The use of comprehensive software components for research hardware is necessary to help developers to focus on the development of new techniques and not the handling of common problems in ultrasound imaging.
With this in mind we designed a portable C++ and C# (Microsoft .NET framework) software architecture including imaging and analysis for both online and offline tools. Nevertheless the programming of the framework and its interfaces is easy to learn and can be adapted to any additional tasks.
Included in the software framework are routines for data acquisition via USB, signal processing (i.e. logarithmic compressions, envelope detections, bone or tissue detection filters, channel-data to delay-and-sum-data reconstruction, scanconversion, 3-D reconstruction,...), image analysis and processing (i.e. speckle reduction imaging, measurements,..), data export (i.e. RF-raw data, images, videos, DICOM, text output for external analysis,...) and data import from different measurement systems in-house and externally.
The algorithms support multi-core, multi-threaded or GPU accelerated (using OpenCL) system architectures and benefit from parallelization.
Support for 3D tracking systems to acquire and reconstruct volume datasets is implemented by position and orientation measurement for optical, mechanical and electromagnetic 3D tracking systems (i.e. NDI Polaris, Spectra, Vicra, Microscribe, Ascension Flock of Birds,...).
The signal and image processing of this research platform provides a unique feature by the closed-loop device control that simplifies the development of new techniques and algorithms. Custom filters developed according to the filter software interfaces can control the beamforming and system parameters automatically, conserving the medical certification for the important tests at clinical sites.
All this is implemented using an open filter framework, open data structures, a plug-in concept and closed-loop control. Programming an implementation of a new filter algorithm for both online and offline processing can be developed in just 5 minutes using templates for RF-based or image-based processing.
The picture above shows the possibilities of signal generation and data acquisition using the "DiPhAS" research platform, and gives an overview of access to channel data, beamformed 16 bit RF-data, 8 bit amplitude data after envelope detection and logarithmic compression before scan conversion, image data after scan conversion and processed imaged data for use in the closed-loop control.