Digitize handwritten growth plots
Use document scanning, image processing, and custom object detection to extract age and measurement points from the Kenya MOH chart format.
About NeoScan
A research project connecting paper records with digital care.
NeoScan is the Newborn Growth Chart Digitizer—an academic mobile and cloud health information system project from JKUAT and JHUB AFRICA.
Mission
Make paper growth histories easier to preserve, understand, and share—without losing human review.
Why this problem matters.
The Kenya Road-to-Health booklet is the primary growth monitoring record for children from birth to five years. It is useful, familiar, and portable—but also vulnerable to loss, damage, inconsistent plotting, and limited remote access.
NeoScan explores a practical bridge: photograph the chart already in use, convert its plotted points into structured measurements, compare them with WHO growth standards, and let a person verify the result before it becomes part of the digital history.
The project is designed for parents, guardians, nurses, community health workers, pediatricians, researchers, and institutions working in Kenya’s maternal and child health context.
Project goals
Focused on a clear, testable gap.
Apply WHO references
Calculate day-precise, sex-specific growth positions using WHO LMS reference data.
Support careful review
Show the extracted record to a person, log corrections, and flag defined patterns for qualified review rather than diagnosis.
Work through weak connectivity
Protect scans locally while offline and synchronize them when network access returns.
Give parents sharing control
Require explicit, revocable consent before a registered clinic receives a child’s data.
Produce research evidence
Deliver an accuracy report, clinical validation study, parent manual, and clinic API guide.
Institutional context
JKUAT · JHUB AFRICA
The supplied SRS identifies JHUB AFRICA as the institution and Jomo Kenyatta University of Agriculture and Technology (JKUAT) as the university context for this June 2026 academic project.
This site does not add unlisted partnerships, credentials, deployments, or endorsements.
Prototype boundary
Full clinical deployment, national HMIS integration, real-time video scanning, and telemedicine video calls are outside the stated version scope.
Project team
A multidisciplinary student team.
Roles and deliverables below are reproduced from the supplied SRS without adding biographies or credentials.
Michael MburuBackend & ML / Team LeadSRS, SADD, ML pipeline, API
Josphat ThumiFrontend DeveloperReact Native application, UI/UX
Livia NafulaStorytelling & Technical WritingClinical Validation Study, User Manual
Sharlene MutheuResearch, Communications & Project ManagementPMP, KDPA work, data collection coordination
Stella WamboiData AnalystDataset validation, performance tracking, reporting
Guiding principles
Useful technology needs restraint.
- Do not replace clinical judgment
Present review flags as prompts for qualified assessment.
- Keep verification visible
Let users inspect and correct extracted measurements.
- Design for context
Support limited connectivity, older devices, and simple parent-facing language.
- Protect sensitive data
Build around consent, role boundaries, encryption, and audit history.
- Report limitations honestly
Separate requirements and research targets from validated outcomes.
Learn more