Handheld Skin Scanner Could Reveal a Baby’s True Gestational Age in Seconds

In the delivery rooms of Manaus, in remote riverside communities deep in the Brazilian Amazon, and in the teaching hospitals of Tegucigalpa, a quiet technological experiment is about to unfold. Researchers have unveiled the full protocol for the APPLE study, a multicenter prospective cohort study designed to validate a portable optoelectronic device called PreemieTest, which estimates a newborn’s gestational age within the first 24 to 48 hours of life by simply reading the optical properties of the infant’s skin. The study, published in BMC Pediatrics, addresses one of the most stubborn gaps in global neonatal medicine: in much of the world, nobody actually knows how premature a baby is, and that uncertainty can be lethal.

The biological premise behind the device is elegant. Fetal skin does not mature in a linear fashion; instead, it follows a sigmoid trajectory, meaning that its structure and composition change dramatically and predictably across the course of gestation. As the fetus approaches term, the skin’s maturation alters the way it reflects light. A noninvasive optical sensor can detect these shifts in reflectance and translate them into an estimate of gestational age. Because the measurement is passive, painless, and requires no blood draw, no imaging equipment, and no specialized operator, it is precisely the kind of technology that could function in a birth center with no physician on site, or in a boat-accessed village where the nearest ultrasound machine is hundreds of kilometers away.

Why does gestational age matter so much? The answer, according to the research team led by Roberta Lins Gonçalves of the Federal University of Amazonas, lies largely in the respiratory system. Gestational age is the single most important factor for stratifying a newborn’s early respiratory risk. Babies born significantly preterm are at elevated risk of respiratory distress syndrome, a condition in which underdeveloped lungs lack sufficient surfactant and struggle to keep the air sacs inflated. Knowing whether a newborn is 32 weeks or 37 weeks along changes everything about how clinicians triage that infant: whether to monitor closely, whether to transfer to a neonatal intensive care unit, whether to prepare ventilatory support. Yet in resource-poor settings, the gold standard for dating a pregnancy, a first-trimester obstetric ultrasound, is often unavailable or unreliable, and the date of the last menstrual period is frequently uncertain or misremembered.

The APPLE study, whose name derives from the Portuguese phrase Acompanhamento de Bebês pelo PreemieTest, meaning follow-up of babies using the PreemieTest, will enroll newborns across a strikingly diverse network of birth settings in Brazil and Honduras. Participating sites include the Getúlio Vargas University Hospital in Manaus, the Sofia Feldman Hospital in Belo Horizonte, the Maria Aparecida Pedrossian University Hospital and the Florescer Humanized Birth Center in Mato Grosso do Sul, the Januário Cicco Maternity School and Ana Bezerra University Hospital in Rio Grande do Norte, and, crucially, the Vó Mundoca Hospital in Borba and the Edith Mendes Weckner Hospital in Novo Aripuanã, both riverside and remote municipalities in the Amazon. In Honduras, the study spans Hospital Escuela, the Instituto Hondureño de Seguridad Social, and Hospital General San Felipe in Tegucigalpa. This deliberate spread across tertiary hospitals, maternity schools, birth centers, and remote communities is the point: the researchers want to know not whether the device works in an idealized laboratory, but whether it works everywhere a baby might actually be born.

Methodologically, the validation is rigorous and multipronged. The gestational age estimated by PreemieTest will be compared against three traditional reference approaches: first-trimester ultrasound when available, which the researchers treat as the gold standard; the date of the last menstrual period; and a postnatal clinical score, a standardized assessment based on physical and neurological signs of maturity. Agreement between the device and these references will be quantified using correlation coefficients, Bland–Altman plots, which reveal whether a device systematically overestimates or underestimates across the measurement range, and concordance analyses. The team will also report the mean absolute error, a clinically intuitive metric that expresses, in weeks, how far off the device typically is. Performance will be examined across different skin phototypes, a critical consideration in a study population spanning the wide range of skin tones found in Brazil and Honduras, where pigmentation could plausibly influence optical readings if the algorithm were not properly calibrated.

What sets this study apart from a typical device-validation exercise is its longitudinal ambition. The researchers will follow the infants and model respiratory outcomes, including hospitalizations, clinical diagnoses, and the use of ventilatory support, at birth, three months, and six months of age. Using multivariable logistic regression adjusted for prespecified confounders such as birth weight, sex, prematurity, mode of delivery, and ultrasound-based gestational age, and accounting for clustering by study location, they aim to determine whether the device’s gestational age estimates actually predict clinically meaningful outcomes. This is the difference between a device that produces a number and a device that changes care. If optical gestational age assessment stratifies respiratory risk as well as ultrasound dating does, the case for deployment in low-resource settings becomes compelling.

The protocol also takes usability seriously, which is often the Achilles heel of medical devices intended for frontline use. Midwives, who in many of the participating settings are the primary birth attendants, will rate the device using the System Usability Scale, a validated instrument that yields a standardized score of perceived ease of use. A device that is accurate but cumbersome will simply not be adopted in a busy maternity ward or a community health post. By measuring usability alongside accuracy, the APPLE study is effectively testing the entire implementation pathway, not just the sensor. The study is registered in the Brazilian Clinical Trials Registry under RBR-10ch623r, was approved by ethics committees at all participating institutions in both countries, and will be conducted in accordance with the Declaration of Helsinki, with written informed consent obtained from parents or legal guardians.

Data management reflects a similarly careful design. All information will be captured in Research Electronic Data Capture, the widely used REDCap platform, with standardized quality checks and continuous monitoring for completeness. Missing data, an inevitable feature of any study spanning the Amazon to urban Honduras, will be handled through multiple imputation and sensitivity analyses, so that the conclusions are robust to realistic imperfections in follow-up. The study is funded by Birthtech Dispositivos para a Saúde, the device’s manufacturer, together with Grand Challenges Canada, and supported by Brazilian federal agencies including CAPES, CNPq, and the Amazonas Research Foundation, a funding mix that pairs commercial development with public health philanthropy.

The stakes are enormous. The World Health Organization has long identified prematurity as the leading cause of death in children under five, and the first step in managing a preterm infant is knowing that the infant is preterm. In settings where early ultrasound is a luxury, clinicians currently rely on last menstrual period recall and postnatal scoring, both of which carry substantial error. A low-cost, handheld optical device that delivers a reliable gestational age estimate in the first day or two of life, across skin tones and across levels of care, would give midwives and physicians in the most remote corners of the world the same dating information that wealthy health systems take for granted. The APPLE study will not deliver that answer until its results are in, but its protocol, spanning the Amazon’s river communities to Tegucigalpa’s teaching hospitals, shows what honest, real-world validation of a global health technology looks like: diverse sites, hard reference standards, longitudinal outcomes, and a willingness to test whether a promising gadget survives contact with the places that need it most.

Subject of Research: Post-market validation of a portable optical device for newborn gestational age assessment

Article Title: APPLE study protocol: post-market validation of the PreemieTest® device for gestational age assessment in diverse birth settings

Article References: Gonçalves, R. L., da Costa Neto, S. S., de Souza, E. K. S., da Silva, A. V., Soares-Marangoni, D., Zambrano, L. I., Neves, G. S., Miralha, A. L., Lopes, T. C. P., de Andrade Vieira, J. A. R., da Silva, A. V., Costa, A. J. T., Lima, E. C. C., Bessa, N. B., Bahia, B. L., de Souza Rocha, P. G., de Sousa Júnior, F. S., Lima, J. C. B., Lima, H. L. O., … Pereira, S. A. (2026). APPLE study protocol: post-market validation of the PreemieTest® device for gestational age assessment in diverse birth settings. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07640-6

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07640-6

Keywords: gestational age, PreemieTest, prematurity, neonatal care, skin reflectance, optical device, Brazil, Honduras, Amazon, respiratory distress, global health, study protocol