AI Doctor: Cardiologists develop AI to detect heart conditions, predict attacks with 99.6% accuracy

A new AI tool has the potential to revolutionize the diagnosis of heart attacks, providing faster and more accurate results. Researchers have created an algorithm that aims to alleviate the burden on emergency departments and offer reassurance to patients experiencing chest pain.

According to a recent study, the algorithm outperformed existing testing methods by accurately ruling out heart attacks in over twice as many patients, achieving an impressive accuracy rate of 99.6 per cent. The research team, hailing from the University of Edinburgh, believes that this capability to swiftly exclude a heart attack could significantly reduce hospital admissions and quickly identify patients who can safely return home.

Current practices lead to misdiagnosis, especially among women
Currently, the standard procedure for diagnosing a heart attack involves measuring troponin protein levels in the bloodstream. However, this approach employs a universal threshold for all patients, disregarding factors such as age, gender, and other underlying health conditions that can influence troponin levels. As a result, the accuracy of heart attack diagnoses is compromised.

Previous studies have indicated that women are 50 per cent more likely to receive an incorrect initial diagnosis, and individuals who experience misdiagnosis face a 70 per cent higher risk of mortality within 30 days.

The research team has introduced a new algorithm named CoDE-ACS as a potential solution to address this issue. The algorithm was developed utilizing data from 10,038 patients in Scotland who presented at the hospital with suspected heart attacks.

CoDE-ACS, the game changer
CoDE-ACS incorporates various patient information routinely collected, including age, gender, electrocardiogram (ECG) results, medical history, and troponin levels, to predict the likelihood of an individual having suffered a heart attack. The algorithm generates a probability score ranging from 0 to 100 for each patient.

Professor Nicholas Mills, leading the research, emphasized the significance of early diagnosis and treatment for patients experiencing acute chest pain resulting from a heart attack. He acknowledged the challenges in differentiating between various conditions that manifest similar symptoms, making accurate diagnosis a complex process.

Professor Nilesh Samani, the Medical Director of the British Heart Foundation, which provided funding for the research, expressed the immense potential of utilizing data and artificial intelligence to enhance patient care and increase efficiency in busy Emergency Departments.

Chest pain is a common complaint leading people to seek care in Emergency Departments worldwide. Doctors face the daily challenge of distinguishing between patients experiencing chest pain due to a heart attack and those with less serious causes.

Big boon for emergency wards and departments
The development of CoDE-ACS, employing advanced data science and AI techniques, holds the promise of more accurate identification and exclusion of heart attacks compared to current methods. This tool has the potential to revolutionize Emergency Departments by expediting the diagnostic process and providing better outcomes for patients.

In the UK alone, there are approximately 100,000 hospital admissions annually due to heart attacks, equivalent to one admission every five minutes. Clinical trials are currently underway in Scotland to evaluate the effectiveness of the AI tool in helping doctors alleviate the strain on overcrowded emergency departments.

The research findings have been published in the journal Nature Medicine, highlighting the significance and credibility of the study.

Read all the Latest News, Trending News, Cricket News, Bollywood News,
India News and Entertainment News here. Follow us on Facebook, Twitter and Instagram.