Yunsub Jung MS, Presenter: Nothing to Disclose

Hwan Shim PhD, Abstract Co-Author: Employee, Samsung Electronics Co Ltd

Beunggeun Cheon MS, Abstract Co-Author: Employee, Samsung Electronics Co Ltd

Hyungjoon Lim MS, Abstract Co-Author: Employee, Samsung Electronics Co Ltd

Youngtae Kim PhD, Abstract Co-Author: Employee, Samsung Electronics Co Ltd

Measuring the blood velocity with Pulsed wave (PW) Doppler is affected by parameters such as Doppler angle, positioning of the sample volume (SV), and Doppler gain and these parameters are generally decided by human. To overcome error by human, operation complexity and time consumption in measurement, we present an automation process for PW Doppler.

Proposed automation method was designed based on vessel and bloodstream information and include the following steps; 1) vessel detection and aliasing area removal 2) region of interest (ROI) searching and SV positioning 3) B, C-mode image fusion 4) vessel angle estimation using virtual ray 5) computing of Doppler and steering angle 6) measurement of blood velocity. As a result of mentioned process, SV position and vessel angle are automatically computed without user intervention. Then, the vessel angle obtained from automation method were compared with manual results by two experts, also repetition test was performed to evaluate a reproducibility of each way. We collected 300 pair images (B-mode : 300, C-mode : 300) including the carotid artery from three subjects without vascular disease for clinical assessment.

Vessel angle in collected images is distributed approximately from -40 to 40 degree. To evaluate the vessel angle accuracy of automation method, we calculated the mean absolute difference (D) between computed angle and manual measurement angle, at this time, each angle was estimated under same conditions such as ROI and SV position. Automation method compared to manual results by two experts shows D1 = 1.2 ± 0.7 and D2 = 1.7 ± 0.9 respectively. In the case of repetition test, automation method always present D = 0 ± 0, however, manual measurement (D1 = 1.7 ± 0.4, D2 = 2.3 ± 0.7) shows a variation definitely.

Automation method for PW Doppler allows the robustness measurement of blood velocity without variance caused by subjective human factors, moreover, enhance the accuracy and reproducibility of the results.

The described technique might be useful for current clinical environment in aspect of time reduction and reproducibility.

Jung, Y, Shim, H, Cheon, B, Lim, H, Kim, Y, Automation Process for Pulsed Wave Doppler Measurement Using Bloodstream Information.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13011913.html