Artificial intelligence in bone age assessment: accuracy and efficiency of a novel fully automated algorithm compared to the Greulich-Pyle method

European Radiology Experimental

Table 2 Summary of main results

Characteristics	Overall	Boys	Girls
Number of patients	514	252	262
Mean chronological age ± SD	10.20 ± 4.85	10.31 ± 3.93	10.10 ± 3.65
Mean reference BA ± SD	9.91 ± 2.85	9.95 ± 2.89	9.87 ± 2.87
Mean AI-derived BA ± SD	9.98 ± 2.94	9.99 ± 2.98	9.96 ± 2.91
Accuracy of AI-derived BA	MAD 0.34 RMSD 0.38	MAD 0.34 RMSD 0.40	MAD 0.33 RMSD 0.36
Mean GP-based BA ± SD	10.18 ± 3.94	10.21 ± 3.88	10.15 ± 3.97
Accuracy of GP-based BA	MAD 0.79 RMSD 0.89	MAD 0.79 RMSD 0.90	MAD 0.80 RMSD 0.88

Apart from number of patients, all values are given in years
AI Artificial intelligence, BA Bone age, GP Greulich-Pyle, MAD Mean absolute deviation, RMSD Root mean square deviation, SD Standard deviation