Mobile application technology choice
We consider the pros and the cons of the following mainstream mobile application environment:
native Android mobile application
- (+) almost fully OpenSource framework
- (+) large and reactive community
native iOS mobile application
- (-) proprietary environment
- (+) large community
cross-platform framework for mobile application
- (-) no acces to low level interface, framework is not optimized for required optimizations on backend (real time image rendering) and front-end (precise gestures)
According to the points listed above we choose to implement first an Android native mobile application.
Rendering pipeline technology choice
We have developed multiple ScanConversion filters (transformation from polar to cartesian coordinate in an image) relying on differents technologies. This provided us good feedbacks on which technology we should use in order to get great performances to generate images from probe data.
RenderScript
The best performances as of (07.05.17) are obtained with RenderScript. We achieved a performance of around 90-100 FPS for a 512x512 image. However, the limit is that the user does not control the allocation to GPU or CPU, and it seems that on most modern architectures, GPU is consistently used, which explains the high frame rate. In a near future, we would like to carry the scan conversion on Vulcan. For the moment (07.05.17), it is not widely supported or implemented on smartphones.
OpenCV
To process scan conversion in our implementation, we use OpenCV. The OpenCV implementation scan conversion process is processed from the opencv_interpolation() in the ScanConversion class, under the preproc package.
OpenCV uses essentially the number of lines and the number of samples in order to run its scan conversion method Imgproc.linearPolar (see here for further informations about number of lines and of samples).This method can be found in the opencv_interpolation() in the ScanConversion class.
We chose to use cubic interpolation, which is done by passing to method the Imgproc.INTER_CUBIC constant. There are many other scan conversion algorithms choice available. You'll find here some of them such as Imgproc.INTER_NEAREST for nearest neighbour interpolation or Imgproc.INTER_LINEAR for bilinear interpolation
To understand further the technological internals of the ultrasound probe, we refer the reader to our start-kit GitBook
We reach a performance of 30 FPS on a 512x512 image.
Home-made scan conversion
The home made implementation scan conversion process is processed from the compute_interpolation() in the ScanConversion class, under the preproc package.
Contrary to the OpenCV implementation, there are much more paramaters to set. See here for further informations about these parameters.
We reach a performance of 10-12 FPS on a 512x512 image.
OLPC implementation
In 2007, some echOpen team members worked on an early implementation of an ultrasound probe on an OLPC device. It faced scan conversion issues, you'll find her.e a short story of this and some technical explanation about how the problem was tackled.