What we know. What still needs testing.

The current product proposition

The planned workflow uses guided front, side and back smartphone pictures to estimate height. Height may be considered with age and sex for stunting screening, paired with Bluetooth-connected weight for one wasting-screening route, or used alongside a separate image-based MUAC offering.

The website’s human images are AI-generated illustrations. Any sample measurement shown in an image is illustrative, not a result from a validated clinical study.

Claims this website does not make

GrowthSight has not published a field validation study, a quantified accuracy or sensitivity result, or evidence of improved health outcomes here. The product is not presented as ready for independent field deployment or as a medically certified device. Screening does not replace a full assessment.

What a useful pilot should establish

A pilot should compare image-derived measurements with appropriate reference methods and document performance across the children and conditions in which the product would be used. It should measure unusable captures, repeatability, staff time, training needs and how workers handle uncertain or concerning results.

A programme should also define consent, image and child-data handling, review responsibilities, escalation and referral pathways before field use. Results should be reported by measurement and setting rather than compressed into a single unqualified accuracy claim.

How to read the rest of this site

Descriptions such as “aims to”, “developing” and “pilot capability” refer to intended features or evaluation work. They are not evidence that a capability is clinically validated. We will update this page when evidence suitable for public review is available.