Healthcare: Diagnosing anaemia with a photo

A smartphone app for the detection of anaemia is presented in Nature Communications. The app detects anaemia by analysing photos of fingernail beds taken with a smartphone, and estimates haemoglobin levels. This approach could potentially lead to the replacement of blood-based laboratory tests, which are currently required to diagnose and monitor anaemia.

Anaemia affects over two billion people globally. Existing clinical approaches to measure haemoglobin levels require specialized equipment and represent trade-offs between invasiveness, accuracy, infrastructure requirements, and cost, all of which are problematic in rural and low-resource settings, where anaemia is most prevalent.

Wilbur Lam and colleagues developed an algorithm that calculates the concentration of haemoglobin in the blood by analysing colour and technical metadata from fingernail bed images. In this way, the algorithm, incorporated into a mobile app, can function without the need for any equipment other than the smartphone itself. In a clinical assessment involving 100 participants, the app estimated haemoglobin concentrations with high sensitivity and accuracy comparable to currently available diagnostic tools for the detection of anaemia. In a further assessment involving four participants the app showed high levels of accuracy when used as a monitoring tool.

The authors suggest the app may allow for screening of anaemia in regions lacking specialized equipment and trained personnel, and enable patients with the condition to monitor their haemoglobin levels remotely in less than a minute. Nevertheless, additional studies with a larger number of participants are needed to confirm the high level of diagnostic accuracy necessary to replace blood-based anaemia testing.

doi:10.1038/s41467-018-07262-2

Return to research highlights