DIT: Fake tissue to aid training in ultrasound imaging
Innovation Profile Complex mimicking of a real breast is set to boost the efficiency of scanning and the skills of experts
Dr Jacinta Browne of DIT with co-researcher Prof Andrew Fagan of St James’s Hospital and Minister Sean Sherlock.
Ultrasound scanning is a very powerful diagnostic tool and plays a particularly important role in the detection of conditions such as breast cancer. However, as some recent cases in this country have demonstrated, the technology must be used properly if it is to be effective.
Researchers from the school of physics in DIT and the Centre for Advanced Medical Imaging in St James’s Hospital are addressing this issue through the development of a highly advanced tool for training radiologists and sonographers in the use of ultrasound scanners.
The problem to be addressed lies in the very nature of ultrasound scanning itself, according to Dr Jacinta Browne, who leads the Medical Ultrasound Physics and Technology Group at DIT. “Ultrasound scanning is very different from an X-ray, for example,” she explains. “With an X-ray, you put the patient into the required position and take an image at the required power setting. It’s quite straightforward and easy for the patient. The images can be viewed later and further images taken if needed.”
With ultrasound, there is no snapshot picture. Instead, the scanner operator uses a transducer pressed against the skin of the patient to send sound waves into the tissue and detect their reflections or echoes. These reflections are displayed on a screen in the grainy black -and-white images familiar to all who have experience of foetal ultrasound scans. But the image is built up as a combination of very small thin “slices” just a few millimetres wide and centimetres thick produced by the transducer.
“A radiologist has to be very skilled when using the equipment,” Browne notes. “They have to press the transducer against the skin of the patient with the right force and keep it at the right angle to get usable images and they have to do this while looking at the screen and adjusting various controls to adjust and enhance the image. It’s very difficult and requires a lot of skill.”
Deterioration of image
And the level of concentration and expertise required to pick up on lesions or other abnormalities which may be displayed on screen is also quite considerable. There is a 90 per cent loss of energy per centimetre of tissue which the sound waves pass through so the quality of the image deteriorates quite dramatically with the depth of the scan.
There are ways to boost the quality of the image by changing the frequency of the sound but higher frequencies do not penetrate as deeply as lower ones. There are also ways of amplifying the sounds that come back but this can result in other distortions.
The skill level required is evident but training is still very much on-the-job for radiologists.
“Currently, radiologists are trained in a live-patient environment, with all the associated problems this brings,” says Browne. “Prior to this, their training is typically limited to a simulation using stocks of preprepared images and lectures on the underlying physics. This offers no tactile feedback and does not help with co-ordination skills, nor does it afford trainees the opportunity to master the complex ultrasound scanning technology used for breast imaging.”
Browne and her team have addressed this problem through the development of what can be described as an anatomically realistic false breast (or an anatomically correct anthropomorphic breast phantom to give it its correct technical term).
This has been designed to exactly mimic the complex nature of the breast, complete with the different tissue types contained within it including glandular tissue, pectoral muscle, subcutaneous fat and Cooper’s ligaments, together with a variety of pathologies and other complex features randomly distributed throughout the breast such as malignant and benign lesions, fibroadenoma, cysts, and calcifications.
It works as a training device because, when scanned using a modern breast ultrasound scanner, the phantom’s images very closely resemble those from actual patients. “We achieved this through a process of tweaking the material properties of each constituent element in the phantoms so that they matched the various tissues and pathologies being mimicked, considering the physics of the interaction of the ultrasound energy with the tissue-mimicking materials.
Properties such as the speed of sound, the acoustic attenuation and the acoustic impedance were adjusted and optimised for each anatomical structure and pathology mimicked, while the phantoms themselves were designed and constructed to replicate the internal structure and shape of the breast as closely as possible.”
The device has been devised to be used in conjunction with a specially designed training protocol, whereby the ability of the trainees to detect and correctly characterise all suspicious targets in the phantom can be assessed before and after a training period of four weeks.
The new device is already close to becoming a commercial reality and a small-scale pilot study carried out with registrars in St James’s Hospital, the National Breast Screening Programme (BreastCheck) Dublin, and University College Hospital Cork was recently completed with promising results.
“We saw a greater than 4 per cent improvement in performance,” she says. “This would assist with both improving detection of abnormal lesions in the breast but also help reduce over-diagnosis which leads to unnecessary treatment and anxiety for patients. At present, it is estimated that for every patient whose life is saved by breast cancer screening, up to 10 others will undergo unnecessary treatment.
“We are continuing work on the device and we are investigating commercialisation possibilities with the help of Dermot Tierney here in the DIT hothouse, the technology transfer office,” she says. “We have received assistance from Enterprise Ireland to carry out a feasibility study to assess the commercial potential for the product and the feedback has been excellent.
“We are now investigating different materials for the surface of the phantom to make it robust and we are very hopeful of having a product commercially available on the market within the next year or two.”