Carpal bone segmentation using fully convolutional neural network
Background: Bone Age Assessment (BAA) refers to a clinical procedure that aims to identify a discrepancy between biological and chronological age of an individual by assessing the bone age growth. Currently, there are two main methods of executing BAA which are known as Greulich-Pyle and Tanner-Whit...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Published: |
Bentham Science Publishers
2019
|
| Subjects: |
| _version_ | 1825720872979136512 |
|---|---|
| author | Liang, Kim Meng Khalil, Azira Nizar, Muhammad Hanif Ahmad Nisham, Maryam Kamarun Pingguan-Murphy, Belinda Hum, Yan Chai Salim, Maheza Irna Mohamad Lai, Khin Wee |
| author_facet | Liang, Kim Meng Khalil, Azira Nizar, Muhammad Hanif Ahmad Nisham, Maryam Kamarun Pingguan-Murphy, Belinda Hum, Yan Chai Salim, Maheza Irna Mohamad Lai, Khin Wee |
| author_sort | Liang, Kim Meng |
| collection | UM |
| description | Background: Bone Age Assessment (BAA) refers to a clinical procedure that aims to identify a discrepancy between biological and chronological age of an individual by assessing the bone age growth. Currently, there are two main methods of executing BAA which are known as Greulich-Pyle and Tanner-Whitehouse techniques. Both techniques involve a manual and qualitative assessment of hand and wrist radiographs, resulting in intra and inter-operator variability accuracy and time-consuming. An automatic segmentation can be applied to the radiographs, providing the physician with more accurate delineation of the carpal bone and accurate quantitative analysis. Methods: In this study, we proposed an image feature extraction technique based on image segmentation with the fully convolutional neural network with eight stride pixel (FCN-8). A total of 290 radiographic images including both female and the male subject of age ranging from 0 to 18 were manually segmented and trained using FCN-8. Results and Conclusion: The results exhibit a high training accuracy value of 99.68% and a loss rate of 0.008619 for 50 epochs of training. The experiments compared 58 images against the gold standard ground truth images. The accuracy of our fully automated segmentation technique is 0.78 ± 0.06, 1.56 ±0.30 mm and 98.02% in terms of Dice Coefficient, Hausdorff Distance, and overall qualitative carpal recognition accuracy, respectively. © 2019 Bentham Science Publishers. |
| first_indexed | 2024-03-06T05:41:43Z |
| format | Article |
| id | um.eprints-16692 |
| institution | Universiti Malaya |
| last_indexed | 2024-03-06T05:41:43Z |
| publishDate | 2019 |
| publisher | Bentham Science Publishers |
| record_format | dspace |
| spelling | um.eprints-166922020-03-17T03:04:16Z http://eprints.um.edu.my/16692/ Carpal bone segmentation using fully convolutional neural network Liang, Kim Meng Khalil, Azira Nizar, Muhammad Hanif Ahmad Nisham, Maryam Kamarun Pingguan-Murphy, Belinda Hum, Yan Chai Salim, Maheza Irna Mohamad Lai, Khin Wee R Medicine Background: Bone Age Assessment (BAA) refers to a clinical procedure that aims to identify a discrepancy between biological and chronological age of an individual by assessing the bone age growth. Currently, there are two main methods of executing BAA which are known as Greulich-Pyle and Tanner-Whitehouse techniques. Both techniques involve a manual and qualitative assessment of hand and wrist radiographs, resulting in intra and inter-operator variability accuracy and time-consuming. An automatic segmentation can be applied to the radiographs, providing the physician with more accurate delineation of the carpal bone and accurate quantitative analysis. Methods: In this study, we proposed an image feature extraction technique based on image segmentation with the fully convolutional neural network with eight stride pixel (FCN-8). A total of 290 radiographic images including both female and the male subject of age ranging from 0 to 18 were manually segmented and trained using FCN-8. Results and Conclusion: The results exhibit a high training accuracy value of 99.68% and a loss rate of 0.008619 for 50 epochs of training. The experiments compared 58 images against the gold standard ground truth images. The accuracy of our fully automated segmentation technique is 0.78 ± 0.06, 1.56 ±0.30 mm and 98.02% in terms of Dice Coefficient, Hausdorff Distance, and overall qualitative carpal recognition accuracy, respectively. © 2019 Bentham Science Publishers. Bentham Science Publishers 2019 Article PeerReviewed Liang, Kim Meng and Khalil, Azira and Nizar, Muhammad Hanif Ahmad and Nisham, Maryam Kamarun and Pingguan-Murphy, Belinda and Hum, Yan Chai and Salim, Maheza Irna Mohamad and Lai, Khin Wee (2019) Carpal bone segmentation using fully convolutional neural network. Current Medical Imaging, 15 (10). pp. 983-989. ISSN 1573-4056, DOI https://doi.org/10.2174/1573405615666190724101600 <https://doi.org/10.2174/1573405615666190724101600>. https://doi.org/10.2174/1573405615666190724101600 doi:10.2174/1573405615666190724101600 |
| spellingShingle | R Medicine Liang, Kim Meng Khalil, Azira Nizar, Muhammad Hanif Ahmad Nisham, Maryam Kamarun Pingguan-Murphy, Belinda Hum, Yan Chai Salim, Maheza Irna Mohamad Lai, Khin Wee Carpal bone segmentation using fully convolutional neural network |
| title | Carpal bone segmentation using fully convolutional neural network |
| title_full | Carpal bone segmentation using fully convolutional neural network |
| title_fullStr | Carpal bone segmentation using fully convolutional neural network |
| title_full_unstemmed | Carpal bone segmentation using fully convolutional neural network |
| title_short | Carpal bone segmentation using fully convolutional neural network |
| title_sort | carpal bone segmentation using fully convolutional neural network |
| topic | R Medicine |
| work_keys_str_mv | AT liangkimmeng carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT khalilazira carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT nizarmuhammadhanifahmad carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT nishammaryamkamarun carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT pingguanmurphybelinda carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT humyanchai carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT salimmahezairnamohamad carpalbonesegmentationusingfullyconvolutionalneuralnetwork AT laikhinwee carpalbonesegmentationusingfullyconvolutionalneuralnetwork |