Artificial Intelligence Can Predict Heart Failure
- When the blood pressure is high in the lungs, it damages the heart.
- About a third of patients die within 5 years of being diagnosed.
- The available options for treatment currently include drugs, injections straight into the blood vessels, and lung transplants.
- But doctors need to know how much time a patient is left with before recommending the right treatment.
- Researchers at MRC London Institute of Medical Sciences have developed an AI to predict when a patient's heart is about to fail.
- The software analyzes blood tests and scans of the heart to look for signs of heart failure.
- Training the AI
The AI was trained using MRI scans of 256 patients’ heart and blood test results.
- The data contained the info of movement of almost 30,000 different points on the heart during each heartbeat.
- Plus, this data of 8 years of patient records was provided.
- Using this data, the machine learnt to predict heart failure 5 years before it happens.
- High Accuracy
The machine was accurate 80% of the time in predicting who would still be alive after one year.
- The team says technology will help many patients who require aggressive treatment for heart disease.
Post a Comment