Handheld 3D scanner delivers poor anthropometric measurements in children

Use of the AutoAnthro scanning technology, a handheld 3D imaging system, in children aged <60 months has yielded many poor results for anthropometric data quality, accuracy, and precision compared to expert manual measurements of length or height, midupper arm circumference (MUAC), and head circumference (HC), especially for those aged <24 months, reports a recent study.

“Understanding the factors that cause the many poor scan results and how to correct them will be needed prior to using this instrument in routine, population-based survey and surveillance systems,” said the researchers, led by Karim Bougma from the Centers for Disease Control and Prevention Foundation in Atlanta, Georgia, USA.

Bougma and colleagues assessed the accuracy of the AutoAnthro system to measure the anthropometry of children aged 0‒59 months using population-based surveys and surveillance systems in households in Guatemala and Kenya and in hospitals in China.

Field team anthropometrists and their assistants collected manual and scan anthropometric measurements, including length or height, MUAC, and HC (China only) from each child. Then, an anthropometry expert and assistant collected both manual and scan measurements on the same child. The expert manual measurements, rather than field team scans, were considered the standard.

For inter-rater accuracy, the respective average biases in Guatemala, Kenya, and China were ‒0.3, ‒1.9, and ‒6.2 cm for length or height; 0.9, 1.2, and ‒0.8 cm for MUAC; and 2.4 cm for HC in China only. The corresponding intertechnical errors of measurement (inter-TEMs) were 2.8, 3.4, and 5.5 cm for length or height; 1.1, 1.5, and 1.0 cm for MUAC; and 2.8 cm for HC. [Am J Clin Nutr 2022;116:97-110]

For intrarater precision, the absolute mean difference and intra-TEM were 0.1 cm for all manual measurements for all countries. For scans, the absolute mean differences were 0.4‒0.6 cm for length or height, 0.1‒0.1 cm for MUAC, and 0.4 cm for HC. For intra-TEM, length or height was 0.5 in China and Guatemala and 0.7 cm in Kenya; other measurements were ≤0.3 cm.

These results were different from the efficacy evaluation of the AuthAnthro system and other evaluations of 3D technology obtained in controlled settings via different 3D devices, which displayed better accuracy and precision. [Obes Open Access 2016;2(3); Peer J 2017;5:e2980; PLoS One 2018;13:e0205320; Sci Rep 2016;6:26672]

Contextual differences

“The inter-rater, intermethod accuracy results from this effectiveness study were generally poor both overall and for the <24-month and ≥24-month age groups... However, for children ≥24 months, relative inter-TEMs for height were acceptable below the novice threshold,” the researchers said.

“These poor results could be because high accuracy and precision are more challenging in effectiveness field settings due to contextual differences (eg, more anthropometrists, less controlled setting),” they added.

For instance, the field team in Guatemala and Kenya were allowed inside households and moved furniture and closed window dressings to limit lighting and resolve issues affecting scans.

In China, the setting was similar to controlled efficacy settings with dedicated rooms and allowed the team to easily control the lighting, temperature, and equipment placement.

“Overall, this effectiveness study found many poor results and suggests that information on the factors causing the poorer quality data and what steps could be taken to correct these factors is needed before using the AutoAnthro system in population-based survey and surveillance system settings to assess young child anthropometry,” the researchers said.