COVID-19 identification in X-ray images by Artificial intelligence
[ad_1]
Novel Software to identify COVID-19 from X-Ray images
A novel approach to identify COVID-19 with high confidence from other viral and bacterial infections using X-ray will help to make Coronavirus diagnosis faster and more accurate with the help artificial intelligence-based program.
Flint, North Wales May 2020,
Flint based team of mathematicians and computer scientists has successfully developed an artificial intelligence-based program to identify COVID-19 pneumonia from other virus and bacteria caused pneumonia. Impressively, the software has managed to make the identification with 95% accuracy. The software takes less than 30 seconds to analyse a Chest X-Ray image to reach its conclusion. It is expected that this technology will reduce the workload on Radiologists and speed up the triage of COVID-19 patients in combination with other clinical data.
Professor Sabah Jassim, who oversaw the development commented “The unique nature of the development cannot be underestimated and its contribution to the medical teams is clear. What is needed now is to work with many hospitals to evaluate the technology and validate its output”.
Dr Shakir Al-Zaidi, managing director of Medical Analytica Ltd added “It is exciting to offer such unique approach which can support the speedy triage of suspected COVID-19 patients and also to differentiate, with high confidence, COVID-19 infection from pneumonia caused by other viruses and bacteria”.
The company is seeking collaboration with radiologists to work together to test and validate the software performance on X-ray images and furthermore to test the technology to identify other critical medical conditions.
The development team is part of Medical Analytica Ltd, a new start-up company based in Castle Park, Flint. It focuses on the application of artificial intelligence in medical imaging. Software for the identification of COVID-19 using CT scans has already been developed. Another software for the identification of Malignant ovarian tumour from Ultrasonic images has already been developed and hospital-based evaluation studies are on course to start in June 2020.