Computer algorithm for automated detection of intramedullary rod hole position and orientation / Ahmad Zulhilmi Mohd Ziyadi

In minimal invasive orthopedic procedure such as distal locking, a lot of surgeon's time is usually wasted to determine the hole's position of intramedullary rod before screw can be insert at desire place. Often patient expose to x-ray until the surgeon get it right. The gap between manual distal locking procedure and fully automatic distal locking procedure is huge. The biggest and hardest is the system must able to recognize an X-ray image and all the relevant content inside it. Therefore to close the gap between manual and automatic, automatic image recognition system using machine vision algorithm must be developed. To obtain the intramedullary rod's holes screw coordinate (X, Y, Z) in cost effective and reasonable real time period using machine vision. Taken into account this is early research period, real surgery x-ray image is not use because of complexity and uncertainty involved. Therefore simulated x-ray image is use in control lab environment. This research is to automatically determine the holes position, distance and orientation of intramedullary rod in distal locking procedure by using in-process 3D machine vision image recognition system. It is achieve by using special algorithm that detect holes 2D coordinates and average width of intramedullary rod. After that the algorithm converting the average width value to Z-coordinate. Angle Z and angle Y are than determined through the X,Y,Z coordinates obtain earlier.