QuEST Global, a global product engineering and lifecycle services company, announced that it has developed a robust AI-powered solution that will enable healthcare professionals to accelerate COVID-19 screening of patients with pneumonia symptoms.
Using advanced deep learning models, the AI-powered diagnostic solution can sort and identify chest X-rays of patients with COVID-19. With an accuracy of more than 95%, the solution can be deployed on the cloud as a service, making it easily accessible on edge for healthcare professionals and end-users. The solution is backed by Microsoft Azure Machine Learning, which helps accelerate the development and deployment on machine learning models in a highly secure and trusted fashion.
The Medical Devices engineering team at QuEST developed this technology demonstrator using chest X-rays of healthy individuals, patients with symptoms of pneumonia and COVID-19. These X-rays were used to train and build a deep neural network model that could discriminate the radiological patterns of pneumonia related to COVID-19 and highlight the suspicious ones, thus leading to a faster screening of the disease.
Krish Kupathil, Global Head, Hi-Tech and Digital, QuEST Global said, “Since the fight against COVID-19 is all about faster screening and immediate isolation of maximum number of people, we aim to accelerate the screening time as much as possible. The AI-based solution will make radiography examinations much faster by leveraging modern image diagnostic systems. As we continue to add more features, we aim to reduce the screening time to less than a minute”.
Michael Kuptz, General Manager – Americas IoT & Mixed Reality Sales, Microsoft, said, “Microsoft’s collaborations with product engineering leaders like QuEST can go a long way to driving a more positive outcome. For example, QuEST’s AI-driven diagnostic solution, built on Microsoft Azure, empowers healthcare personnel in the fight against COVID-19 by reducing screening time, thereby enabling more testing capacity.”
Disclaimer: This media release is auto-generated. The CSR Journal is not responsible for the content.
