensor, and then present this data in an interactive 3-D model. In addition, this method can be used by a physician, to perform preliminary diagnostic tests to determine if further medical tests are required.

Project Summary
Subsurface modeling of deformable objects, such as soft tissues and organs, involves the use of non-destructive methods of determining the properties of an object encased by a material. The ability to determine these properties is based on the accuracy of the method used and the properties of the surface encasing the object. As computers become more powerful and are able to produce even more realistic graphics, it will be possible to store and recreate precise duplicates of the original for later analysis.

Being able to detect and localize hard abnormal objects within tissue is a very essential procedure in medicine today, and is a reason why palpation is performed on patients (Peine et al. 1998). As of yet, the technology is nonexistent to combine the palpation process with the field of medical imaging. This paper will present a method of approximately modeling both the surface and an object below the surface of the skin by a method of palpation and then present this data in an interactive 3-D model using C++ and Open-GL.

The ModelGlove (patent pending) consists of a force sensor and a position sensor attached to the index finger, in which each device is interfaced to the computer via a serial port. Together, these devices allowed the program to determine the validity of the each data sample as it is collected, as well as the position at which the valid points are located.

Creation of both the surface and subsurface model involved similar steps. The only difference occurs in the way the data points are determined to be ‘valid’. The five basic steps required to generate both models are: data collection and validation, sort data, arrange data into groups, calculate control points, and generate NURB surface.

To validate the results obtained from the program developed, 4 experiments were performed that would test different aspects of the program. The experiments involved: tracing a 2-D curve with varying force threshold values, repeatedly reconstructing a 3-D surface, modeling surface and subsurface objects, and modeling general surfaces. An example of an object and its resulting model is shown on the left, while an example of a surface and subsurface model is shown on the right.


Cricket Bat (left) and resulting model (right)


Rubber Spleen under cotton batting (left) and resulting surface/subsurface model (right)

From the results obtained, it has been shown that it is possible to generate an approximate NURB model of a surface and subsurface based on scattered data that is acquired by a force and position sensor, while allowing the model to be saved for later viewing and comparison. In addition, it has been show that a physician can use the presented method as a preliminary diagnostic procedure, to determine if further medical exams are necessary or required.

Finally, combining the method presented in this thesis with the field of haptics can result in a method to not only help doctors analyze a situation, but help train future doctors. By adding the ability to feel what was modeled at a previous time, a doctor can make a better diagnosis of the problem or pass this information onto a specialist for later analysis.


People
Dr. T. Kesavadas (Dept. of Mechanical Engineering)
Kevin Smalley (Dept. of Mechanical Engineering)

Publications
[1] Smalley, K. & Kesavadas, T. "Development of a Method for Surface and Subsurface Modeling Using Force and Position Sensors", 2004 MMVR (in process)


VR LAB 809 FURNAS HALL, THE STATE UNIVERSITY OF NEW YORK AT BUFFALO, BUFFALO, NY 14260 | (716) 645-2593
VR LAB (C) 2003-2004, ALL RIGHTS RESERVED