27 Nov 2018 15:35
27 November 2018
MedaPhor Group plc
("MedaPhor" or the "Company")
MedaPhor's AI obstetric image database exceeds 1 million images
MedaPhor, the intelligent ultrasound software and simulation company, is pleased to announce that its database used for training its artificial intelligence (AI) products now exceeds 1 million obstetric ultrasound images.
Large image libraries are a prerequisite to creating AI solutions. These images have helped MedaPhor to develop its ScanNav® AI-based clinical software for ultrasound professionals. The one million mark is a significant milestone for the Company, as it will enable it to build on its ScanNav software for the global obstetric ultrasound market.
MedaPhor's ScanNav intelligent ultrasound technology, which uses AI algorithms and deep learning techniques to automatically assess ultrasound images, is currently being piloted in UK hospitals to support sonographers carrying out the 20 week anatomy scan. ScanNav assists sonographers to ensure that the images they are taking conform to the UK Fetal Anomaly Screening Programme (FASP) protocol. In future, the software will also be capable of automatically recording required images during the ultrasound scan.
To develop ScanNav, MedaPhor has been working with anonymised ultrasound scans taken throughout pregnancy in eight countries. The latest addition to the ScanNav image library comes from a collaboration with University College London Hospitals NHS Foundation Trust (UCLH) in the UK, which is using the ScanNav software to audit its routine obstetric ultrasound practice.
By continuing to increase the image library, MedaPhor is able both to improve the discrimination of existing software; develop ScanNav to a range of other global obstetric protocols at different stages of fetal development; and push forward automatic recognition and recording of images during routine scans.
Commenting, Nicholas Sleep, CTO at MedaPhor said: "Our ScanNav artificial intelligence software is the first AI based software that can assess in real-time the quality of scans taken during a 20-week pregnancy scan. Importantly, these automated results are as good as an experienced human sonographer. This significant increase in our image library will allow us to train ScanNav to meet all the 20-week global scanning protocols, as well as expanding the range of obstetric applications for ScanNav."
Enquiries:
MedaPhor Group plc | www.medaphor.com | |
Stuart Gall, CEO Wilson Jennings, CFO | Tel: +44 (0)29 2075 6534 | |
|
| |
|
| |
Cenkos Securities - Nominated Advisor and broker | Tel: +44 (0)20 7397 8900 | |
Mark Connelly / Cameron MacRitchie (Corporate Finance) |
| |
Michael Johnson / Julian Morse (Sales) |
| |
|
| |
|
| |
Walbrook PR | Tel: +44 (0)20 7933 8780 or medaphor@walbrookpr.com | |
Anna Dunphy / Paul McManus | Mob: +44 (0)7876 741 001 / Mob: +44 (0)7980 541 893 | |
About MedaPhor (www.investors.medaphor.com)
MedaPhor (AIM: MED), the intelligent ultrasound software and simulation company, develops artificial intelligence-based clinical image analysis software tools, augmented reality-based needle guidance software and advanced hi-fidelity haptic and manikin-based training simulators for medical practitioners.
Based in Cardiff and Oxford in the UK, Atlanta in the US and Hong Kong in Asia, MedaPhor operates two divisions:
Intelligent Ultrasound Simulation Division
Focuses on hi-fidelity ultrasound education and training through simulation. Its three main products are the ScanTrainer OBGYN and General Medical simulator training platform, the HeartWorks echocardiography simulator platform and the BodyWorks Eve Point of Care and Emergency Medicine Simulator. Over 500 MedaPhor simulators have been sold to over 300 medical institutions in over 30 countries around the world.
Intelligent Ultrasound Clinical Division
Focuses on augmented reality and deep-learning based algorithms to make ultrasound machines smarter and more accessible. Products in development include ScanNav and NeedleGuide. ScanNav uses machine-learning based algorithms to automatically identify, grade and capture good ultrasound images. NeedleGuide aims to simplify ultrasound-guided needling by using deep learning and augmented reality to provide the user with pathway guidance and automated tracking for a range of medical procedures.
Some products in the pipeline may require US FDA approval, as such this material should be considered informational only and does not constitute an offer to sell or infer claims or benefits.