New AI-Powered Software Can Analyze and Report Vascular Ultrasound Images
Sat, April 17, 2021

New AI-Powered Software Can Analyze and Report Vascular Ultrasound Images

See-Mode Technologies developed “Augmented Vascular Analysis (AVA)” that will help medical professionals interpret images in vascular ultrasound scans / Photo Credit: Zapp2Photo via Shutterstock

 

Artificial intelligence in healthcare is a reality for the past few years already. It brings more tools and systems that address certain issues or problems faced by the industry.

More and more researchers are developing AI tools to further expand the current usage of the technology in healthcare. Recently, Singapore announced that its Health Sciences Authority has finally approved the use of AI-powered software for the automated analysis and reporting of vascular ultrasound scans.

Scans are typically analyzed by a sonographer or radiologist after manually reviewing between 50 and 150 images for each patient. The results are handwritten documents that are filled with drawings, numbers, and measurements. This can take as long as 20 minutes per patient for severe cases. Not only can this be time-consuming, but it is also error-prone.

According to ZDNet, a business technology news website that covers breaking news, analysis, and research and keeps business technology professionals in touch with the latest IT trends, issues, and events, the new AI-powered software uses deep learning, text recognition, and signal processing technologies to help medical professionals interpret such images. It was developed by See-Mode Technologies, which raised $1 million in its seed funding round earlier this year from investors.

The software, which is called “Augmented Vascular Analysis (AVA),” was classified as a Class B medical device by the Health Sciences Authority. This means that it was deemed to have low-to-mid risk. Sadaf Monajemi, See-Mode's co-founder, stated that their research team found that mistakes in handwritten ultrasound worksheets could potentially result in the wrong treatment plan for a patient. "Given that these reports are produced by clinicians, who acquire and review hundreds of images per day, human error is inevitable," Monajemi said. 

At the same time, the AI-powered software not only improves and minimizes potential errors but also generates an analysis report in under a minute. As of now, See-Mode is looking to secure regulatory approval for the AI-powered software in other regions, including TGA in Australia and FDA in the US.