Artificial Intelligence to Be Used in Heart Failure Diagnosis

Heart Failure Diagnosis to Be Made Using Artificial Intelligence The "Artificial Intelligence and Digital Transformation in Healthcare" project, implemented by Mersin University Faculty of Medicine with the support of AstraZeneca Turkey, has been presented. The project results were also shared at a press conference attended by stakeholders. In the "Artificial Intelligence and Digital Transformation in Healthcare" project implemented by Mersin University Faculty of Medicine with the support of AstraZeneca Turkey, artificial intelligence technology is being used in early diagnosis of heart failure. According to initial findings, the program tested on chest X-rays of 57 patients resulted in heart failure diagnosis in 49 patients. The innovative diagnostic protocol's initial results were also presented at the introduction meeting. The initial results of the "Artificial Intelligence and Digital Transformation in Healthcare" project, implemented by Mersin University Faculty of Medicine with the support of AstraZeneca Turkey, were shared at a press conference. At the conference, Mersin University Rector Prof. Dr. Ahmet Çamsarı, Faculty Member of Mersin University Cardiology Department Prof. Dr. Ahmet Çelik, Chair of the Board of Turkish Thoracic Radiology Society Prof. Dr. Recep Savaş, and AstraZeneca Middle East and Africa Region Medical Director Dr. Viraj Rajadhyaksha presented the research findings to the public. The project was initiated in January 2021 by Prof. Dr. Ahmet Çelik, Faculty Member of Mersin University Cardiology Department. Within the scope of this project, chest X-rays of patients who presented to Mersin University Faculty of Medicine Hospital for any reason were stripped of personal information and uploaded to a dedicated platform independently from the hospital's main server. Qure AI, an artificial intelligence solutions provider that cooperates within the framework of the company's "Emerging Markets Health Innovation Centers Program," conducted analysis by retrospectively scanning X-rays stripped of personal information. Subsequently, it simultaneously identified suspicious chest X-rays including heart enlargement and fluid accumulation around the lungs—"cardiomegaly abnormalities and pleural effusion."

Expected to be Beneficial in Early Detection of Other Diseases

Cardiologists evaluated these suspicious X-rays and contacted patients with abnormalities for advanced analysis such as echocardiography and BNP-ProBNP level testing for definitive diagnosis. Due to this method's application using artificial intelligence technology screening 29 different parameters such as tuberculosis, lung nodules, and cavities, it is expected to be beneficial not only in detecting abnormalities related to heart failure but also in early detection of other diseases such as lung cancer and tuberculosis.

Difficult-to-Diagnose Heart Failure Could Be Detected

Last week, Prof. Dr. Ahmet Çelik shared the project's outputs at the American College of Cardiology Congress. Çelik stated: "At the beginning of this project, in which artificial intelligence scanned chest X-rays of approximately 10,000 patients, artificial intelligence diagnostic results were verified by Mersin University Radiology Department. From the X-rays of the first 5,623 patients scanned with artificial intelligence, we detected in 119 cases a combination of fluid in the lungs and heart shadow enlargement observed in some heart failure patients. We reached 57 of the 119 patients invited to the Cardiology Department of Mersin University Hospital for advanced testing and definitive diagnosis. Following advanced analysis and definitive diagnostic tests such as NT-proBNP and echocardiography, a definite diagnosis of heart failure was made in 49 out of 57 patients in accordance with heart failure guideline diagnostic criteria. In this study conducted for early diagnosis of heart failure with the aid of artificial intelligence, artificial intelligence achieved the ability to predict heart failure by examining chest X-rays with 89.1 percent sensitivity and 86.4 percent specificity. More importantly, among 49 patients diagnosed with heart failure, 32 cases showed Heart Failure with Preserved Ejection Fraction, which is called difficult-to-diagnose heart failure."

Will Be Among the First Projects Using Artificial Intelligence

Mersin University Rector Prof. Dr. Ahmet Çamsarı, referring to studies conducted at his faculty, shared that Mersin University closely follows technological developments and supports research and development work carried out in this field. Çamsarı stated: "The results we heard at today's conference are quite gratifying for us. Our greatest achievement in this project was to improve patients' quality of life and extend their lifespans through early treatment opportunities. On the other hand, thanks to this project, early diagnosis of heart failure patients can now reduce hospitalizations due to heart failure. Our project will be among the first projects in our country and even in the world where artificial intelligence is used in early diagnosis of patients who are suspected of having heart failure and have not been diagnosed. In line with the results obtained, we aim to expand the project nationally and implement it in every chest X-ray taken. We also hope that these systems can be used in other fields such as oncology from a radiological perspective and that artificial intelligence projects that touch patients' lives can be implemented."

Artificial Intelligence May Help Improve Lung Cancer Detection

Turkish Thoracic Radiology Society Board Chair Prof. Dr. Recep Savaş also made statements. Savaş said: "Artificial intelligence is successfully applied in various healthcare settings. Indeed, according to a report prepared by Signify Research, the global market for medical imaging artificial intelligence applications is estimated to reach approximately USD 1.2 billion by 2025, with a compound annual growth rate of 26 percent. The majority of the medical imaging artificial intelligence market, 85 percent, is comprised of four clinical application branches: cardiology, neurology, breast, and pulmonology. It is estimated that these four clinical applications will still account for 82 percent of total revenue by 2025. Additionally, studies show that an artificial intelligence algorithm trained to detect lung nodules can help improve lung cancer detection in chest radiographs."

"We Will Continue Our Work to Best Benefit from Artificial Intelligence and Data Sources"

AstraZeneca Middle East and Africa Region Medical Director Dr. Viraj Rajadhyaksha emphasized the importance of collaboration between scientists and healthcare professionals in fighting diseases. Rajadhyaksha stated: "Early diagnosis in heart failure patients is very important in treatment planning and predicting the course of the disease. This project, by implementing advanced artificial intelligence and machine learning approaches in patients who go to different units for various reasons, will enable patients with early diagnosis to have their lives touched and meet with correct treatments much earlier. The results emerging from this research have the potential to create the first early diagnostic tool for heart failure in the world. Life-altering diseases that are still the leading cause of cancer deaths globally and affect more than 2 million people, such as lung cancer, can be prevented through early diagnosis. As a company, we will continue our efforts to best benefit from artificial intelligence and data sources, as in this project where we provided unconditional support to Mersin University Faculty of Medicine Hospital."