![]() ![]() Our customer is optimistic about proceeding further with this project using our pathfinding and proof-of-concept materials to demonstrate the viability of this approach. Going forward we are now emphasizing the depth of the catheter tip in our display, using fiducials to track the catheter tip in real time. This pivot involved tracking the tip of the catheter in the presence of full occlusion compared to previous work tracking the microwire. We were able to take feedback on board through a pivot near the end of our project. Customer Feedback Our response from the customer, Dr. Compared to the 5 dot scenario, the depth error is better on average, with the median error being below the 1 mm target, showing that the goal is indeed met. In both cases, the error is near or below 5 mm. The error is portrayed in a box plot with the depth error on the left and position error on the right. This figure represents the error within our program during the analysis of the dot with 0.5 cm of spacing between the dot, with a total of 9 dots on the catheter. However, it must be noted here that our considerations for further development (indicated in the next section) may drastically improve this outcome. We found that the maximum percentage of time the catheter tip will be located with 1mm accuracy is 70% of the time. Figure X below shows a plot of chances ranging from 10% to 90%, using the calculated error data. In other words, we wanted to find the maximum percentage of time that our algorithm will locate the catheter tip with 1mm accuracy (error<1mm). Once errors were obtained for each frame (process described below), we performed a binomial distribution for a range of chance values. The overall goal was to assess if the ground truth catheter location was within 1mm accuracy of our software. Testing Results We tested our software to assess if our code was able to accurately locate the catheter tip. This video illustrates the system detecting, extrapolating and overlaying an estimated location of the hidden catheter tip in the aneurysm phantom setup.Ĭatheter Imaging System Prototype In-Action Final Prototype Setup and Outputs The final output of our prototype is demonstrated in the video below. Overall, our system works by pre-processing, detecting the circular markers on the catheter, and using this marker data and extrapolation to best estimate the position of the occluded catheter tip. The third image below demonstrates the final setup for our physical system in which the phone camera is taking images/videos of the aneurysm on the light table while the catheter is being moved in and out of the aneurysm with occlusions are being added and removed.Ĭomputational Detection and Estimation Process Below is a flowchart that outlines our image analysis and estimation process using MATLAB. Next, the aneurysm phantom was placed on a light table facing up towards a iPhone camera mount. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |