Summary: Researchers at the University of Technology Sydney (UTS) and Middle East University have developed an innovative AI system that uses deep learning algorithms to rapidly and accurately detect COVID-19 cases. By analyzing X-ray images, the custom convolutional neural network (Custom-CNN) achieves a 98% accuracy rate, outperforming other AI diagnostic models. This breakthrough technology could significantly reduce the delays, costs, and errors associated with traditional PCR testing methods.
Reliable and timely diagnosis of COVID-19, alongside other respiratory illnesses like the flu or pneumonia, is crucial for effective treatment. However, distinguishing between these conditions based on common symptoms, such as fever, cough, and breathing difficulties, can be challenging. PCR tests, the standard method for confirming COVID-19 infections, are expensive, slow, and occasionally yield false positive results.
The new AI system developed by UTS researchers offers a potential solution to these issues. Professor Amir H. Gandomi from UTS’ Data Science Institute explains that the AI model provides a high degree of accuracy in detecting COVID-19 cases from chest X-ray images. This system can assist radiologists in making accurate and efficient diagnoses, particularly in regions with a shortage of medical professionals or high COVID-19 prevalence.
One of the notable advantages of using X-ray images for COVID-19 diagnosis is their portability, allowing tests to be conducted in local areas with limited access to specialized equipment. Moreover, compared to CT scans, chest X-rays result in less exposure to ionizing radiation, minimizing potential risks for patients.
The Custom-CNN model significantly expedites the diagnosis process by eliminating the need for manual examination of X-rays. This AI technology offers a faster and more accurate alternative, providing patients with earlier confirmation of COVID-19 infections. Timely detection enables prompt treatment with antivirals, which are most effective within the first five days of symptom onset.
To demonstrate the system’s superior performance, researchers conducted a comprehensive comparative analysis, validating the Custom-CNN model against other AI diagnostic models. Deep learning techniques used in this AI system offer an end-to-end solution, eliminating the manual search for biomarkers and enhancing efficiency in COVID-19 detection.
In conclusion, the development of this AI system represents a significant advancement in COVID-19 diagnosis. With its rapid and accurate results, it has the potential to revolutionize the way we detect and combat the spread of the virus, particularly in regions with limited resources. The implementation of AI technology in healthcare continues to show promise, providing valuable support to medical professionals and improving patient outcomes.
The source of the article is from the blog reporterosdelsur.com.mx