Digital image display manipulation and observer variability in pediatric chest radiograph interpretation

Chris Neil James; Dinesh Kumar Narayanasamy; Vasanthan Tanigasalam; Jeyaraj Munusamy; Harish Vijayakandeepan Kumaresan

doi:10.18203/2349-3291.ijcp20263819

Authors

Chris Neil James JIPMER, Karaikal Puducherry, India https://orcid.org/0009-0002-3076-5680
Dinesh Kumar Narayanasamy Department of Pediatrics, JIPMER Karaikal, Puducherry, India
Vasanthan Tanigasalam Department of Pediatrics, JIPMER Karaikal, Puducherry, India https://orcid.org/0000-0003-0629-3644
Jeyaraj Munusamy Department of Pediatrics, JIPMER Karaikal, Puducherry, India https://orcid.org/0000-0001-7305-1456
Harish Vijayakandeepan Kumaresan Department of Respiratory Medicine, Indira Gandhi Medical College & Research Institute, Puducherry, India https://orcid.org/0009-0002-9317-372X

DOI:

https://doi.org/10.18203/2349-3291.ijcp20263819

Keywords:

Paediatric, Chest radiography, Observer variability, Agreement, Digital radiography, Lung opacity

Abstract

Background: Interpretation of pediatric chest radiographs is subject to observer variability, and digital image processing may further influence visual assessment. We evaluated interobserver agreement and the effect of controlled contrast manipulation on the interpretation of pediatric chest radiographs.

Methods: Sixty anonymised pediatric chest radiographs, comprising 30 classified as normal and 30 as abnormal, were presented in three display conditions: original (OI), contrast-reduced (CRR), and contrast-enhanced (CER). Five pediatricians independently assessed all 180 image presentations for image quality, radiographic features, and final interpretation as normal or abnormal. Interobserver agreement was assessed using Fleiss' kappa and intraobserver consistency using Cohen's kappa. Display-related changes in abnormal and lung-opacity assessments were compared at the underlying radiograph level using paired Wilcoxon signed-rank tests.

Results: Overall interobserver agreement for final interpretation was moderate (κ=0.528), with complete agreement in 48.3% of presentations. Agreement was κ=0.496, 0.594, and 0.479 for OI, CRR, and CER, respectively. Among radiographs classified as normal, lung-opacity identification increased from 8.7% with OI to 32.7% with CRR (p=0.012) and 27.3% with CER (p=0.018). Final normal/abnormal classification did not differ significantly between display conditions.

Conclusions: Controlled contrast manipulation influenced recognition of lung opacity despite unchanged underlying radiographic content. Standardised display conditions and awareness of perceptual variability may be important in pediatric chest radiograph interpretation.

References

Arthur R. Interpretation of the paediatric chest X-ray. Paediatr Respir Rev. 2000;1(1):41-50.

Johnson J, Kline JA. Intraobserver and interobserver agreement of the interpretation of pediatric chest radiographs. Emerg Radiol. 2010;17(4):285-90.

Elemraid MA, Muller M, Spencer DA, Rushton SP, Gorton R, Thomas MF, et al. Accuracy of the interpretation of chest radiographs for the diagnosis of paediatric pneumonia. PLoS One. 2014;9(8):e106051.

Voigt GM, Thiele D, Wetzke M, Weidemann J, Parpatt PM, Welte T, et al. Interobserver agreement in interpretation of chest radiographs for pediatric community acquired pneumonia: findings of the pedCAPNETZ-cohort. Pediatr Pulmonol. 2021;56(8):2676-85.

Cherian T, Mulholland EK, Carlin JB, Ostensen H, Amin R, de Campo M, et al. Standardized interpretation of paediatric chest radiographs for the diagnosis of pneumonia in epidemiological studies. Bull World Health Organ. 2005;83(5):353-9.

Kheddache S, Denbratt L, Angelhed JE. Digital chest radiography-optimizing image processing parameters for the visibility of chest lesions and anatomy. Eur J Radiol. 1996;22(3):241-5.

Brady AP. Error and discrepancy in radiology: inevitable or avoidable? Insights Imaging. 2017;8(1):171-82.

Rotem-Grunbaum B, Scheuerman O, Tamary O, Lakovsky Y, Shkalim Zemer V, Goldberg L, et al. Pediatric chest radiograph interpretation in a real-life setting. Eur J Pediatr. 2024;183(10):4435-44.

Fleiss JL. Measuring nominal scale agreement among many raters. Psychol Bull. 1971;76(5):378-82.

Geijer M, Båth M, Wessman C. Some common statistical methods for assessing rater agreement in radiological studies. Acta Radiol. 2025;66(6):675-83.