| YANG Anning,LU Na,JIANG Huayong,CHEN Diandian,YU Yanjun,WANG Yadi,Wang Qiusheng,Zhang Fuli. Automatic delineation of organs at risk in non-small cell lung cancer radiotherapy based on deep learning networks. Oncol Transl Med, 2022, 8: 83-88. |
| Automatic delineation of organs at risk in non-small cell lung cancer radiotherapy based on deep learning networks |
| Received:January 15, 2022 Revised:April 19, 2022 |
| View Full Text View/Add Comment Download reader |
| KeyWord:non-small cell lung cancer; organs at risk; medical image segmentation; deep learning; DenseNet |
| Author Name | Affiliation | Postcode | | YANG Anning | School of Automation Science and Electrical Engineering,Beihang University | 100191 | | LU Na | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | | | JIANG Huayong | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | | | CHEN Diandian | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | | | YU Yanjun | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | | | WANG Yadi | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | | | Wang Qiusheng | School of Automation Science and Electrical Engineering,Beihang University | | | Zhang Fuli | Radiation Oncology Department,the Seventh Medical Center of Chinese PLA General Hospital | 100700 |
|
| Hits: 2929 |
| Download times: 3359 |
| Abstract: |
| Objective To introduce an end-to-end automatic segmentation method for organs at risk (OARs) in chest
computed tomography (CT) images based on dense connection deep learning and to provide an accurate
auto-segmentation model to reduce the workload on radiation oncologists.
Methods CT images of 36 lung cancer cases were included in this study. Of these, 27 cases were
randomly selected as the training set, six cases as the validation set, and nine cases as the testing set. The
left and right lungs, cord, and heart were auto-segmented, and the training time was set to approximately
5 h. The testing set was evaluated using geometric metrics including the Dice similarity coefficient (DSC),
95% Hausdorff distance (HD95), and average surface distance (ASD). Thereafter, two sets of treatment
plans were optimized based on manually contoured OARs and automatically contoured OARs, respectively.
Dosimetric parameters including Dmax and Vx of the OARs were obtained and compared.
Results The proposed model was superior to U-Net in terms of the DSC, HD95, and ASD, although there
was no significant difference in the segmentation results yielded by both networks (P > 0.05). Compared
to manual segmentation, auto-segmentation significantly reduced the segmentation time by nearly 40.7%
(P < 0.05). Moreover, the differences in dose-volume parameters between the two sets of plans were not
statistically significant (P > 0.05).
Conclusion The bilateral lung, cord, and heart could be accurately delineated using the DenseNetbased deep learning method. Thus, feature map reuse can be a novel approach to medical image autosegmentation . |
| Close |
|
|
|