SciDev.Net

A web-based microscopy tool that could improve diagnosis of parasite-related diseases has been developed by Finnish and Swedish scientists.

The system enables local researchers or doctors to compare what they are viewing down their microscope with a 'gold standard' online and thus improve the quality of their diagnoses.

Ewert Linder, lead author of the study and a researcher at the Sweden-based Karolinska Institute, told SciDev.Net that there is tremendous over-diagnosis of infections such as schistosomiasis and malaria in the developing world.

"In the laboratory, people are very eager to come up with a diagnosis," he says. "But [diagnosis] is not always founded on observations that are reliable."

This has implications not only for patients, but also for the reliability of data that is used to calculate the prevalence of neglected tropical diseases and the effects of interventions, the researchers write in PLoS Neglected Tropical Diseases.

The team captured up to 50,000 images of a series of parasite specimens on ordinary glass slides at high magnification, and digitally stitched them together to create a representation of the slide. They then 'stacked' the images to allow for three-dimensional navigation online.

The tool uses the viewing system employed by geospatial imaging systems such as Google Maps, meaning that while the quantities of data are large, a viewer does not need a powerful computer to access the images on the Internet.

Alternatively, a local hospital or university with limited connectivity to the Internet could install a local server, which would overcome data transfer problems cheaply and without complex technical requirements, say the researchers.

Linder says a future direction for the project could involve remote diagnostics for diseases such as malaria. "It would be possible to take a picture with your local microscope somewhere in the bush and send this to a central laboratory. We actually have a program now — an algorithm — which can interpret these pictures so that we can digitally say how many parasites there are in this malaria patient."

Link to full paper in PLoS Neglected Tropical Diseases