Diagnostic ultrasound: the geophysical way
Kelvin3D is a revolutionary new concept in ultrasound imaging, intended primarily for the medical imaging and diagnostic market. It is protected by patents either granted (US patent) or pending.
It is revolutionary in its approach to three-dimensional ultrasound imaging, because it uses an approach which is diametrically opposed to current ultrasound techniques.
However, it is in fact a very well-tried and tested methodology; - that is, it uses the technique developed over many decades by the oil exploration industry. This is 3-D seismic reflection imaging, a sort of 'scanning' into the Earth using low-frequency sound, to search for oil and gas. In effect, the proof of concept for Kelvin3D is there already in the oil exploration industry.
I developed the US patent, granted in 2004, in collaboration with maxillofacial surgeon Mr Duncan Campbell. The original patent, which has recently expired, was very broad but contained few details.
We tested the concept by having an array built (550 channels, 2 mm pitch, 2.4 MHz). It uses the geophysical philosophy scaled down by 10,000 spatially (and upwards by the same factor in frequency). I processed the data, comprising various phantoms and bone samples, using standard seismic industry software. In short, our proof of concept results show that the geophysical way of doing ultrasound does work, and is potentially far superior.
Here is a summary of Kelvin3D results to date:
- Hardware comprises a distributed areal array, with omnidirectional point sources and receivers.
- Imaging into and through bone is no problem (in the earth sciences we deal routinely with rocks varying in velocity by a factor of three or four).
- Furthermore, we can extract bone elastic properties (including anisotropy) by reflection off the top surface, all as a by-product of the acquired imaging data.
- The image quality (precision) does not degrade with depth; there is no 'windscreen wiper' effect, as with conventional 2D/3D ultrasound images.
- Another novel application is in indirect, non-invasive measurement of intracranial pressure, by imaging through the eye.
- Because the energy density of our surface is low, we can inject a lot of ultrasound energy without risk of tissue damage. Depth penetration (e.g. through obese patients) is not a problem.
- The imaging method uses the instantaneous attributes of the source wavelet and not just the energy envelope, so for a typical short pulse the imaging precision is several times better. Our relatively low frequency prototype frequency of 2.4 MHz is thus equivalent to about 10 MHz in the orthodox ultrasound field.
- Because the hardware is relatively simple, and most of the image construction is done in software, medical devices will be relatively inexpensive, and open to continuous development by software improvements.
- Because no ionising radiation is used, scans can be carried out as often as necessary.
Mr Campbell and I incorporated Kelvin Biophysics Limited (KBL) in 2005 to develop the device. As KBL we obtained funding to build and test a prototype Kelvin3D array. KBL was dissolved in 2010. It was replaced firstly by Kelvinbio Limited and subsequently by Kelvin3D Limited. During this period we had several unsuccessful attempts to collaborate with small niche companies, but funding had become scarce after the financial downturn of 2008-09.
Kelvin3D was dissolved in 2016 but revived in October 2019, because we are now in negotiations with at least one major medical ultrasound company. We filed a new US provisional patent application in November 2019. This covers the new IP that we have developed and tested since the original patent was granted.
The fourth dmension
Kelvin3D is not only intrinsically designed for 3D scanning, but also for four-dimensional imaging. The fourth dimension is time, in addition to the three spatial dimensions of 3D. What this means simply is a 3D image which is animated, or turned into a movie. 4D, which Kelvin3D is capable of, means that we can see an animated 3D image in real time, without any significant delay. This will be invaluable in medical intervention, as well as in diagnosis.