Page 67 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

for ongoing field data collection, device maintenance, and training programs. Cloud storage
solutions will be essential for securely backing up and managing large datasets.

In terms of collaborations and expansions, there are plans to work with regional healthcare
organizations and NGOs to pilot the system in more rural communities, potentially broadening 4.1-Healthcare
its reach. The AI system could also be expanded to detect other blood-borne diseases, such
as sickle cell anemia or bacterial infections, further improving healthcare in remote areas.
Additionally, partnerships with technology companies or research institutions will be explored
to optimize the TinyML models and investigate more efficient deployment on low-cost devices.

3 Use Case Requirements

• REQ-01: Smartphone with high-resolution camera capability and sufficient processing
power to run TinyML models.
• REQ-02: Optical microscope compatible with smartphone mounting for clear blood slide
imaging.
• REQ-03: Trained TinyML model optimized for real-time malaria parasite (Trochozoite)
and WBC detection.
• REQ-04: Reliable power sources (e.g., battery banks or solar chargers) to support
smartphone use in areas with unstable electricity.
• REQ-05: Secure data storage and management system for saving and labeling new blood
slide images collected during fieldwork.
• REQ-06: Basic training materials and sessions for community health workers on operating
the smartphone-microscope setup and interpreting results.
• REQ-07: Collaboration framework with Bolgatanga Technical University for continuous
dataset expansion and validation.

4 Sequence Diagram

31