AI-based computer modelling has produced a new way to diagnose a common heart condition much earlier than patients or doctors may spot the signs.
Millions of people worldwide live with atrial fibrillation, a condition that makes their heart beat irregularly and can increase the risk of stroke and longer-term heart problems later in life. Diagnosing the condition can be tricky because the irregularity of the patient's heartbeat can be intermittant and the effect subtle, but when it's spotted the condition is easy to confirm.
Several companies have been working on wearable heart monitors that can potenitally provide all-day data on heart activity, but sifting through that data for evidence of the medical condition is a real problem. Researchers at the Mayo clinic in the US have now reported that they've successfully used AI to diagnose the condition in a new computer model.
Diagnosis currently relies on catching an atrial fibrilation in the act while hooked up to heart moniitoring equipment, but the AI model was able to detect much more subtle indicators of the condition and could be used to identify early warning signs in asymptomatic individuals in a non-invasive test.
The researchers behind the breakthrough said that it could lead to earlier detection of the condition and that may help patients get the right treatment earlier, potentially saving lives. Heart attack and stroke are two of the biggest killers in the UK, and the uptake in wearable tech here may provide a wealth of data for training AI-based solutions.
Source: BBC News
