Research on detection method of bleeding point in two-dimensional DSA image based on parametric color imaging

Abstract

Digital Subtraction Angiography (DSA) has become an essential method in diagnosing and detecting vascular bleeding points. According to the frame number sequence, doctors subjectively analyze gray-black DSA data to diagnose vascular bleeding points. This study carried out parametric color imaging processing on the original two-dimensional DSA image sequence to assist doctors in judging blood vessel bleeding more accurately and quickly on angiography results and obtain quantitative standards. Five parameter values are calculated by enhancing the two-dimensional DSA image, segmenting the blood vessel area of the two-dimensional DSA image, and fitting the time-concentration curve. The parameters include AUC, time to peak, development time, transit time, and flow rate of contrast agent. The parameter value determines the gray value of the image pixel point and generates a color image with corresponding parameters through image pseudo-color processing. The color imaging results of each parameter of 33 groups of simulated hemorrhage data in-vitro and DSA data of 42 cases of vascular bleeding points determine the selection of parameters to assist in diagnosing vascular bleeding points, and the experimental results finally select the development time parameter color imaging result map to assist doctors in diagnosing blood vessel bleeding points.

Keywords: Digital subtraction angiography; Image enhancement; Image segmentation; Parametric imaging; Time-concentration curve.