Artificial intelligence algorithms developed by Near East University can diagnose COVID-19 in a short time by analyzing PCR tests. The system, which was proven to work with 100 percent accuracy by experienced molecular microbiology experts, can categorize PCR results as positive, low positive, negative, and uncertain.
Analyzing the results of PCR tests, which is considered the gold standard in the diagnosis of COVID-19, had put a strain on health systems all over the world during the pandemic due to the intensity they caused. In order to get results in a short time, determining the results of PCR tests, which creates a great burden, especially on laboratory staff, also requires experienced staff.
The artificial intelligence algorithms developed by the Near East University analyze the PCR test data with an interface that can be integrated with PCR devices, allowing the test result to be determined within seconds without the need for human intervention.
The study, co-designed by Near East University Acting Rector Prof. Dr. Tamer Şanlıdağ, DESAM Research Institute researchers, Near East University Faculty of Medicine Department of Medical Microbiology and Clinical Microbiology Lecturer Assoc. Prof. Dr. Buket Baddal and Cyprus International University Engineering Faculty Computer Engineering Faculty Member Assist. Prof. Dr. Emre Özbilge, is of great importance in terms of shortening the analysis and reporting processes of COVID-19 PCR tests, and initiating the isolation and treatment processes of patients rapidly. The developed artificial intelligence application can also be used in possible future epidemics.
It can categorize PCR results as positive, low positive, negative and indeterminate.
With the developed deep learning algorithm, the fluorescent radiation data obtained from the PCR device is loaded into the pre-trained program for each patient. The artificial intelligence system, which has been tested with the PCR results of hundreds of patients, can define the graphics of the patient samples and give the result in seconds.
The artificial intelligence system developed at the Near East University using a multi-layered neural network model can categorize PCR data as positive, low positive, negative or uncertain without laboratory personnel. In case of uncertain results, the system warns staff that the patient sample needs to be reworked.
The system, which was provided by experienced molecular microbiology experts during the development stages, can give results with 100 percent accuracy. Thanks to the system, it is aimed to reduce the workload of laboratory personnel and to ensure efficient use of personnel during pandemic periods when sample flow is very intense.
The developed artificial intelligence system was shared with the scientific world at the 32nd European Congress of Clinical Microbiology and Infectious Diseases held in Portugal.
The study presented by Assoc. Prof. Dr. Buket Baddal was met with great interest by the microbiology community. The study presented at the “COVID-19 Diagnosis: New and Newer” session, where new technologies were discussed at the congress, which brought together experts in the field of infectious diseases and microbiology, was appreciated by scientists because it can be applied to different pathogens and can be integrated into laboratories all over the world.
The study was shared with the whole scientific community by being included in the “Applied Artificial Intelligence” special issue of the “Applied Sciences” magazine of the prestigious Multidisciplinary Digital Publishing Institute, published in October 2022.
Near East University Acting Rector Prof. Dr. Tamer Şanlıdağ, on the other hand, stated that they have been conducting scientific research in a multi-faceted way since the first day during the COVID-19 process, and said, “While transforming the studies carried out by our scientists within our university into products that will allow the COVID-19 process to be managed more effectively, we continue to share the results of our studies with the scientific world with scientific articles.”
Reminding that Near East University scientists have published 375 articles in the world’s leading scientific publications during the COVID-19 process, Prof. Dr. Şanlıdağ said, “We shared our artificial intelligence application, which analyzes PCR tests and determines the results with 100 percent accuracy, with the scientific world at the European Clinical Microbiology and Infectious Diseases Congress held in Portugal. At the same time, we published our work in the ‘Applied Artificial Intelligence’ special issue of the journal Applied Sciences.
DESAM Research Institute researcher, Near East University Faculty of Medicine, Department of Medical Microbiology and Clinical Microbiology Lecturer Assoc. Prof. Dr. Buket Baddal, on the other hand, said that PCR tests, which entered our lives intensively with the COVID-19 pandemic, are also used in the identification of many pathogens that cause infectious diseases, and said, “The artificial intelligence application we have developed can also be used in the diagnosis of new infectious disease agents that may arise in the future and viruses that may lead to pandemics.” Assoc. Prof. Dr. Baddal said, “With the COVID-19 pandemic, it was revealed how unprepared the health systems were for such pandemics. With the artificial intelligence algorithms we have developed, we will be prepared for future pandemics. By diagnosing the disease early, we can isolate these people early and prevent the spread of the disease in the community.”