Breast Tomosynthesis

Low Dose 3D Digital Mammography

Breast cancer is the most common cancer for women in this country. Each year, there are over 41,000 new cases in the UK. Each year there are over 12,500 deaths from this disease. This cancer accounts for almost one in three of all cancer cases in women, and the lifetime risk for breast cancer in women is one in nine.

Early detection of breast cancer using mammography significantly reduces mortality. However, conventional mammography still has a significant rate of both false negatives and false positives, particularly in younger women. Recently developed methods for 3D breast imaging, known as Digital Breast Tomosynthesis (DBT), promise to significantly reduce both false negatives and false positives, allowing the earlier detection of cancer while reducing the occurrence of callbacks. DBT is also more effective for imaging the dense breast tissue of women in their 30s and 40s. Dexela has developed a leading intellectual property position covering powerful methods to acquire, process and visualize data from 3D breast imaging. Dexela plans to play a central role in this field through collaborative agreements with medical device manufacturers.

The purpose of this research project is the proof of concept of the core image processing engine (reconstructor). The project’s principle deliverable is the creation of a bench top working model to implement advanced iterative, non-linear algorithms to generate 3D reconstructions of breasts from a plurality of 2D x-ray images. The project will develop algorithmic constraints and heuristics to produce high quality 3D reconstructions with minimal computational artefacts and blurring and maximum resolution and diagnostic efficacy.

The project’s deliverables are:
1) an improved algorithm with effective constraints and heuristics
2) a documented and optimised software implementation of the algorithm to meet clinical requirements for processing time
3) an experimental working model of the reconstructor that can rapidly generate a reconstruction from x-ray data and deliver it to Dexela’s proprietary workstation
4) One or more patent applications

The reconstructor would be used by our partners in clinical research projects along with our workstation. The technology would be implemented using an industry standard computing platform (64-bit Windows Server 2003 running on multi-core 64 bit x86 processors). Image data would be captured by clinical partners using DBT devices built by industry partners. Initially, these will be modified versions of existing machines. Substantial clinical data has been made available to us.

The global market for Mammography devices is large: there are approximately 40,000 installed globally and 2D digital devices cost around £200,000. The market for 2D Digital devices is growing rapidly and is currently worth over £200 million per annum. When DBT is perfected it is likely to have a similar price to 2D digital mammography (perhaps 15% higher) but yield significant advantages. The benefits of DBT are both social and economic: it will reduce false negative tests, saving lives, and also reduce false positive tests, reducing anxiety for patients and saving money. For these reasons DBT will rapidly displace 2D digital mammography and eventually lead to the replacement of the installed base of film-screen mammography devices.

Dexela’s marketing plan anticipates sales of reconstructor units and workstations to existing Medical Device Manufacturers as well as licensing IPR and software. Initial feedback from the manufacturers in this sector has been positive and they are receptive to both the technology and the business model. Revenue from these activities could reach £20m per annum if the technology is widely adopted for screening and Dexela is able to secure licensing and other sales with manufacturers.