INFECTIOUS DISEASE OPEN SCREENER

Elisee JAFSIA

Funding Area: Capacity building / Construcción de capacidad

Problem Statement: Infectious diseases are illnesses due to pathogens or their toxic products arising through transmission from an infected person, an infected animal, or a contaminated inanimate object to a susceptible host. Mainly prevalent in tropical areas, they are responsible for 2.2 million death per year in Africa. Infectious diseases are endemic among marginalized populations, due to the unavailability or inadequacy of diagnostic tests that cause empirical misdiagnosis. Microscopy of clinical specimens is a rapid and inexpensive method for the presumptive diagnosis of certain infectious diseases. Examination of stained smear gives information on the inflammatory response, as well as the bacteria involved. Accurate interpretation of the smear, however, is often time-consuming, error-prone and requires some training and experience, which limit the access of the method to low-income communities. Infectious disease testing has seen a surge in miniaturization, automation, and increasing computing power, creating a unique opportunity to exploit Machine Learning (ML). Thanks to the tight collaboration with the DIDA network, we contributed to the significant progress made toward obtaining a diagnostic tool for Malaria and Typhoid Fever (https://github.com/Mboalab/Open-Diseases-Screener-App). The amazing results obtained, is an invitation to maintain, improve and build a community around the infrastructure in order to help healthcare practitioners in their diagnostic tasks

Proposed Activities: Implementation Phase (2 months) • Assemble a multidisciplinary team including software engineers, designers, and healthcare experts. • Collaborate and configure the infrastructure using open-source technologies. • Deploy and test the infrastructure in a controlled environment. • Engage open-source experts to provide technical support during the pilot project. Training and Education Phase (5 months) • Provide training sessions, workshops, and online resources to build the necessary skills and knowledge to run and maintain the infrastructure. Documentation and Knowledge Sharing Phase (2 months) • Develop comprehensive documentation, guidelines, and best practices for the open disease infrastructure implementation and management. • Create a knowledge-sharing platform or internal community to facilitate discussions and exchange of expertise. • Establish a feedback loop to continuously update and improve the documentation based on user input.

Openness: The open disease screener has been developed as an open-source project, making its underlying code freely available to the public. By using an open-source license, the project enables anyone to view, modify, and distribute the infrastructure. This openness encourages contributions from developers worldwide, who can enhance the application, fix bugs, and add new features. The project has established channels for collaboration and communication. This includes hosting public forums and chat channels where people can discuss ideas, ask questions, and contribute to the project. Regular community meetings or webinars can be held to share progress, gather feedback, and foster a sense of ownership and inclusivity among participants. The project adopts an open approach to documentation, providing publicly accessible guides, tutorials, and technical documentation. This empowers developers, researchers, and users to understand the project's architecture, functionality, and usage. Furthermore, project reports, research findings, and evaluation results can be openly shared, promoting transparency and enabling peer review and collaboration.

Challenges: • Building an engaged and active community can be challenging. Encouraging individuals to contribute, provide feedback, and collaborate requires effective community management, regular communication, and creating a welcoming environment. Overcoming barriers such as language barriers, time zone differences, or varying levels of technical expertise can be additional challenges. • Developing an open disease screener involves addressing technical complexities related to software development, system integration, and scalability. • Validating the accuracy, reliability, and effectiveness of the disease screener is crucial. Conducting rigorous testing, validation studies can be time-consuming and resource-intensive. • Encouraging healthcare providers, organizations, and individuals to adopt the open disease screener may face resistance or skepticism. Overcoming barriers such as resistance to change, concerns about accuracy, and limited awareness of the benefits will require robust communication, education, and demonstration of the screener's value and impact. • The Open disease screener initiative must navigate legal and ethical considerations, including data ownership, informed consent, liability, and compliance with local regulations.

Neglectedness: CZI used to fund these kinds of initiative but on Bioimaging. Our infrastructure deals why biomedical data and can’t be fund by the CZI fund.

Success: A combination of quantitative and qualitative methods can provide a comprehensive understanding of the success of the open disease screener. • Assess the level of community engagement, collaboration, and contributions to the screener's development and improvement. Measure the number of external contributors, code contributions, bug reports, or feature requests from the broader community. • Measure the level of openness achieved in the project, such as the number of external contributors, the adoption of open-source practices, and the extent of community engagement and involvement.

Total Budget: 5620

Affiliations: MBOALAB

LMIE Carveout: The project and team are based in Cameroon, it fits within the category of low- and middle-income countries (LMICs). Cameroon is classified as a lower-middle-income country by the World Bank. Therefore, the project would be eligible for consideration under the funding category reserved for LMIEs.

Team Skills: Elisee Jafsia is a Data Scientist as well as a web developer. He has interests in Artificial Intelligence and Machine Learning. He has a strong experience in coding. He is the Project Lead and his research spans AI and Healthcare diagnostics. Jafsia project on typhoid fever has been ranked top 100 projects in AI by UNESCO/IRCAI for the year 2022 (https://ircai.org/top100/entry/improve-diagnostics-of-typhoid-through-open-science-an-artificial-intelligence-based-technique/). Jafsia will serve as project lead Stephane Fadanka is a Molecular Biology researcher with a particular interest in synthetic biotechnology. He has prior experience on improving existing typhoid diagnosis test. He is also the quality assurance leader of the project. Roswilo Alexandre is a computer engineering with extensive experience on developing healthtech solutions. He was part of the Lab team that worked on the OpenFlexure Microscope (OFM) to automate malaria diagnostics and translated the Open connect app into French. He will coordinate technical aspects of the aspect and work closely with the hired staff. Dr Thomas Mboa is deeply engaged in promoting responsible AI in the MboaLab. He supports the scale up and implementation of the infrastructure and will also ensure that all ethical aspects are well respected.

Submission Number: 18