Malaria System: a New Tool for Automatic Diagnosis of Malaria in Mobile Devices
|
If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL). |
Abstract
Background: Malaria is a worldwide public health problem and it is mainly related to remote areas. In this way, low cost systems for automatic diagnosis has become a priority investigation in several research groups. On the other hand, there are new cases due to climate changes which allow the survival of Anopheles in areas previously not inhabited. In this way, the coming years will see a great demand for in loco diagnostic and evaluation systems to these new areas.
Objective: Mobile devices have been viable alternatives in several health systems and epidemiological control. The research groups EpiSchisto Risk Modeling (www.epischisto.org) and Discrete Modelling and Simulation of Biological Systems (http://mosimbio.upc.edu) have developed automatic diagnostic tools for diseases including schistosomiasis and malaria. In this work, a detection system for Plasmodium parasites (malaria) at low cost with mobile devices is presented.
Methods: This work is under the computer vision area that theoretically simulates the human visual (SHV). The SHV is responsible for this visual perception of a human being (in recognizing objects, shapes, among others). Computer vision techniques seek to apply this knowledge to a computerized machine. This operation is not as simple one since there are some perceptions of the human visual system that has not yet been completely explained. The use of techniques of computer vision and artificial intelligence in this project is the theoretical foundations of the detection system for low-cost malaria presented here. The development phases of the system are: capture image - the image is captured using a coupled experimental mobile device system and a microscope; segmentation - operation to reduce the computational cost. This step consists in removing the background around blood cells, thereby reducing the search scope of the recognition algorithm; training and classification - sorting through artificial intelligence techniques for training with positive and false positive images (with and without malaria parasites) partitioned into 80% training and 20% for validation or testing.
Results: An Android based system was developed, it finds the Plasmodium infected parasites in the image. The execution time has an average time of 1000ms to detect each parasite. The system is limited on the identification of the parasite into P.falciparum specie (step trophozoite). In this phase the parasite has an annular shape. The system were tested in a set of 465 images with 50 positive ones and 415 negative ones. The hit rate of the system is currently at 60%, where so every 100 infected 60 parasites are identified.
Conclusions: The main contribution of the Malaria System is its low cost and its usability. This is special when used in remote areas by Public Health agents. In these places, the human capital is also a scarse resource and nevertheless the results can be seen in loco and in real time. The results will be sent to a central unit control once the device get internet connection.
Objective: Mobile devices have been viable alternatives in several health systems and epidemiological control. The research groups EpiSchisto Risk Modeling (www.epischisto.org) and Discrete Modelling and Simulation of Biological Systems (http://mosimbio.upc.edu) have developed automatic diagnostic tools for diseases including schistosomiasis and malaria. In this work, a detection system for Plasmodium parasites (malaria) at low cost with mobile devices is presented.
Methods: This work is under the computer vision area that theoretically simulates the human visual (SHV). The SHV is responsible for this visual perception of a human being (in recognizing objects, shapes, among others). Computer vision techniques seek to apply this knowledge to a computerized machine. This operation is not as simple one since there are some perceptions of the human visual system that has not yet been completely explained. The use of techniques of computer vision and artificial intelligence in this project is the theoretical foundations of the detection system for low-cost malaria presented here. The development phases of the system are: capture image - the image is captured using a coupled experimental mobile device system and a microscope; segmentation - operation to reduce the computational cost. This step consists in removing the background around blood cells, thereby reducing the search scope of the recognition algorithm; training and classification - sorting through artificial intelligence techniques for training with positive and false positive images (with and without malaria parasites) partitioned into 80% training and 20% for validation or testing.
Results: An Android based system was developed, it finds the Plasmodium infected parasites in the image. The execution time has an average time of 1000ms to detect each parasite. The system is limited on the identification of the parasite into P.falciparum specie (step trophozoite). In this phase the parasite has an annular shape. The system were tested in a set of 465 images with 50 positive ones and 415 negative ones. The hit rate of the system is currently at 60%, where so every 100 infected 60 parasites are identified.
Conclusions: The main contribution of the Malaria System is its low cost and its usability. This is special when used in remote areas by Public Health agents. In these places, the human capital is also a scarse resource and nevertheless the results can be seen in loco and in real time. The results will be sent to a central unit control once the device get internet connection.
Medicine 2.0® is happy to support and promote other conferences and workshops in this area. Contact us to produce, disseminate and promote your conference or workshop under this label and in this event series. In addition, we are always looking for hosts of future World Congresses. Medicine 2.0® is a registered trademark of JMIR Publications Inc., the leading academic ehealth publisher.
This work is licensed under a Creative Commons Attribution 3.0 License.