Smart Modality Group

The Smart Modality team formed in 2022, bringing together technical, scientific, and clinical expertise in the field of medical imaging equipment and operation.

Key goals of the Smart Modality team are to improve clinical throughput and reduce the burden on clinical personnel. We aim to achieve this by simplifying, automating, and increasing the efficiency of scanner imaging workflows using Machine Learning, Image Analysis, Computer Vision, and System Integration.

The Smart Modality team researches and develops new technologies that power innovative features of new generations of imaging systems. Our algorithms are integrated directly into the target imaging systems and tested at clinical sites across the world. We collaborate closely with other Canon research sites and with clinical experts in the field to ensure that our technology addresses genuine clinical needs, and that our understanding of current trends in medical imaging are up to date.

Project Spotlight

Scan Planning. For every CT scan performed, the radiographer defines the boundary of the scanning field (the start, end and sides). These areas are patient-specific and are currently set manually based on two quick 2D scans (or ‘scanograms’) in a frontal and a side view. This ensures the anatomy of interest is covered by the scan and limits the inclusion of other anatomical features.

Our Scan Planning algorithm applies a mixture of traditional and machine learning-based algorithms to find the relevant anatomy in a single very low-dose 3D scanogram. This anatomy is then used to automatically set the patient’s optimal scanning field. This type of automation is more reproducible and has been shown to generate tighter bounds than when manually set. Thus, automated Scan Planning helps reduce the radiation exposure of the patient. It also reduces the radiographer’s time interacting with the machine, allowing more attention for the patient. With algorithms like CT Scan Planning, the Smart Modality team brings ‘one-click imaging’ a step closer to reality.

Representation of scan planning steps

User interface guiding through the scan steps

Scotland’s Life Sciences Annual Awards Finalist 2023

SmartFusion is a suite of applications which are integral to Canon Medical’s Ultrasound scanners. Our software engineers and scientists work closely with clinical collaborators and colleagues in Europe, Japan, India and China to develop new advanced features and to maintain our products currently in use by customers.

SmartFusion provides the alignment (registration) of real-time ultrasound images with previously acquired 3D images from another modality (e.g. CT, MRI, PET etc.). Advanced software initially registers the two modalities and then this registration is maintained using Canon’s 3D magnetic sensor technology.

Registration brings the advantages of different modalities together: For example, during an ultrasound examination, a tumour that is not visible to ultrasound can be accurately located and treated using guidance and information provided by the other modality.

Image below highlights the effectiveness of Smart Fusion. The target lesion is clearly visible on the MRI image (left), while not visualised on the UL image (right). By enabling the registration of the two modalities, Smart Fusion allows the practitioner to confidently biopsy the lesion by providing the green circle overlay which precisely indicates the ROI, alongside needle guidance indicator (dotted line) for accurate placement. 

What our people say

“I work daily to solve complex, clinical challenges through creative engineering. Employing a variety of cutting-edge techniques and innovative tools to advance and solve these problems effectively. I am humbled by the impact this technology has to the lives of people, and pride myself on being able to play any part in its creation. Perhaps the only thing more humbling is the quality of the people I am able to work with and be inspired by.”

Peter, Software Engineer

“Starting with what seems to be a clear-cut and easily described problem, we usually then see the clinical reality turn it into a really interesting technical challenge. To build something that deals with all the complexities of real-world usage, to make it robust so that it works as expected every single time, that’s exciting to see. To then see it go to market and be appreciated by its users is really the icing on the cake.”

Corné, Technology Team Lead