Diagnostic experts, such as pathologists, are in high demand, particularly in many low-resource countries, but the problem has also spread to Finland and other western countries during the past few years. In many cases it would be more efficient, faster and cheaper to consult a pathologist or microbiologist over a remote connection to acquire a diagnosis, potentially with the help of machine intelligence.
Johan Lundin’s research group at the University of Helsinki’s Institute for Molecular Medicine (FIMM) has risen to this challenge and developed a small, low-cost mobile microscope device. The device connects wirelessly to a cloud server to which it transmits images for machine vision analysis.
“To make mobile diagnostics possible, we first need a cheap and reliable device which is easy to move around and can digitise samples taken from the patient. However, just developing this device is not enough. The entire process, from digitisation and transmission of the image to diagnostics and returning the information back to the patient’s contact person, must be flawless,” Johan Lundin explains.
Lundin’s group designed the device, known as MoMic, to use very cheap components originally intended for mobile phones. However, MoMic differs from many add-on devices which seek to generate microscope-like magnification when installed on top of a mobile phone’s camera lens. The new device can attain laboratory-levels of microscopic definition and scan a sufficient area of the sample to enable diagnosis. The device is also small enough to fit in a typical camera bag.
Lundin, who works at both the Karolinska Institutet in Sweden and at FIMM, has recently received significant funding from two organisations to support the further development of the mobile microscope. The device and the full process of the remote diagnosis is currently being tested in field conditions in Tanzania, and another test period will be conducted later this year in Finnish operating rooms. Project partners include the University of Oulu, the University of Muhimbili in Tanzania and a few small Finnish companies.
In Tanzania, the research group will focus on developing remote diagnostics for cancer and parasitic diseases such as malaria. The country has approximately 45 million inhabitants, but only two dozen trained pathologists. Consequently, the ability to receive diagnostic services through a remote connection would significantly improve services to patients.
The Swedish Research Council, the largest funder of science in Sweden, has granted funding for the three-year field test in Tanzania. The field testing will involve several locations, including the village of Yombo near Dar es Salaam.
“The health clinic at Yombo is very modestly equipped. Electricity services are unreliable, but there are two mobile phone towers nearby,” explains Lundin.
Previously tied to the laboratory, the microscope can now be taken to the patient.
In Finland, Lundin’s group has recently received one year of funding from Tekes through its “New knowledge and business ideas from research” programme. With this funding, the research group will test the mobile microscope to see whether it can be used to support the necessary consultations with pathologists during cancer surgeries.
At hospitals involved in the one-year pilot programme, patient samples will be diagnosed both on location and remotely based on the digitised sample. The group hopes to use feedback from operating room staff and comparisons of on-site diagnoses to remote diagnoses acquired with the mobile microscope to obtain information that will help them develop the device further. Thanks to the Tekes funding and with help from Helsinki Innovation Services (HIS), the commercial potential of the invention is also being explored.
According to Lundin, the mobile microscope would also be well suited for research work, including the study of tissue and cell samples to identify the expression of biomarkers in cancer research.
An article published in the PLOS ONE journal in December confirmed the suitability of the mobile microscope for cancer diagnostics. The results indicated that the number of oestrogen receptors in breast cancer samples digitised with MoMic could be determined with the same accuracy as in samples digitised with significantly more expensive equipment.
Lundin is particularly interested in developing machine intelligence applications that can also help make diagnoses.
“I believe that we will be able to start testing the suitability of the new methods based on machine vision for imaging diagnostics already this year,” he states.
Research Director Johan Lundin
Institute for Molecular Medicine Finland (FIMM)
Tel: +358 50 415 5459