Automated Prediction of Hepatic Arterial Stenosis
Several thousand life-saving liver transplants are performed each year. One of the causes of early transplant failure is arterial stenosis of the anastomotic junction. Early detection of transplant arterial stenosis can help prevent transplant failure and the need to re-transplant. Doppler ultrasound with manual measurements is the most common screening method, but it suffers from poor specificity when thresholded to reduce false negatives. Positive screening cases proceed to angiography, which is an invasive and expensive procedure. A more accurate test could decrease the number of normal patients who would have to undergo this invasive diagnostic procedure. Machine learning models have shown promise in determining stenosis in the carotid artery; however, they have yet to be tested on the less ideal data hepatic arteries generate. Software has been created to extract liver artery Doppler ultrasound information in an automated fashion to predict stenosis. A turnkey approach is utilized to refine the region prior to extraction. Current methods of extraction generate waveforms with an average percent error per pixel of 6.5 percent from a human drawn waveform. Single feature models and machine learning models performed similarly when predicting stenosis; however, when thresholded for high sensitivity (greater than 0.90), random forest models had the highest specificity at 1.0 sensitivity and 0.60 specificity.
Table of Contents
Introduction -- Literature review -- Methodology -- Results -- Future work