Development of a computerized system for esophageal bouginage.
| Project/Area Number |
02670534
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
General surgery
|
|---|
| Research Institution | Dept. of Medical Technology, College of Medical Technology and Nursing, University of Tsukuba |
|---|
Principal Investigator |
SAKANIWA Misao Dept. of Medical Technology, College of Medical Technology and Nursing, University of Tsukuba Professor, 医療技術短期大学部・衛生技術学科, 教授 (40134233)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
OHTA Michio Institute of Engineering Mechanics, University of Tsukuba Professor, 構造工学系, 教授 (10016446)
|
|---|
| Project Period (FY) |
1990 – 1991
|
| Project Status |
Completed (Fiscal Year 1991)
|
| Budget Amount *help |
¥2,200,000 (Direct Cost: ¥2,200,000)
Fiscal Year 1991: ¥900,000 (Direct Cost: ¥900,000)
Fiscal Year 1990: ¥1,300,000 (Direct Cost: ¥1,300,000)
|
|---|
| Keywords | esophageal stenosis / neural network / bouginage / viscoelasticity / microcomputer / Neural Network / ファジィ理論 |
|---|
| Research Abstract |
The aim of this research is to develop a new system for esophageal bouginage. The basic idea for this system is that the viscoelasticity of the esophageal wall, especially the stricture lesion, is to be measured before and during the dilatation procedure. This is valuable to monitor the effectiveness and the safety. A three-lumen silastic balloon was used as the dilator. A computerized control system that inflates and deflates the dilator with water, measures pressure changes and calculates the elasticity of the esophageal wall was newly assembled. The active dilating force produced by the balloon and the effects of the dilatation were needed to be calculated from the balloon pressure changes. Accordingly the neural network was introduced into the system. The network was a software basis, and had two sets of three layers ; input layers, intermediate layers and output layers. 'Back propagation' was employed as the Teaming algorithm. Sets of three parameters, i. e., pressure values at the stricture lesion and at the healthy area and the amount of water injected into the balloon were fed into the neural network. Then the neural network calculated the extension forces and extents of the stretch in the esophageal wall. The calculated values were within 5% approximately the measured values. The canine model experiments showed that this system with the neural network worked successfully. Further experiments are under way.
|
Report
(3 results)
Research Products
(6 results)
