Low-cost mobile microscopes developed in Finland can be utilised for diagnosis of tropical parasitic infections and cancer. This year, prototypes of the mobile microscopes will be tested in field trials in Tanzania and Kenya. Rapid remote diagnostics supported by machine intelligence is the researchers' main objective.

Shortage of experts in the field of diagnostics is particularly prevalent in low-resource countries. Expert opinions given via a remote connection could significantly facilitate the diagnosis and proper treatment of patients in Sub-Saharan Africa.

A research group led by Johan Lundin, Research Director at the University of Helsinki’s Institute for Molecular Medicine Finland (a part of Helsinki Institute of Life Science), is studying devices and applications that enable artificial intelligence (AI) supported automated diagnostics. The MoMic mobile microscopy device is based on low-cost components, originally developed for mobile phones and here used for digitisation of biological samples. The device is connected wirelessly to a cloud server where the sample pictures are being analysed with machine vision or further transferred to a remote expert for diagnosis.

“If everything goes according to plans, the devices and applications developed by our group can significantly facilitate diagnostic processes in the health care systems of low-resource countries, which in turn makes it easier to provide proper treatment to patients”, said Oscar Holmström, a medical student completing his dissertation on mobile microscopy-based diagnostics.

The latest results of the studies were just recently published in the Global Health Action journal. Holmström analysed stool and urine samples digitised with the mobile microscope and showed that resolution is adequate for detection of schistosomiasis and a series of other parasitic infections. Also, preliminary results of AI-supported automated analysis are reported. The research group has already previously found that the mobile microscope is suitable for the analysis of breast cancer samples.

Field conditions in Africa, however, will be the real challenge for the methods. The research group has already visited the Fogo Fogo elementary school close to Lake Victoria in Northern Tanzania where a trial will be conducted in cooperation with Muhimbili University. During a visit to the field in February, the group collected samples from schoolchildren in the local school for diagnosing schistosomiasis. With the help of these samples, the group has further been developing the diagnostic algorithms based on machine learning. Later in the autumn, the researchers will return to the village to test the diagnostic methods in practice.

“During the first field visit we also tested sample scanning and transfer for remote diagnostics. Local mobile phone networks were fast and reliable enough to transfer pictures, but lack of electricity at the school came as a surprise. For the next field trial, the mobile microscope will be equipped with an improved rechargeable battery”, says Lundin, who, in addition to the University of Helsinki, works at Karolinska Institutet in Sweden.

There are less than one pathologist per one million people in average in the African countries. Thus a very limited number of Pap smears to identify premalignant conditions in the cervix are performed. This lack of cytology-based screening programs leads to delayed cancer diagnoses; cervical cancer is, in fact, the most common cause of death in African women.

According to Nina Linder, MD, PhD who works in Lundin’s research group, the mobile microscope could also be used for screening Pap smear samples.

“We are currently using biobank samples to develop an algorithm that can identify the most suspicious cell changes. The aim is to test Pap smear sample analysis with the mobile microscope in field conditions in cooperation with the Kinondo Hospital in Ken