LONDON: AI scans that measure the size of a person’s chest muscles could help doctors pinpoint patients at risk of heart failure or death.
Researchers used the technology to identify those with sarcopenia, an age-related progressive loss of muscle mass and strength.
The condition is common in older people with health problems, but is rarely diagnosed as accurate measurements require specialised scans, leaving doctors to use simpler scores such as body mass index.
To overcome this, experts at Queen Mary, University of London developed a tool to give patients a sarcopenic obesity index, which is calculated using a person’s weight and size of the pectoral muscles in the chest.
To get a score, a person’s weight in kilograms is divided by their pectoral muscle mass in grams.
For the study, published in the Journal of the American College of Cardiology, AI was used on the anonymised scans of more than 55,000 people.
After tracking patients’ health for an average of four years, the study found that among those with a sarcopenic obesity index score in the top 15%, more than half faced a higher risk of death from heart problems.
They also had a 37% risk of death from any cause, the study suggests, as well as a 31% higher chance of heart failure.
"Ageing and obesity are two of the most central health problems of our time, and both are key risk factors for heart disease," said lead author Dr Mihir Sanghvi, clinical lecturer in cardiology at Queen Mary, University of London.
"Sarcopenic obesity represents a meeting between the two – it is a crucial sign of health deterioration in obesity, especially for older people," Sanghvi said.
"We have developed a deep learning tool to measure this, in the hope that it will allow doctors and patients to take action to protect their health," he said.
"This new tool can help identify people who most need interventions to prevent further muscle loss and a worsening of their condition.”
Those with higher sarcopenic obesity index scores tended to be older, have higher blood pressure and blood sugar levels, were less likely to be active and more likely to come from a deprived area.
Researchers are hopeful the technology can be used to detect risks in people having a heart MRI scan, which are performed on millions of people around the world every year.
The AI tool will be tested in clinical practice later this year.
Professor Bryan Williams, chief scientific medical officer at the British Heart Foundation – which funded the study – said: "This intriguing study demonstrates how the latest technologies could be used to expand our abilities to predict risk and save lives.
"By finding a new, easier way to measure sarcopenic obesity, this observational study has opened up many possibilities for spotting the condition and assessing the damage it can do to patients’ heart health.
"It is important to now test whether it can be an effective addition to a doctor’s toolkit as they work to reduce heart failure.” – dpa
