Ayni Lab (from the Quechua word that means collaboration or mutual help), the social innovation laboratory of the Ministry of Development and Social Inclusion of Peru, promoted a call to find innovative solutions for an important public health issue in the country: anemia. This condition is suffered by more than 43% of children between six months and almost three years of age. The World Health Organization (WHO) classifies this condition as “serious” when it exceeds 40%.
Anemia is suffered when there are not enough red blood cells in the blood. It implies that not enough oxygen is transported to the tissues and organs of the body. In practice, for a three-year-old child, the condition could result on life-time physical and mental underdevelopment, even though the condition is overcome later.
Before curing this condition, there is an essential step: properly diagnose the affected population. José Enrique Velásquez, technical secretary of Ayni Lab, explains that there are different types of detection obstacles: from families that reject the punctures with which blood samples are obtained, to insufficient resources (human and physical) to properly collect and store those samples in remote areas.
That is why Ayni Lab posed the challenge of finding a non-invasive diagnostic method, which does not require puncture; and the answer to this challenge resulted Professor Mirko Zimic, head of the bioinformatics laboratory at the Universidad Peruana Cayetano Heredia. “When I reviewed the call, an image of childhood crossed my mind,” says Zimic. “The doctor used a flashlight, lowered the eyelid and watched us.” Indeed, the inner face of the lower eyelid, depending on its paleness, may reveal anemia. “We set out to move that classic procedure to one that was more objective,” adds Zimic.
The result –after winning the call and accessing about US$90,000 of funding– is an application used on a smartphone, which uses photos of the so-called palpebral conjunctiva (the inner face of the lower eyelid). According to its coloration, an algorithm determines if the diagnosed person suffers from severe or moderate anemia. In addition, it is not necessary for technical or specialized personnel to do the procedure.
In a first phase they have tested this technology with almost 600 children from the department of Lima. To contrast the results, blood samples were taken from the same group of children that underwent the cyanmethemoglobin measurement, the most reliable test for the quantitative determination of hemoglobin. This whole process allowed them to calibrate the system thanks to the use of artificial intelligence.
The first conclusions are exciting: Professor Zimic’s detection method coincides more than 90% of the time with the laboratory test in cases of severe or moderate anemia. For the detection of mild anemia (which is the least severe) so far they have reached 87% effectiveness. In order to improve these numbers, they prepare a second phase of data collection with 500 more children.
Source: Agencia ID