Novel support system "man and machine" decision for the diagnosis of malaria infection -
A Finnish-Swedish research group at the Institute of Molecular Medicine Finland (FIMM), University of Helsinki and Karolinska Institutet, Stockholm, has developed a new system of "man and machine" decision support for diagnosis of malaria infection. This innovative diagnostic aid was described in PLoS One journal today August 21 the method is based on computer vision algorithms similar to those used in facial recognition systems combined with visualization of only the most relevant areas for diagnosis. Tablet computers can be used in the visualization of images.
in this new method, a thin layer of blood spread on a microscope slide is first scanned. the algorithm analyzes more than 50,000 red blood sample and the class based on the probability of infection. Then, the program creates a panel containing images of more than one hundred very probably infected cells and presents this panel to the user. The final diagnosis is made by a health care professional based on the displayed images.
Using a set of samples already existing diagnosed, the researchers were able to show that the precision of this method is comparable to the quality criteria defined by the World Health Organization. In the test frame, more than 0% of infected samples were accurately diagnosed based on the panel. The few problematic samples were of poor quality and in true diagnostic framework would have led to further analysis.
"We are not saying that the process of diagnosis of the whole malaria could or should be automated. Rather, our goal is to develop methods that are much less work than the traditional and have a potential significantly increase the flow in the diagnosis of malaria, "said the research director Johan Lundin (MD, PhD) of the Institute for molecular medicine Finland FIMM.
"the equipment necessary for the sample scan is a challenge in developed countries. In the next phase of our project, we will test the system in combination with mobile devices microscopy cheap as our group has also developed, "said first author shared section Nina Linder (MD, PhD) of FIMM.
support system developed can be applied in various other fields of medicine. In addition to other infectious diseases such as tuberculosis, the research group plans to test the system comes the diagnosis of cancer in tissue samples.
"There is also a strong need for rapid and accurate methods to measure the load of the malaria parasite in a sample. Different malaria drug testing programs are underway and parasite load in large numbers samples must be quantified to determine the efficacy of potential drugs. We also develop computer algorithms used in this study to address this need, as well, "Dr. Linder continued.
There are over 0 million new cases of malaria each year. the high quality microscopy is still the most accurate method for the detection of malaria infection. However, microscopy requires a well-trained and can take time when it is run as recommended highly. in 2012, less than half of suspected malaria in sub-Saharan Africa received a diagnostic test. the workload of health personnel is excessive thus contributing to demonstrate a low accuracy microscopy.
"The new imaging and analytical method can revolutionize the point of care diagnostics not only malaria, but also several diseases where the diagnosis depends on the microscopy. Action can lead to" market disruption "in the diagnosis of the disease," says Professor Vinod Diwan of Karolinska Institutet.
EmoticonEmoticon