01
Make information easier to explore
Study how clinical indicators might be organised into a clear, transparent research workflow.
Responsible AI · Public health
Care begins with understanding.
UpLift21 is an early-stage research concept exploring how machine learning and selected clinical indicators might support more informed conversations about prenatal screening.
Research concept only. Not a diagnostic product, clinically validated screening service, or medical advice.
UPLIFT21 · RESEARCH CONCEPT
Evidence, access, and people at the centre.
An exploration of how data and responsible AI might support care—not replace it.
AccessTransparencyHuman-led care
What guides the work
EquityTransparencyPrivacyClinical oversight
01 / Mission
Explore the possibility. Respect the complexity.
Prenatal screening involves complex information and important personal decisions. UpLift21 explores whether machine learning could help organise selected clinical indicators in a way that supports—not replaces—qualified healthcare professionals.
02
Keep access in view
Consider accessibility and resource constraints when exploring possible future digital health tools.
03
Keep people at the centre
Protect informed choice, clinical judgement, counselling, and appropriate follow-up care.
02 / Interactive demo
A demo that explains limits—not risk.
Explore how a responsible interface might communicate data completeness, privacy, and possible next steps. This demonstration does not use patient data or make predictions.
Illustrative workflowNon-clinical
This is an interface prototype only. It calculates no risk score, estimates no condition, and provides no clinical recommendation.
03 / Research
An early-stage machine-learning concept.
The project has explored a gradient-boosting ensemble using seven clinical biomarkers. This description does not establish clinical validity, safety, generalisability, or readiness for deployment.
No clinical performance figures or patient outcomes are claimed here.
Evidence before implementation.
Any potential medical use would require robust evaluation, independent review, and appropriate clinical and regulatory oversight.
01
Define the intended use
Be specific about the research question, setting, and limits.
02
Validate independently
Evaluate with appropriate datasets and transparent methods.
03
Assess equity and uncertainty
Study data gaps, bias, and performance across relevant populations.
04
Protect people and data
Use strong privacy safeguards and suitable clinical oversight.
04 / Team
A public-health question, explored through technology.
MD
Mihir Deshpande
Project contributor · AI and research
Exploring how applied machine learning can be investigated responsibly in healthcare, with attention to accessibility, evidence, and limitations.
Confirm current project roles and obtain permission before publishing additional names, affiliations, or collaborator claims.
Questions or collaboration?
For project enquiries, reach out by email.