Accuracy of a handheld 3D imaging system for child anthropometric measurements in population-based household surveys and surveillance platforms: an effectiveness validation study in Guatemala, Kenya, and China

Exploring the Use of a Length AI Algorithm to Estimate Children's Length from Smartphone Images in a Real-World Setting: Algorithm Development and Usability Study

BACKGROUND

Length measurement in young children younger than 18 months is important for monitoring growth and development. Accurate length measurement requires proper equipment, standardized methods, and trained personnel. In addition, length measurement requires young children's cooperation, making it particularly challenging during infancy and toddlerhood.

OBJECTIVE

This study aimed to develop a length artificial intelligence (LAI) algorithm to aid users in determining recumbent length conveniently from smartphone images and explore its performance and suitability for personal and clinical use.

METHODS

This proof-of-concept study in healthy children (aged 0-18 months) was performed at KK Women's and Children's Hospital, Singapore, from November 2021 to March 2022. Smartphone images were taken by parents and investigators. Standardized length-board measurements were taken by trained investigators. Performance was evaluated by comparing the tool's image-based length estimations with length-board measurements (bias [mean error, mean difference between measured and predicted length]; absolute error [magnitude of error]). Prediction performance was evaluated on an individual-image basis and participant-averaged basis. User experience was collected through questionnaires.

RESULTS

A total of 215 participants (median age 4.4, IQR 1.9-9.7 months) were included. The tool produced a length prediction for 99.4% (2211/2224) of photos analyzed. The mean absolute error was 2.47 cm for individual image predictions and 1.77 cm for participant-averaged predictions. Investigators and parents reported no difficulties in capturing the required photos for most participants (182/215, 84.7% participants and 144/200, 72% participants, respectively).

CONCLUSIONS

The LAI algorithm is an accessible and novel way of estimating children's length from smartphone images without the need for specialized equipment or trained personnel. The LAI algorithm's current performance and ease of use suggest its potential for use by parents or caregivers with an accuracy approaching what is typically achieved in general clinics or community health settings. The results show that the algorithm is acceptable for use in a personal setting, serving as a proof of concept for use in clinical settings.

TRIAL REGISTRATION

ClinicalTrials.gov NCT05079776; https://clinicaltrials.gov/ct2/show/NCT05079776.

Reliability of anthropometric measurements of a digi-board in comparison to an analog height board in Namibian children under 5 years

Poor measurement quality has set back the utility of anthropometry in defining childhood malnutrition, prompting calls for alternative measurement techniques. This study aimed to assess the reliability of anthropometric measurements using a digital height board in comparison to an analog height board in Namibian children under 5 years of age. A cross-sectional, descriptive study was conducted (n = 425) between the age of 6 and 59 months, using anthropometric measurements of weight, height and mid-upper arm circumference. Two trained enumerators each collected four height measurements of each child: two using an analog height board and two using a digi-board. The repeated height measurements between and within the enumerators were used to determine intra- and interobserver reliability. Reliability of the digi-board was assessed using the technical error of measurement (TEM), relative TEM (%TEM), intraclass correlation and a Bland-Altman analysis to assess the agreement between the two methods. In all these assessments, the analog height board was considered as the gold standard and used for comparison. The digi-board showed superiority to the analog height board in terms of reliability (analog TEM = 0.22, digi-board TEM = 0.16). Although the digi-board has potential to improve child anthropometry, further clinical and large survey studies are needed to validate the used of this tool in routine population-based surveys.

Impact of anthropometry training and feasibility of 3D imaging on anthropometry data quality among children under five years in a postmortem setting

BACKGROUND

The Child Health and Mortality Prevention Surveillance Network (CHAMPS) identifies causes of under-5 mortality in high mortality countries.

OBJECTIVE

To address challenges in postmortem nutritional assessment, we evaluated the impact of anthropometry training and the feasibility of 3D imaging on data quality within the CHAMPS Kenya site.

DESIGN

Staff were trained using World Health Organization (WHO)-recommended manual anthropometry equipment and novel 3D imaging methods to collect postmortem measurements. Following training, 76 deceased children were measured in duplicate and were compared to measurements of 75 pre-training deceased children. Outcomes included measures of data quality (standard deviations of anthropometric indices and digit preference scores (DPS)), precision (absolute and relative technical errors of measurement, TEMs or rTEMs), and accuracy (Bland-Altman plots). WHO growth standards were used to produce anthropometric indices. Post-training surveys and in-depth interviews collected qualitative feedback on measurer experience with performing manual anthropometry and ease of using 3D imaging software.

RESULTS

Manual anthropometry data quality improved after training, as indicated by DPS. Standard deviations of anthropometric indices exceeded limits for high data quality when using the WHO growth standards. Reliability of measurements post-training was high as indicated by rTEMs below 1.5%. 3D imaging was highly correlated with manual measurements; however, on average 3D scans overestimated length and head circumference by 1.61 cm and 2.27 cm, respectively. Site staff preferred manual anthropometry to 3D imaging, as the imaging technology required adequate lighting and additional considerations when performing the measurements.

CONCLUSIONS

Manual anthropometry was feasible and reliable postmortem in the presence of rigor mortis. 3D imaging may be an accurate alternative to manual anthropometry, but technology adjustments are needed to ensure accuracy and usability.

Portable digital devices for paediatric height and length measurement: A scoping review and target product profile matching analysis

BACKGROUND

Routine anthropometry of children, including length/height measurement, is an essential component of paediatric clinical assessments. UNICEF has called for the accelerated development of novel, digital height/length measurement devices to improve child nutrition and growth surveillance programs. This scoping review aimed to identify all digital, portable height/length measurement devices in the literature or otherwise available internationally. We also assessed identified devices against the UNICEF Target Product Profile (TPP) to identify those of highest potential for clinical and public health use.

METHOD

We searched four databases (Medline, Embase, CINAHL and Global Health) and the grey literature between 1st January 1992 and 2nd February 2023. We looked for studies or reports on portable, digital devices for height or length measurement in children up to 18 years old. Citations were screened independently by two reviewers, with data extraction and quality assessment performed in duplicate and disagreements resolved. Devices were evaluated and scored against the 34 criteria of the UNICEF TPP.

RESULTS

Twenty studies describing twelve height/length measurement devices were identified, most of which used prospective validation designs. Additional devices were found in the grey literature, but these did not report key performance data so were not included. Across the twelve devices, only 10 of 34 UNICEF criteria on average could be fully assessed. Six met UNICEF's ideal accuracy standard and one device met the minimum accuracy standard. The Leica DistoD2 device scored highest (41%), followed by Autoanthro in a controlled environment (33%) and GLM30 (32%). These devices may be high potential for further assessment and development, though further research is required.

CONCLUSION

While 12 portable, digital devices exist for child height/length measurement, insufficient data are available to fully assess whether they meet the industry's needs. Although some devices show promise, further research is needed to test the validity of these devices in varying contexts, and continued development and commercialization will be important to improve reliability and precision of these devices for widespread use.

Acceptability and Experiences with the Use of 3D Scans to Measure Anthropometry of Young Children in Surveys and Surveillance Systems from the Perspective of Field Teams and Caregivers

Background

Portable systems using three-dimensional (3D) scan data to calculate young child anthropometry measurements in population-based surveys and surveillance systems lack acceptability data from field workers and caregivers.

Objective

The aim was to assess acceptability and experiences with 3D scans measuring child aged 0-59 mo anthropometry in population-based surveys and surveillance systems in Guatemala, Kenya, and China (0-23 mo only) among field teams and caregivers of young children as secondary objectives of an external effectiveness evaluation.

Methods

Manual data were collected twice and 12 images captured per child by anthropometrist/expert and assistant (AEA) field teams (individuals/country, n = 15/Guatemala, n = 8/Kenya, n = 6/China). Caregivers were interviewed after observing their child's manual and scan data collection. Mixed methods included an administered caregiver interview (Guatemala, n = 465; Kenya, n = 496; China, n = 297) and self-administered AEA questionnaire both with closed- and open-ended questions, and 6 field team focus group discussions (FGDs; Guatemala, n = 2; Kenya, n = 3; China, n = 1). Qualitative data were coded by 2 authors and quantitative data produced descriptive statistics. Mixed-method results were compared and triangulated.

Results

Most AEAs were female with secondary or higher education. Approximately 80-90% of caregivers were the child's mother. To collect all anthropometry data, 62.1% of the 29 AEAs preferred scan, while 31% preferred manual methods. In FGDs, a key barrier for manual and scan methods was lack of child cooperation. Across countries, approximately 30% to almost 50% of caregivers said their child was bothered by each manual and scan method, while ≥95% of caregivers were willing to have their child measured by scans in the future.

Conclusions

Use of 3D scans to calculate anthropometry measurements was generally at least as acceptable as manual anthropometry measurement among AEA field workers and caregivers of young children aged <60 mo, and in some cases preferred.