MDR-TB Treatment & Prevention
ICT tools for early diagnosis of tuberculosis
Contact coordinator:
------------------
Instituto de Medicina Tropical Alexander von Humboldt in Peru (Universidad Peruana Cayetano Heredia) collaborates with the CTIC Foundation developing in the project TEDS-TB (ICT Tools for Early Diagnosis Support) aims to contribute to research in artificial intelligence technologies and computer vision early warning of tuberculosis and strengthen health education through a platform virtual for the training workshops in the area rural and urban.
The "ICT tools for Early Diagnosis Support (TB) - TEDS" Project helps research using artificial intelligence tools and computer vision for early identification of prevalent and neglected illnesses (EPO). It also provides useful tools to enhance trainingi for health professionals that are not experts in microbiology for the TB diagnosis.
Artificial Intelligence:
Given this problem, the TEDS-TB poses a form of systematic part of the process by analyzing photographic images of sputum samples through a series of algorithms to isolate and quantify the bacilli. This is achieved by issuing an automatic outcome estimate of the degree of infection or severity of infection with a high degree of reliability. This is because the images of sputum, in addition to the rods, we can see remnants of dyes and also, sometimes, color characteristics similar to those of the bacilli. Therefore, this system is a helpful tool for health workers, especially for those who are not skilled in detecting bacilli, maximizing prevents discrimination and other areas that may be interpreted as potential bacilli.
Therefore, the contribution of the TEDS-TB in the field of health ICT is to provide an early warning tool on-line for diagnostic support. Since it may take preventive decisions about the possible contagion from the analysis of biological samples from a more objective view than just microscopy.
The project is I+ D "ICT Tools for Early Diagnosis Support (TB) - TEDS" contribute to research in artificial intelligence tools and computer vision applied for early warning and Neglected Illness (EPO - Prevalent and Neglected Diseases), and provide tools contribute to expand the training of health personnel are not experts in microbiology in diagnosis of TB. Of all the EPO i Tuberculosis (TB) is selected as a test case to test the system in a real scenario. However, although the ultimate goal of this project is to improve and expand the options for diagnosing TB, the system's success in the future will extend the proposed approach to other countries and EPO as an early warning and diagnosis which can be based in image analysis of smear samples. This system is based on the acquisition, analysis and feature extraction of digital images related to the samples (sputum) collected at health centers in the case of this project the real environment for the collection of samples is the district of San Juan Lurigancho in Lima, Peru.
The process is based on the method of diagnosis to the date recommended by the World Health Organization (WHO). The diagnosis of active TB from smear test is done by manual identification and counting of the population of TB bacilli in the sputum.
