The researchers plan to continue testing their approach with the ultimate goal of developing a free online platform for medical image classification.
A recent study published in EEE / CAA Journal of Automatica Sinica, a joint publication of the IEEE and the Chinese Automation Society, has found that X-rays in conjunction with an artificial intelligence device could be a key method for detecting the COVID-19 virus.
The research was conducted by researchers from the Universidade de Fortaleza in Brasil, in collaboration with the Universidade Federal do Ceará and the Instituto Federal do Ceará.
“When the COVID-19 pandemic arose, we agreed to use our knowledge to help tackle this new global problem,” said Victor Hugo C. Albuquerque, a researcher at the applied image processing, signaling, and computing laboratory and the Universidade de Fortaleza said: “We decided to investigate the possibility of automatically detecting COVID-19 infection using X-ray imaging.”
The researchers behind the study have previously researched the detection and classification of lung pathologies, such as fibrosis, emphysema and lung nodules, using medical imaging. Common symptoms of COVID-19 such as respiratory distress, cough and pneumonia, can be seen through medical imaging such as CT scans or X-rays.
For this study, the team focused on developing X-ray machines with a machine learning algorithm to see if it could detect COVID-19 in breast scans. Using a number of different methods, they trained the model with a large data set to detect lungs that are susceptible to the virus.
Findings showed that the AI device was able to detect COVID-19 in chest X-rays with an accuracy of 95.6 to 98.5%.
“Because X-rays are very fast and cheap, they can help test patients in areas where the healthcare system has collapsed or in locations far from major centers with access. more complex technologies, “Albuquerque said.” This approach to automated detection and classification of medical images can help doctors identify, quantify the severity of the disease.