AI-Driven Early Detection and Treatment of Preeclampsia Using Low-Dose Aspirin
AspirishaAI is a community-ready digital platform that identifies pregnant women at risk of preeclampsia early, recommends timely low-dose aspirin (LDA), supports referral pathways, and empowers community health promoters and clinicians with interpretable AI decision support.
Why AspirishaAI Matters
Preeclampsia remains a major, preventable contributor to maternal and perinatal morbidity and mortality in Kenya and other low-resource settings. AspirishaAI responds to this challenge by combining local data, machine learning, community health systems, and low-dose aspirin pathways into one scalable maternal health platform.
16%
Estimated share of maternal deaths in sub-Saharan Africa linked to preeclampsia and other hypertensive disorders of pregnancy.
6%
Estimated prevalence of preeclampsia among pregnancies in Africa, contributing significantly to maternal and neonatal complications.
2M
Estimated community health workers across Africa’s health systems who can support screening, maternal education, referral, and treatment adherence.
70%
Estimated maternal deaths from hypertensive disorders of pregnancy that are preventable with early detection and timely treatment
Preventive care is not reaching women early enough
Low-dose aspirin is a simple and relatively affordable intervention that can prevent preeclampsia in women at risk, yet uptake remains low where maternal deaths are highest. Screening is often late, ANC attendance is irregular, and vulnerable women in rural and informal settlements are least likely to receive timely preventive care.
01 Late or missed identification of women at risk
Many women at high risk of preeclampsia are not identified early enough for timely intervention, especially outside well-equipped facilities.
02 Irregular antenatal attendance among vulnerable populations
Low-income, rural, marginalized, and informal-settlement populations often face interrupted or delayed ANC attendance.
03 Weak integration of LDA into community-based care
Low-dose aspirin prevention pathways are not yet fully embedded within frontline maternal health and CHP workflows.
04 Limited predictive analytics and policy evidence
Health systems still lack scalable AI-supported screening tools and strong implementation evidence for national LDA scale-up.
Why Early Detection Matters
Preeclampsia is preventable, but delayed identification can lead to severe maternal complications, fetal growth restriction, miscarriage, preterm birth, stillbirth, and avoidable neonatal deaths. A community-deployable AI model linked to LDA can move care upstream before disease progression becomes severe.
The Solution
AspirishaAI is a machine learning-enabled mobile and EMR-integrated platform that predicts preeclampsia risk, recommends timely low-dose aspirin initiation, supports community health promoter action, and strengthens referral pathways between households, primary care, and hospitals.
Predictive AI Model
Uses clinical, demographic, obstetric, laboratory, physiological, and lifestyle data to identify women at risk before severe symptoms emerge.
Mobile App for CHPs
Enables field-level screening, decision support, patient education, referral initiation, and adherence follow-up in low-resource settings.
EMR Integration
Connects risk scores, alerts, LDA plans, ANC records, and follow-up notes into health facility workflows and dashboards.
Dashboard Analytics
Tracks uptake, missed doses, aspirin stock levels, referral trends, and population-level maternal health insights for planning and scale-up.
How the AspirishaAI Pipeline Works
The platform is designed as a practical end-to-end workflow from community screening to treatment monitoring, with interpretable AI and operational integration at every step.
Capture Risk Parameters
CHPs or clinicians collect demographic, obstetric, medical, clinical, laboratory, and psychosocial variables through the mobile app or facility interface.
Age, parity, gravidity, ANC access, preeclampsia history, chronic conditions, blood pressure, BMI, proteinuria, biomarkers, and CHP observations.
AI Risk Stratification
The model processes multimodal inputs and provides interpretable risk categories and recommendations suitable for both facility teams and community health workers.
Clinical Decision Support
Once risk is generated, the system can automatically flag LDA eligibility, issue alerts, recommend dosage and timing, generate treatment orders, and populate a monitoring register.
Community Health Promoter Actions
CHPs educate the pregnant woman, support LDA initiation, record start and end dates, set reminders, track adherence, report side effects, and initiate referral when needed.
Monitoring, Dashboard, and Stock Management
The platform tracks treatment uptake, missed follow-ups, side effects, blood pressure trends, and aspirin stock levels while feeding analytics into the NextGenHIMS dashboard.
Core Platform Features
AspirishaAI is designed for both research and real-world deployment, balancing predictive accuracy, explainability, community usability, and health system integration.
Locally Relevant AI
Built using locally sourced Kenyan data, including datasets on determinants of preeclampsia and eclampsia, to improve contextual accuracy and generalizability.
Explainable Predictions
Uses explainable AI methods so providers and CHPs can understand which risk factors are driving the prediction and trust the recommendation.
Community-Based Delivery
Extends early detection and treatment initiation beyond facilities into households and underserved communities through the CHP network.
Referral Pathway Support
Links community screening to timely referral and facility follow-up, strengthening continuity of care from first contact to treatment monitoring.
EMR + App Ecosystem
Supports integrated workflows between the mobile app and electronic medical records, reducing duplication and improving data continuity.
Population-Level Insights
Provides dashboards for LDA coverage, risk distribution, missed doses, stock levels, and program performance for counties and national stakeholders.
Path to Impact
AspirishaAI is designed not only as a digital tool, but as a maternal health systems intervention that can influence policy, financing, service delivery, and long-term community health outcomes.
Reduce preventable maternal deaths
Improve early detection and preventive treatment among women at risk of preeclampsia.
Strengthen community health systems
Equip CHPs to deliver screening, education, LDA follow-up, and timely referral.
Generate policy evidence
Provide cost-effectiveness, feasibility, and implementation evidence for national scale-up.
Support sustainable scale
Enable financing models, stakeholder mobilization, and broader integration into maternal care programs.
From community screening to national policy impact
AspirishaAI brings together AI, low-dose aspirin, community health promoters, and digital health systems to create a scalable model for preventing preeclampsia-related maternal and neonatal harm in low-resource settings.
- Predictive model for early warning
- LDA initiation and adherence support
- CHP training and referral workflows
- EMR integration and health system dashboards
- Economic evaluation for scale-up
Contact
Connect with the AspirishaAI team to explore partnership, deployment, research collaboration, or app integration opportunities.
Deployment Context
Kenya and similar low-resource settings
Phone
+254722 779 456
info@aspirishaai.org | okungu008@gmail.com