Design and evaluation of perception reconstruction with sensor system for artificial anal sphincter based on vector similarity.

Abstract

Artificial organs are playing an increasingly important role in medical field. Artificial anal sphincter, as an example, is a widely used medical device for fecal incontinence. Though it could help patients maintain continence, the issue of perception reconstruction is still unsolved, which means that patients cannot control defecation as desired. In this paper, a novel sensor system based on vector similarity has been analyzed and evaluated. The sensor system is a self-packaged strain gauge sensor with a small size 14.5 mm × 6 mm, especially for medical device implanted in body to reconstruct perception function. In order to overcome the disadvantages of single threshold judgment, a more accurate and reliable judgment standard for content pressure detection in rectum is developed by using vector similarity method. Therefore, in the novel sensor system, standard multi-parameters are comprehensive considered. Furthermore, weight-optimization of sensors are investigated since each sensor has played a different role in detection. Finally, sensor calibration, in vitro and in vivo experiments are established. In sensor calibration, the result R2 > 0.99 presents the strain gauge pressure sensor has a good linearity. After that, a series of in vitro experiments have been conducted. The optimized sensor system shows a high accuracy rate in perception reconstruction, which is 87.5% on early warning and 95% on the alarm. In the following in vivo experiments. The results indicate that the average accuracy of the proposed perception reconstruction module has increased by 17.7%, compared with that without optimized sensor system. All of these have proven that the novel perception reconstruction module with sensor system based on vector similarity is more acceptable and reliable.