The Greulich-Pyle and Gilsanz-Ratib atlas method versus automated estimation tool for bone age: a multi-observer...

Area of Science:

• Medical Imaging
• Pediatric Endocrinology
• Radiology

Background:

• Accurate bone age estimation is crucial for diagnosing and managing pediatric growth disorders.
• Inter-observer variability in bone age assessment can lead to diagnostic and treatment discrepancies.

Purpose of the Study:

• To compare bone age estimation agreement between observers using Greulich-Pyle (GP) and Gilsanz-Ratib (GR) methods.
• To evaluate agreement across four medical specialties: radiology, pediatrics, pediatric endocrinology, and pediatric radiology.
• To assess the agreement between human observers and an automated bone age estimation tool.

Main Methods:

• 99 observers participated in a questionnaire-based study.
• The BoneXpert software was used as the automated tool for bone age estimation.
• Agreement was assessed using coefficient of variance (CV), intraclass correlation coefficient (ICC), root-mean-square-error (RMSE), and Bland-Altman analysis.

Main Results:

• Intraclass correlation coefficients (ICC) were high, ranging from 0.963 to 0.990 across methods and specialties.
• The automated BoneXpert tool demonstrated the lowest RMSE values (0.504 years for GP atlas).
• CV values ranged from 4.98% to 22.08%.

Conclusions:

• Automated bone age estimation using BoneXpert provides results comparable to established GP and GR methods.
• The utilization of automated tools has the potential to significantly decrease inter-observer variability in bone age assessment.
• This technology offers a promising avenue for improving the consistency and accuracy of bone age determination.