Key Technology And Experiment Forlocalizing An Implantable Detecting System In The Gastrointestinal Tract

The research of gastrointestinal motility depends on the monitoring and analysis of gastrointestinal motility parameters. However, there is lack of methods to long-time and dynamically monitor gastrointestinal motility parameters under normal physiological condition. Supported by Specialized Research Fund for the Doctoral Program of Higher Education, the National Natural Science Foundation of China and the 863 program, a noninvasive system including an implantable microcapsule for detecting gastrointestinal parameters was developed. Considering the location of the microcapsule in the gastrointestinal tract is invisible to doctors, which influences the diagnosis and treatment, a novel localization system was developed to track the microcapsule.The main purpose was to provide a noninvasive method to detect the gastrointestinal motility parameters under normal physiological condition. The overall scheme of the system was designed and the modular design ideas were determined. Main technical problems in the process of design and realization of this system were analyzed. The principle and methods of implementing the functions under rigid restriction of space and power consumption were presented. In order to test and verify the performace of the whole system, human experimentation was carried out by volunteers.In order to track the implantable microcapsule, a miniature object locating method was researched. To find the suitable localization method, dynamically monitoring methods which were used to measure transit process of gastrointestinal tract were discussed. At first, an ultrasonic localization method was presented. In the real test, the author found the insufficiency of the ultrasonic method. Then, three localization methods including a magnetic marker localization method, an excitation localization method using pulse current and a telemetric localization method by alternating magnetic field were focused on.In the magnetic marker method, a proper localization model was built up. According to magnetic finite element method, the dimension and the material of the magnetic marker were determined. Then a system sheme was determined. A complete system hardware and software platform was built. After preparation of the data processing program, the verifying experiment was performed. In the system design, a new type of magnetoresistive sensor was used to expand the localization distance. Furthermore, the impact of system noise to the objective locating precision was analyzed. A difference method was used to eliminate background magnetic field interference and improve localization accuracy. The experiment showed that the localization theory was feasible. However, the practical localization system failed to work reliably because of poor anti-jamming capability, small localization distance and bad stability.Then an excitation localization method using pulse current was researched. The localization principle was presented. Based on the principle, the localization sheme was determined and a system hardware and software platform was built. In the localization system, a magnetic sensor with high resolution was employed to expand the localization scope. The localization model was derived. A neural network algorithm was developed to solve the inverse problem in the magnetic field. The novel algorithm succeeded in resolving the position and orientation of the object. A verifying experiment was carried out. The experiment showed that the localization system could work well under magnetic shielding environment. Under non-magnetic shielding environment, however, the system failed to put into practice because of low localization accuracy and poor anti-jamming capability.A telemetric localization method using alternating magnetic field was newly presented. Focusing on the method, a localization principle was analyzed and a novel localization model was detaily derived. A simulation experiment was carried out to prove the correctness and superiority of the novel localization model. The localization algorithm is also researched. Because of rapid convergence speed and high precision, particle swarm optimization algorithm was determined as the localization algorithm to solve the complex problems in localization. The particle swarm optimization avoided some disadvantages of conventional algorithms, such as dependence of the initial value, trapping into local minimum and slow convergence. Then a portable localization system was developed. In the system design, a root-mean-square detection method based onΔΣcomputing technology was employed to extract the characteristics of the alternating signals. The novel technology avoided real-time dynamic waveform acquisition and simplified data collection, storage and processing. The experimental platform was preformed. Baesd on experimental research, the parameters of the system was optimized to improve the performance of the system. Some calibration algorithms were investigated, such as Hardy's multi-quadric interpolation method, high-order polynomial fit and bayesian-regularization neural network based on levenberg-marquart algorithm. Comparing the performance of the above-mentioned calibration methods, the bayesian-regularization neural network method was determined to improve the localization precision. The experimental results demonstrate that the localization system using alternating magnetic field is correct and feasible. The novel localization system has high accuracy, good repeatability, high stability and wide localization range, which can put into practice. The novel localization method can be applied to track similar types of implantable medical micro-devices.