Automatic video tracking

Screen shot of the automatic analysis of marker motion at 85 ⁰C target temperature.

Automatic video tracking is a  technique of automatically locating a  moving object over time within recorded video frames. The image in this post is from a project, where I programmed a tracking algorithm in MATLAB to analyze the shrinkage dynamics of liver tissue during ablative temperature exposures. The algorithm located the tracking markers automatically based on a threshold criteria and identified them on every thirtieth frame (= 1 second). The marker positions (centers) as well as the distance between the two markers were stored and the relative shrinkage in percent was estimated. A video that shows the automatic tracking procedure can be found here: http://www.youtube.com/watch?v=X_SnEULgH7M
The essential tracking function for Maltab can be found here: https://www.dropbox.com/sh/ip306uavm7jpekg/ib65VT633K/Automatic%20video%20tracking

To create the application I used:

• Matlab (http://www.mathworks.com/products/matlab)

 

Patient-specific finite element simulation of radiofrequency ablation

Tetrahedral FE model for patient-specific simulation of radiofrequency ablation in the liver. The model geometry was generated based on patent-specific CT-image data. The image shows liver vasculature tumor, ablation probe, and surrounding bone structure.

Lets get this blog started with one of my favorite images of an python application (http://www.youtube.com/watch?v=jTXbjQ_6RWM) that I created to generate finite element (FE) models for simulation of radiofrequency ablation (RFA). The FE method is an established technique to solve the thermal-electrical differential equations for realistic RFA treatment simulation. Nevertheless, generation of patient-specific FE models is a labor intense task and requires use of multiple software tools. In order to overcome the challenges of modeling I generated a research tool as well as a workflow to generate patient-specific FE models for simulation of RF ablation efficiently. Common techniques of image processing and image meshing are combined with new approaches, such as interactive electrode placement, interfaces to commercial FE software programs including visualization of model results overlaid on imaging data. Such a tool may provide additional information for the treating physician for more effective pre-treatment planning of procedures. Via this tool, I was able to simulate RF ablation based on patient-specific geometry and visualize model results overlaid on CT imaging data. 

To create the application I used:

• Python (http://www.python.org)
• Numpy (http://numpy.scipy.org) 
• Tkinter (http://wiki.python.org/moin/TkInter)
• PyOpenGL; OpenGL,GLSL (http://pyopengl.sourceforge.net)
•  f2py (http://cens.ioc.ee/projects/f2py2e)