AI Detects Anemia Through Fingernail Photos - A Helpful Solution for Billions

According to the World Health Organization (WHO), nearly 2 billion people worldwide are living with anemia, a common blood disorder that occurs when the body does not have enough healthy red blood cells or when the amount of iron-rich protein in red blood cells called hemoglobin (Hgb) is lower than normal.

Hemoglobin is the protein responsible for transporting oxygen in the blood, so when deficient, the patient may feel prolonged fatigue, shortness of breath and reduced ability to concentrate. If not detected and treated promptly, chronic anemia can lead to serious complications such as heart failure, damage to internal organs, and is especially dangerous for pregnant women.

Previously, to diagnose anemia, patients often had to go to medical facilities to perform specialized blood tests such as complete blood count (CBC), hemoglobin and hematocrit levels, or evaluate red blood cell morphology through peripheral blood smears, procedures that are both time-consuming and require specialized medical equipment.

But what if you could just take a photo of your fingernails with your phone to detect your risk of anemia? Researchers at Chapman University (USA) have turned this idea into reality with a mobile application that uses artificial intelligence.

This application can analyze fingernail images, which contain many small, easily observable capillaries, to detect signs of anemia in a non-invasive, convenient and potential way for at-home health screening.

How effective is it really?

This AI application has been widely deployed in a medical study with more than 200,000 users across the US, performing more than a million tests.

According to the research team at Chapman University, the initial results are very positive, the application is not only easy to use but also shows the potential to become a large-scale anemia surveillance tool, especially suitable for communities with limited access to traditional health services.

Specifically, the app’s AI model achieved 89% sensitivity and 93% specificity, numbers that are considered close to the gold standard compared to laboratory tests. This means that the app can accurately identify most cases of anemia, while minimizing the risk of misdiagnosis.

Notably, once users were diagnosed with anemia and started using the app in a personalized mode, the error rate continued to drop significantly thanks to the AI-driven real-time analysis of personal data.

An accessible digital platform like this could help hundreds of millions of users self-monitor their hemoglobin (Hgb) levels quickly and conveniently, without having to visit a hospital or undergo traditional blood tests that are costly and time-consuming.

Previously, in 2020, the US healthcare startup Sanguina also introduced an application called AnemoCheck for patients with chronic anemia. However, this application is oriented as a personal health monitoring tool, not pursuing the official approval process from health regulatory agencies.

Additionally, a similar app was piloted in India's public health system two years ago, and was found to be accurate enough to be used for community screening.

Core Benefits: Proactive Surveillance, Community Health Support

According to the team at Chapman University, the app is not designed to completely replace traditional medical testing methods, nor is it a tool for users to self-diagnose diseases.

Instead, it acts as an early warning system, helping users recognize unusual signs related to anemia, thereby deciding whether to see a doctor or perform in-depth testing, especially during times when the disease shows signs of worsening.

“The app is particularly useful for people with chronic anemia,” the team said. “For example, patients with kidney failure, cancer, or other medical conditions that require regular hemoglobin monitoring. Being able to test for anemia at home, quickly and non-invasively, provides convenience and peace of mind.”

The highlight of the app is the ability to personalize with AI. When this feature is activated, the accuracy of diagnosing anemia in the target user group (i.e. patients with known underlying diseases) increases by up to 50% compared to using the general mode.

This allows for the establishment of continuous health monitoring procedures, thus paving the way for early medical interventions, even without formal laboratory testing.

In addition to the personal benefits, the app also integrates GPS to create what the team calls “the first county-by-county anemia prevalence map in the United States.”

Thanks to its ability to collect data on a large scale, this technology not only serves individual users but also has great significance in the field of public health such as supporting population screening, detecting areas with high anemia rates and allocating medical resources more reasonably.

The study was published inProceedings of the National Academy of Sciences (PNAS)– Proceedings of the National Academy of Sciences of the United States, one of the world's most prestigious scientific journals.