A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams
Abstract Multiple studies have created state-of-the-art liver segmentation models using Deep Convolutional Neural Networks (DCNNs) such as the V-net and H-DenseUnet. Oversegmentation however continues to be a problem. We set forth to address these limitations by developing a an automated workflow th...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-09-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-20108-8 |
| _version_ | 1818016724091404288 |
|---|---|
| author | Jayasuriya Senthilvelan Neema Jamshidi |
| author_facet | Jayasuriya Senthilvelan Neema Jamshidi |
| author_sort | Jayasuriya Senthilvelan |
| collection | DOAJ |
| description | Abstract Multiple studies have created state-of-the-art liver segmentation models using Deep Convolutional Neural Networks (DCNNs) such as the V-net and H-DenseUnet. Oversegmentation however continues to be a problem. We set forth to address these limitations by developing a an automated workflow that leverages the strengths of different DCNN architectures, resulting in a pipeline that enables fully automated liver segmentation. A Pipeline for Automated Deep Learning Liver Segmentation (PADLLS) was developed and implemented that cascades multiple DCNNs that were trained on more than 200 CT scans. First, a V-net is used to create a rough liver, spleen, and stomach mask. After stomach and spleen pixels are removed using their respective masks and ascites is removed using a morphological algorithm, the scan is passed to a H-DenseUnet to yield the final segmentation. The segmentation accuracy of the pipleline was compared to the H-DenseUnet and the V-net using the SLIVER07 and 3DIRCADb datasets as benchmarks. The PADLLS Dice score for the SLIVER07 dataset was calculated to be 0.957 ± 0.033 and was significantly better than the H-DenseUnet’s score of 0.927 ± 0.044 (p = 0.0219) and the V-net’s score of 0.872 ± 0.121 (p = 0.0067). The PADLLS Dice score for the 3DIRCADb dataset was 0.965 ± 0.016 and was significantly better than the H-DenseUnet’s score of 0.930 ± 0.041 (p = 0.0014) the V-net’s score of 0.874 ± 0.060 (p < 0.001). In conclusion, our pipeline (PADLLS) outperforms existing liver segmentation models, serves as a valuable tool for image-based analysis, and is freely available for download and use. |
| first_indexed | 2024-04-14T07:17:32Z |
| format | Article |
| id | doaj.art-1c062265bea345d391b37333cda08fde |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-14T07:17:32Z |
| publishDate | 2022-09-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-1c062265bea345d391b37333cda08fde2022-12-22T02:06:16ZengNature PortfolioScientific Reports2045-23222022-09-0112111110.1038/s41598-022-20108-8A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT examsJayasuriya Senthilvelan0Neema Jamshidi1Department of Radiological Sciences, David Geffen School of Medicine, University of California, Los AngelesDepartment of Radiological Sciences, David Geffen School of Medicine, University of California, Los AngelesAbstract Multiple studies have created state-of-the-art liver segmentation models using Deep Convolutional Neural Networks (DCNNs) such as the V-net and H-DenseUnet. Oversegmentation however continues to be a problem. We set forth to address these limitations by developing a an automated workflow that leverages the strengths of different DCNN architectures, resulting in a pipeline that enables fully automated liver segmentation. A Pipeline for Automated Deep Learning Liver Segmentation (PADLLS) was developed and implemented that cascades multiple DCNNs that were trained on more than 200 CT scans. First, a V-net is used to create a rough liver, spleen, and stomach mask. After stomach and spleen pixels are removed using their respective masks and ascites is removed using a morphological algorithm, the scan is passed to a H-DenseUnet to yield the final segmentation. The segmentation accuracy of the pipleline was compared to the H-DenseUnet and the V-net using the SLIVER07 and 3DIRCADb datasets as benchmarks. The PADLLS Dice score for the SLIVER07 dataset was calculated to be 0.957 ± 0.033 and was significantly better than the H-DenseUnet’s score of 0.927 ± 0.044 (p = 0.0219) and the V-net’s score of 0.872 ± 0.121 (p = 0.0067). The PADLLS Dice score for the 3DIRCADb dataset was 0.965 ± 0.016 and was significantly better than the H-DenseUnet’s score of 0.930 ± 0.041 (p = 0.0014) the V-net’s score of 0.874 ± 0.060 (p < 0.001). In conclusion, our pipeline (PADLLS) outperforms existing liver segmentation models, serves as a valuable tool for image-based analysis, and is freely available for download and use.https://doi.org/10.1038/s41598-022-20108-8 |
| spellingShingle | Jayasuriya Senthilvelan Neema Jamshidi A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams Scientific Reports |
| title | A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams |
| title_full | A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams |
| title_fullStr | A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams |
| title_full_unstemmed | A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams |
| title_short | A pipeline for automated deep learning liver segmentation (PADLLS) from contrast enhanced CT exams |
| title_sort | pipeline for automated deep learning liver segmentation padlls from contrast enhanced ct exams |
| url | https://doi.org/10.1038/s41598-022-20108-8 |
| work_keys_str_mv | AT jayasuriyasenthilvelan apipelineforautomateddeeplearningliversegmentationpadllsfromcontrastenhancedctexams AT neemajamshidi apipelineforautomateddeeplearningliversegmentationpadllsfromcontrastenhancedctexams AT jayasuriyasenthilvelan pipelineforautomateddeeplearningliversegmentationpadllsfromcontrastenhancedctexams AT neemajamshidi pipelineforautomateddeeplearningliversegmentationpadllsfromcontrastenhancedctexams |