by Arto Järvinen, director of research & development, ContextVision

Today's families interact with photo and video in a way that would (and does) boggle our grandparents.

Many of us can capture and share high-resolution video with our mobile phones, recording almost every important moment. With powerful face recognition software we can choose to automatically tag faces to organize our image files.

Fifteen years ago only professionals with an array of special-purpose workstations with specialized software could undertake any complex imaging task, and today we can do many of them with a device that fits in our pocket. The same intricate technologies that have changed how we interact with family photos and video are now also enabling new medical imaging possibilities.

From hardware to software

In the past, whether the modality was X-ray, ultrasound or MR, medical imaging was primarily fueled by equipment advances. Consider the development of X-ray technology in the early twentieth century. From the invention of the first vacuum tube in 1904, the thermionic diode, to the 1913 invention of the Coolidge X-ray tube, the pioneers of X-ray tubes learned from each iteration, experimenting with tube size and temperature until the optimal combination was found. A form of the Coolidge tube remains in use today.

This century, the combination of powerful but inexpensive standard hardware combined with special purpose software will be the dominant development force. This shift can be attributed to four primary drivers.

1. Exponential hardware growth: Thanks to pressure from the consumer electronics industry, such as the gaming industry and mobile communications industry, processing speeds have grown exponentially. The graphics processing unit, or GPU, is used more and more not only for gaming but also for general purpose computing, either replacing the CPU or working with it. As GPU technology rapidly develops and standard programming languages emerge, it increases the potential to fully leverage software advances. Huge volumes of images can be processed at unprecedented speed. With the right combination of hardware and software, real-time processing - which is mandatory for certain applications - becomes possible.

2. Using software to replace expensive hardware and reduce costs: When combined with the appropriate hardware, today's software can cost-effectively deliver advanced capabilities, which previously required expensive special purpose hardware. For example, ultrasound scan conversion was originally a process completed by hardware. This can now be accomplished with advanced image processing software. Additionally, software can be used to manipulate images, without the patient present, revealing additional details and pathologies while maximizing the throughput levels of costly equipment.

3. Rise of ultrasound: Given its low cost, mobility and non-ionizing nature, practitioners today use ultrasound for many clinical examinations that 10 years ago would have involved a more cumbersome and possibly dose-intensive procedure. While equipment evolution prompted much of this increased use, advances in image-enhancement software have made increased diagnostic capabilities possible. First developed to reduce noise in MRI imaging, advanced image processing software has revolutionized ultrasound imaging, with noise reduction and sharpening of organ boundaries, propelling ultrasound to major diagnostic tool status. And stakeholders will continue to push the modality forward.

4. Individual preferences require customization: Clinicians in different hospitals and regions typically have specific preferences in terms of image capture and analysis. Certain individuals or groups prefer soft edges while other users find the elimination of speckle to be the most important image-enhancement task. As successful equipment manufacturers serve a huge number of centers and hospitals, it is critical that they can tailor their equipment for success worldwide. Manufacturers can use software to customize their equipment based on end-user preferences, increasing sales and maximizing the ROI of each product. Ultimately, this customization will deliver improved diagnostic value.

Moore's law

Ten years ago, standard imaging equipment could not support the most advanced developments in software. New imaging technology that relied on complex algorithms to enhance images had to be developed in tandem with the hardware to support it. Because of the significant advances made in standard hardware since 2002, many types of imaging equipment can be upgraded by only replacing the software. Given this change, many medical device manufacturers can access the latest imaging advances without significant capital investment, or bothersome installation requirements of hardware upgrades and work with an independent developer to modify existing tools.

Moore's law, named after Gorden E. Moore, an Intel co-founder, states the processing power of an integrated circuit doubles every 18 months or so. According to the current Moore's Law Wikipedia entry, "Moore's law describes a driving force of technological and social change in the late 20th and early 21st centuries." Consider the proliferation of multi-core GPUs and other powerful hardware. This specific advance has enabled a range of image-processing operations that were impractical in the past, literally adding a new dimension to medical image processing. The real-time adaptive filtering of 3-D ultrasound image volumes possible today gives a superior image quality compared to what we once considered state-of-the-art 2-D filtering. With advanced 3-D rendering software, clinicians can produce the "babyface" images that parents love, adding a whole new chapter to medical imaging history and family photo albums.

Järvinen joined ContextVision in 2011, for the second time. He first joined the company in 1984, having earned degrees in image processing at the University of Linköping and Stanford University. He has held numerous management and consulting positions, including management consultant at a global management consultancy, chief quality officer for a global PACS company and managing director for a medical device company. His expertise includes system development methods, operational excellence, medical device regulations and management of high-tech companies in general.