[Comparison of Three CNN Models Applied in Bone Age Assessment of Pelvic Radiographs of Adolescents].

ABSTRACT: Objective To compare the performance of three deep-learning models （VGG19, Inception-V3 and Inception-ResNet-V2） in automatic bone age assessment based on pelvic X-ray radiographs. Methods A total of 962 pelvic X ray radiographs taken from adolescents （481 males, 481 females） aged from 11.0 to 21.0 years in five provinces and cities of China were collected, preprocessed and used as objects of study. Eighty percent of these X ray radiographs were divided into training set and validation set with random sampling method and used for model fitting and hyper-parameters adjustment. Twenty percent were used as test sets, to evaluate the ability of model generalization. The performances of the three models were assessed by comparing the root mean square error （RMSE）, mean absolute error （MAE） and Bland-Altman plots between the model estimates and the chronological ages. Results The mean RMSE and MAE between bone age estimates of the VGG19 model and the chronological ages were 1.29 and 1.02 years, respectively. The mean RMSE and MAE between bone age estimates of the Inception-V3 model and the chronological ages were 1.17 and 0.82 years, respectively. The mean RMSE and MAE between bone age estimates of the Inception-ResNet-V2 model and the chronological ages were 1.11 and 0.84 years, respectively. The Bland-Altman plots showed that the mean value of differences between bone age estimates of Inception-ResNet-V2 model and the chronological ages was the lowest. Conclusion In the automatic bone age assessment of adolescent pelvis, the Inception-ResNet-V2 model performs the best while the Inception-V3 model achieves a similar accuracy as VGG19 model.

[Automated Assessment for Bone Age of Left Wrist Joint in Uyghur Teenagers by Deep Learning].

OBJECTIVES: To realize the automated bone age assessment by applying deep learning to digital radiography （DR） image recognition of left wrist joint in Uyghur teenagers, and explore its practical application value in forensic medicine bone age assessment. METHODS: The X-ray films of left wrist joint after pretreatment, which were taken from 245 male and 227 female Uyghur nationality teenagers in Uygur Autonomous Region aged from 13.0 to 19.0 years old, were chosen as subjects. And AlexNet was as a regression model of image recognition. From the total samples above, 60％ of male and female DR images of left wrist joint were selected as net train set, and 10％ of samples were selected as validation set. As test set, the rest 30％ were used to obtain the image recognition accuracy with an error range in ±1.0 and ±0.7 age respectively, compared to the real age. RESULTS: The modelling results of deep learning algorithm showed that when the error range was in ±1.0 and ±0.7 age respectively, the accuracy of the net train set was 81.4% and 75.6% in male, and 80.5% and 74.8% in female, respectively. When the error range was in ±1.0 and ±0.7 age respectively, the accuracy of the test set was 79.5% and 71.2% in male, and 79.4% and 66.2% in female, respectively. CONCLUSIONS: The combination of bone age research on teenagers' left wrist joint and deep learning, which has high accuracy and good feasibility, can be the research basis of bone age automatic assessment system for the rest joints of body.

[Advantages and Application Prospects of Deep Learning in Image Recognition and Bone Age Assessment].

Deep learning and neural network models have been new research directions and hot issues in the fields of machine learning and artificial intelligence in recent years. Deep learning has made a breakthrough in the applications of image and speech recognitions, and also has been extensively used in the fields of face recognition and information retrieval because of its special superiority. Bone X-ray images express different variations in black-white-gray gradations, which have image features of black and white contrasts and level differences. Based on these advantages of deep learning in image recognition, we combine it with the research of bone age assessment to provide basic datum for constructing a forensic automatic system of bone age assessment. This paper reviews the basic concept and network architectures of deep learning, and describes its recent research progress on image recognition in different research fields at home and abroad, and explores its advantages and application prospects in bone age assessment.