Page 64 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

Use Case 4: Smartphone-over-Microscope Diagnosis: Empowering

Community Health Workers with TinyML for Malaria Parasite
Detection in Rural Areas

Country: Ghana

Organization: CSIR - Institute for Scientific and Technological Information

Bolgatanga Technical University

Contact Person(s):

Dennis Agyemanh Nana Gookyi dennisgookyi@ gmail .com
Fortunatus Aabangbio Wulnye fortunatuswulnye@ outlook .com
Michael Wilson yboabengwilson@ gmail .com
Roger Kwao Ahiadormey rogerkwao@ gmail .com
Moses Abambire abambiremoses@ gmail .com
Paul Danquah pauldanquah@ yahoo .com
Raymond Gyaang gyaangraymond@ outlook .com

1 Use Case Summary Table

Item Details
Category Healthcare

Problem The primary issue addressed is the lack of reliable and timely malaria diag-
Addressed nosis in rural communities, where access to skilled laboratory technicians and
diagnostic facilities is limited. Traditional microscopy requires trained profes-
sionals and laboratory infrastructure, while Rapid Diagnostic Tests (RDTs) can
be less accurate and prone to false negatives. This gap in diagnostic capability
hampers effective malaria treatment and control efforts in underserved areas.

Key Aspects of • Smartphone-Mounted Microscopy: Use of a smartphone positioned over
Solution a microscope eyepiece to capture high-resolution images of blood slides
for analysis
• TinyML-Powered Real-Time Image Analysis: Deployment of lightweight AI
models on the smartphone to detect malaria parasites (Trochozoites) and
White Blood Cells (WBCs) directly from captured images
• Training with Labeled Dataset: Use of the Lacuna Malaria Detection Chal-
lenge dataset to train and calibrate the AI model for accurate classification
and detection
• Portable and Accessible Diagnostics: Providing community health workers
with an easy-to-use, portable tool that enhances malaria diagnosis without
the need for full laboratory setups

28