Software Defined Ultrasound

The term Software Defined Radio (SDR) was introduced by Joe Mitola in 1991 [1] to describe a new class of radios that are both programmable and reconfigurable. Unlike traditional radios, whose functionality is largely fixed by hardware components, an SDR uses software to implement many radio functions. As a result, the same physical hardware platform can support multiple communication standards, modulation schemes, and operating modes simply by changing the software, even during operation.

In essence, the fundamental objective of SDR technology is to move signal processing and control as close as possible to the antenna or transducer. By shifting traditionally hardware-based functions—such as filtering, modulation, and demodulation—into software, SDR systems achieve higher flexibility, scalability, and futureproofing. This concept applies not only to electromagnetic wave-based communication systems but also to non-electromagnetic systems, such as ultrasonic acoustic wave applications, where similar principles of software-driven signal processing are employed.

Building on the principles of Software Defined Radio, the software‑centric paradigm can be extended to ultrasonic systems, giving rise to Software Defined Ultrasound (SDU). In conventional ultrasound systems, key functions—such as waveform generation, beamforming, filtering, and signal interpretation—are implemented using fixed or semi‑programmable hardware, which limits adaptability, increases upgrade costs, and constrains experimentation.

SDU overcomes these limitations by shifting signal generation, processing, and control into software. In this architecture, the ultrasound transducer serves as a generalized acoustic front end, while transmit waveforms, receive processing, beamforming, and imaging algorithms are implemented in software on programmable platforms such as digital signal processors (DSPs), field‑programmable gate arrays (FPGAs), or general‑purpose processors.

By moving ultrasound system intelligence into software, SDU enables a single hardware platform to support multiple operating modes, frequencies, and applications. For example, the same system can be dynamically reconfigured for medical imaging, non-destructive testing, flow measurement, or therapeutic ultrasound simply by updating the software. This mirrors the flexibility achieved in SDR systems, where different communication standards and protocols are supported without altering the underlying hardware.

Furthermore, Software Defined Ultrasound facilitates rapid prototyping and innovation. Advanced techniques such as adaptive beamforming, coded excitation, real-time parameter optimization, and machine-learning–assisted signal interpretation can be deployed, tested, and refined through software updates alone. This significantly shortens development cycles and enables researchers and engineers to explore novel ultrasonic methodologies that would be impractical in fixed hardware designs.

In summary, Software Defined Ultrasound represents a natural extension of software-defined system concepts into the acoustic domain. By treating ultrasound signal generation and processing as software-defined functions, SDU offers enhanced flexibility, scalability, and futureproofing, similar to the benefits realized by Software Defined Radio. This convergence of software-driven architectures across electromagnetic and acoustic systems highlights a broader trend toward universal, reconfigurable sensing and communication platforms.

As these software‑defined concepts have moved from theoretical framing toward practical implementation, several recurring themes have emerged in both academic literature and practitioner discussions.

SUPRA – The Open‑Source Software‑Defined Ultrasound Pipeline

One of the most widely referenced realizations of software‑defined ultrasound is SUPRA, an open‑source framework created by the Chair for Computer Aided Medical Procedures at the Technical University of Munich, providing a 2D and 3D pipeline from beamforming to B‑mode imaging. [2], [3]

A defining characteristic of SUPRA is that key real‑time processing stages are implemented in software and accelerated using NVIDIA CUDA on GPUs to achieve high throughput, demonstrating that ultrasound pipelines can be executed on consumer GPUs in real time. [2], [3]

SUPRA also emphasizes access to early‑stage data—such as raw channel data and RF data—and explicitly targets improved reproducibility, since pipelines can be shared and executed without relying on proprietary, closed hardware implementations. [2]

Decoupling Hardware and Imaging Algorithms

A recurring SDU theme is decoupling ultrasound acquisition hardware from imaging algorithms: the transducer/front‑end performs transmit/receive, while many downstream steps (receive beamforming, envelope detection, log compression, scan conversion) become software modules that can be changed without redesigning hardware. [2]

This separation is repeatedly motivated by the observation that traditional systems often implement these steps in specialized hardware (e.g., FPGAs/DSPs), making algorithmic modifications difficult and time‑consuming—whereas modern compute (especially GPUs) reduces the need for fixed-function implementations. [2]

GPU‑Accelerated and Heterogeneous Computing

Another major thread is GPU acceleration as an enabling layer for real‑time SDU. SUPRA’s use of CUDA highlights the practical path: parallelize the computationally heavy stages on the GPU while keeping hardware‑bound steps (transmit/receive) on the acquisition device. [2]

A related discussion trend is portability across heterogeneous hardware (GPU/FPGA/CPU). For example, there is published work describing migration of a CUDA‑based ultrasound implementation (SUPRA) into a unified programming environment so it can target multiple architectures. [4]

Rapid Prototyping and Algorithmic Innovation

SDU platforms are repeatedly framed as rapid prototyping engines: once the processing chain is software-defined, researchers can implement and swap processing components (e.g., beamforming variants or pipeline stages) much faster than in a hardware-centric development cycle. [2]

SUPRA’s design goals explicitly include modularity and the ability to alter the pipeline by adding/removing/exchanging processing components—supporting experimentation and algorithm iteration without requiring specialized hardware development workflows. [2]

Toward Universal, Reconfigurable Ultrasound Platforms

The above threads collectively point toward a broader SDU vision: universal, reconfigurable ultrasound systems where hardware provides raw acquisition capability while imaging behavior is largely software-defined—analogous to how software-defined radio abstracted radio hardware into programmable signal-processing blocks. SUPRA explicitly frames itself as following the “software-defined” concept (derived from software-defined radio) and positions software modularity as central to enabling flexible ultrasound pipelines. [2]

References

[1] Joseph Mitola, III, Software Radio Architecture: Object Oriented Approaches to Wireless Systems Engineering. John Wiley and Sons, 2000

[2] Rüdiger Göbl, Nassir Navab, and Christoph Hennersperger, “SUPRA: open-source software-defined ultrasound processing for real-time applications,” arXiv preprint, 2017. [Online]. Available: https://arxiv.org/pdf/1711.06127v3.pdf [arxiv.org]

[3] IFL-CAMP, “SUPRA: Software Defined Ultrasound Processing for Real-Time Applications — An Open Source 2D and 3D Pipeline from Beamforming to B-Mode,” GitHub repository. [Online]. Available: https://github.com/IFL-CAMP/supra [arxiv.org]

[4] Y. Wang et al., “Developing medical ultrasound imaging application across GPU, FPGA, and CPU using oneAPI,” IWOCL’21, ACM Digital Library, 2021. [Online]. Available: https://dl.acm.org/doi/fullHtml/10.1145/3456669.3456680 [ieeexplore.ieee.org]