A dataset of video recordings for respiratory rate counting from chest wall movements in children aged below five years

Abstract

Objectives

This dataset was generated as part of a study aimed at developing a video expert panel (VEP) to serve as a reference standard for respiratory rate assessment in children under five years of age. The overarching goal was to improve the accuracy of diagnosing respiratory illnesses and to support the advancement of automated video-based methods for respiratory rate measurement. Children aged 0–59 months presenting with cough and/or difficulty breathing at different levels of healthcare facilities in Bangladesh were enrolled during 2021–2022.

Data description

The dataset comprises 332 video recordings, each approximately 60 s in duration, documenting chest wall movements of participating children. All recordings were independently reviewed by members of a trained VEP, who determined respiratory rate using a standardized multi-reviewer process. Alongside the video files, a Stata (.dta) file provides structured metadata, including demographic information (age, sex), clinical variables (history and observation of cough and difficult breathing), and respiratory rate counts. This dataset is intended to support the development and validation of automated video-based diagnostic tools, training and standardization of health workers, and quality assurance initiatives. Its open-access availability fosters interdisciplinary collaboration and international research in pediatric respiratory diagnostics.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13104-026-07677-x.

Objective

This dataset was compiled as a part of a broader research initiative to establish a video expert panel (VEP) as a reference standard for respiratory rate assessment in children under five years of age [1]. Respiratory rate is an important clinical sign for diagnosing and managing respiratory illnesses, particularly pneumonia, in children [2]. Accurate measurement of respiratory rate is essential for timely clinical decision-making and effective treatment [3].

In low-resource settings, manual respiratory rate counting remains the standard practice [4]. However, this method is highly susceptible to variability, observer bias, and inconsistency in clinical application, often leading to diagnostic errors [5]. These challenges have driven increasing interest in technology-assisted approaches that can improve the reliability and accuracy of respiratory rate assessment [6]. Advanced systems such as structured light plethysmography and electronic optoplethysmography can record thoracic and abdominal movements to provide detailed respiratory analysis [7, 8]. However, their complexity and cost limit routine use in low-resource settings. Developing and validating simplified, affordable tools for use in such settings therefore requires access to high-quality, standardized datasets that can serve as a robust reference.

To address this need, chest wall movements of children under five years of age were recorded using a standardized video recording protocol across different healthcare facilities in Bangladesh [9]. Each recording was systematically reviewed by a trained VEP to determine respiratory rate values using a structured multi-reviewer process [1]. The dataset consists of both video files and accompanying metadata, offering a comprehensive resource for researchers.

The primary purpose of compiling this dataset was to support innovation in pediatric respiratory care. Specifically, it is intended to facilitate the development and validation of automated video-based respiratory rate counting tools, support training and standardization of healthcare providers, and strengthen quality assurance processes for respiratory illness assessment in children. By making these data openly available, we aim to encourage interdisciplinary collaboration and contribute to global efforts to advance childhood respiratory diagnostics.

Data description

The dataset is organized into two primary components within the data repository: Video Files and Stata File [10]. An overview of these components is provided in Table 1. A structured data-collection tool was specifically developed for this study to systematically capture demographic, clinical, and video-related information (Supplementary File S1).

Table 1.

Overview of data files/data sets

Label	Name of data file/data set	File types (file extension)	Data repository and identifier (DOI or accession number)
Data set 1	Video Files	Video files (.mp4)	Mendeley Data 10.17632/72dd3rkttf.2 [10]
Data file 1	Stata File	Stata file (.dta)	Mendeley Data 10.17632/72dd3rkttf.2 [10]

Video files

This component contains 332 video recordings of chest wall movements from children aged 0–59 months. The recordings were collected between 2021 and 2022 across three types of health facilities in Bangladesh: the inpatient department of the Institute of Child and Mother Health (ICMH) in Dhaka, the Zakiganj sub-district hospital, and three community clinics in Zakiganj, Sylhet.

Each video file is stored in a subfolder structure by health facility and unique child identifier. The recordings were acquired using a Canon EOS M50 digital camera. Where natural light was insufficient, an LED lamp was used. To ensure consistency, children’s clothing was removed from the neck to the umbilicus to expose the chest and abdomen, and videos were recorded only when the child was calm and breathing naturally. To protect privacy, facial features and other identifiable characteristics were intentionally excluded. Each recording has a duration of approximately 60 s and is stored in .mp4 format.

The raw video files were processed before inclusion in the dataset. Processing included trimming to standardized duration, removing sound and any identifiable elements, and reducing file size. Video editing was carried out using Adobe Premiere Pro. The final processed videos capture chest and abdominal movement only, with filename identifiers visible within each recording.

Stata file

The accompanying Stata (.dta) file contains structured metadata corresponding to each of the 332 videos. The variables include are:

• Child identifier.

• Video file name.

• Duration of video file.

• Participating health facility.

• Demographic information (age in months, sex).

• Clinical data (history or observation of cough, history or observation of difficult breathing).

• Respiratory rate (breaths per minute), as determined by a video expert panel.

Respiratory rate was derived through a standardized multi-reviewer process. Each video was independently assessed by two VEP members. When the difference between their counts exceeded ± 2 breaths per minute, the video was reviewed by additional panel members, and consensus was reached using predefined rules. Videos without agreement were excluded.

Data summary

The dataset comprises records from 332 children aged 0–59 months, with 137 enrolled at ICMH, 92 at sub-district hospital, and 103 at community clinics. The median age was 10 months (interquartile range: 3.0–30.5), with 51.8% (172/332) younger than 12 months. Males accounted for 58.7% (195/332) of the sample. Nearly all children (97.3%, 323/332) had a reported history of cough, and 73.5% (244/332) had an observed caugh. A history of difficult breathing was reported in 47.% (157/332), while 20.2% (67/332) had observed difficult breathing. Respiratory rate distribution was as follows: 58.7% (195/332) had < 40 breaths per minute, 18.7% (62/332) had 40–49, 11.5% (38/332) had 50–59, and 11.1% (37/332) had ≥ 60.

Limitations

• The dataset was collected from a limited number of healthcare facilities in Bangladesh (one tertiary hospital, one sub-district hospital, and three community clinics), which may restrict the generalizability of findings both within the country and to other international settings.

• The sample size (332 children) is sufficient for descriptive analysis but may be inadequate for subgroup analyses or more complex statistical modeling.

• Video recording conditions varied across facilities (e.g., lighting, background environment, and child positioning), which could have influenced video quality and interpretability.

• The videos were captured primarily for respiratory rate assessment; therefore, other clinical features such as chest indrawing may not be reliably evaluated.

• The dataset includes only a limited set of clinical variables (cough, difficult breathing, and respiratory rate); other relevant respiratory signs and symptoms were not collected.

• Clinical variables such as cough and difficult breathing were based partly on caregiver reports and health worker observations, which may be subject to reporting bias.

• A proportion of videos were excluded when VEP members could not reach consensus on respiratory rate measurement, resulting in some data loss.

• The study focused exclusively on children presenting with respiratory symptoms; data were not collected from participants without respiratory diseases, which limits comparative interpretation.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

VEP

Video expert panel

ICMH

Institute of child and mother health

Author contributions

AMK conceived the study with guidance from HC, EDM, SC, AHB, and TS, who supervised the overall conception and design. HC also secured funding for the study. AMK drafted the manuscript. RUR conducted video recording and editing. RUR, RB and MSU contributed to video recording. FTJ, MFA, and SA supported project management. MSI carried out data analysis, and NHC managed the data. HC, EDM, SC, AHB, and TS critically reviewed the manuscript. All authors read and approved the final version of the manuscript.

Funding

This research was funded by the UK National Institute for Health Research (NIHR) through the Global Health Research Unit on Respiratory Health (RESPIRE) [16/136/109] using UK aid from the UK Government to support global health research. The views expressed in this publication are those of the authors and not necessarily those of the NIHR or the UK Government.

Data availability

The data described in this Data note can be freely and openly accessed on Mendeley Data under https://doi.org/10.17632/72dd3rkttf.2. Please see Table 1 and references [10] for details and links to the data.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki, and approved by the National Research Ethics Committee of Bangladesh Medical Research Council, Bangladesh (Registration Number: 39315022021), and Edinburgh Medical School Research Ethics Committee (EMREC), Edinburgh, UK (REC Reference: 21-EMREC-040). Informed written consent was taken from the parent or guardian of each child.

Consent for publication

Written informed consent for publication of potentially identifying images and other personal and clinical details was obtained from the parents or legal guardians of all participating children.

Competing interests

The authors declare no competing interests.