According to a new study published in the European Heart Journal, soon sending a “selfie” to the doctor could be a cheap and simple way of detecting heart disease. The study shows that a deep learning computer algorithm could detect coronary artery disease (CAD) by analyzing four photographs of a person’s face.
Although the algorithm requires further refinement and external validation in other populations and ethnicities, it has the potential to be used as a screening tool to identify possible heart disease in people for further clinical investigations.
Humans find it difficult to predict and quantify heart disease risk by looking at certain facial features, including thinning or grey hair, ear lobe crease, wrinkles, xanthelasma and arcus cornea (fat and cholesterol deposits appearing in the outer edges of the cornea).
But the algorithm happened to work exceptionally well on predicting heart disease risk. The algorithm correctly detected heart disease in 80% of cases and also accurately detected heart disease (sensitivity) that was not present in 61% of cases (specificity).
The algorithm demonstrated a moderate performance. Moreover, additional clinical information did not improve its performance, concluding that facial photos could be used to predict potential heart disease.
If we start using selfies as a screening method, it can enable a simple yet efficient way to filter the general population towards more extensive clinical evaluation. Countries or regions that have limited medical capabilities will benefit from this algorithm to screen cardiovascular disease.
However, the algorithm faces some limitations. This includes the low specificity of the test, improving the test and validation among larger populations and also that it raises some ethical questions about “misuse of information for discriminatory purposes”. The technology could be a threat to personal data protection, potentially affecting insurance options.
