DM: What about 4D motion pictures in ultrasound?

JEH: We can achieve a short-duration 4D solution today: on 128 cubic pixels we can run up to 24 volumes per second, which gives an indication [of how far along we are in achieving real-time 4D]. We intend to take it to the full real-time so it becomes a full-fledged 4D solution.


DM: I guess, as in the early days, you are limited only by the capacity of the processors.

JEH: That is correct. Right now we use GPU, a graphical processor unit, which is very popular in the gaming industry. And we see the trends moving toward that type of solution to really get the best possible performance of the processor. The GPU industry is moving faster than the CPU. However, we believe the CPU will win back its leadership in terms of performance growth in the future.

DM: Is ultrasound a modality that is very promising in terms of likelihood to benefit from software innovation in the future?

JEH: What makes ultrasound interesting is that it is the easiest and in many respects the cheapest way of applying some sort of imaging diagnostic. In that sense, ultrasound has grown to be used in many other types of procedures than five years ago. The whole hand-carried ultrasound market has brought the ultrasound devices to other parts of the hospital. So there has been a dynamic growth in usage of ultrasound. When it comes to the software, I would say the software is taking on an increasingly important role in all the modalities. You might say software takes on a slightly larger importance in ultrasound than other heavier modalities -- than MRI or CT -- because those devices have a lot of hardware compared to ultrasound.

DM: What about MRI?

JEH: ContextVision started out of MRI, where we introduced image enhancement into low tesla MRI units. We could see very good impact on the image by using image processing in such cases. By the mid-1990s we dominated the MRI market in providing image enhancement: Out of 10 MRI units sold worldwide, seven used ContextVision image enhancement.

The next big challenge for us after having introduced the 3D volumetric ultrasound, we will do the same for MRI and provide the image enhancement for volumetric data in MRI. And that is very exciting and also a very challenging step to take but we have high expectations in this area and we believe we can contribute a lot of improvement.

DM: Will that effort require a higher field MRI or would a 1.5T benefit from some of these software approaches?

JEH: We are looking at introducing our technology on 3T machines later this year for one of our customers. I think volumetric enhancement will be a very important and powerful solution that we contribute to taking the image to a new level.

More Voices