Ugandan innovators use smartphones to detect malaria

Fun, but at best this article and the linked blog are missing critical details. Other organizations (e.g. http://ianxen.com/) have been working on similar technology involving optics, image processing and a blood sample -- and the associated consumables.

Even assuming the tech is sorted the more interesting questions involve how, from a health systems perspective, these things can be deployed in rural health centers at a scale that matters. Some hard training/security/procedure/money questions to figure out, and I haven't heard much about these issues yet. Unclear if it's because the technology isn't ready for prime time or other reasons.

From the article it looks like they use a technique similar to the 'Heart rate monitor' apps, but instead of trying to look for pulses they look to quantify the blood oxygen content, perhaps using colour? and from that get a diagnosis on a malarial infection.

You've got to love non-invasive tests like this, but the trade-off is lower accuracy.

In this case, for example, it could give false positives for people who have differently shape blood cells - e.g. sickle cell, spherocytosis, elliptocytosis.