RELIABLE AUTOMATED NEEDLE INSERTION SYSTEM FOR MEDICAL APPLICATION

Abstract

Many disorders occur annually as a result of poorly performed stings. This project is an attempt to develop a system that automates blood tests, serum injections and catheter placements, and to identify its basic limitations. Determining parameters are first identified. They include the coordinates of stinging point on the skin, the depth of blood vessel, its radius and the age of patient. The developed module performs the sting process based on the knowledge of these parameters. Automation is based on a neural network which correlates the data to determine insertion angle and needle geometry. Though the insertion process is adapted to patient profile, difficulties still remain concerning correct skin viscoelastic properties as proper input parameters. However, finer analysis of skin-needle system indicates the possibility of a secure and much easier automated sting in a large range of usual parameters with constant speed.