HybridVel

Improved blood velocity measurements with a hybrid image filtering and iterative Radon transform

Description

Red blood cells (RBCs) in plasma labeled with a fluorescent dye present as black disks on microscopy. Moving RBCs create black streaks on the space-time image, e.g., with line-scan imaging on two-photon microscopy. RBC velocity can be calculated from the angle of the streaks. Radon transform can measure the streak angle, but has compuation speed-accuracy trade-off. Moreover, time-invariant and slowly time-varying artifacts prevent accurate angle measurements. The Matlab m-file provided here removes such artifacts from the space-time image while enhancing the streak edge by filtering the image with vertical 3x3 Sobel operator. Iterative application of Radon transforms offers very precise angle measurements with the least number of Radon transforms and thus addresses the computation speed-accuracy trade-off. The m-file is based on the paper published in Frontiers in Brain Imaging Methods titled "Improved blood velocity measurements with a hybrid image filtering and iterative Radon transform" by Pratik Yashvant Chhatbar and Prakash Kara (http://www.frontiersin.org/brain_imaging_methods/10.3389/fnins.2013.00106/abstract).

Acknowledgements

Sortable Profiler Report and Unimodal Regression inspired this file.

• Required Products: Image Processing Toolbox; MATLAB
• MATLAB release: MATLAB 7.5 (R2007b)