PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound
Transrectal ultrasound (TRUS) videos offer valuable histopathologic information about the prostate. Accurate prostate zonal segmentation in TRUS videos is vital for diagnosing prostate cancer and guiding surgery. However, TRUS videos are manually recorded by urologists, resulting in no standardized...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-01-01
|
| Series: | Advanced Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/aisy.202400302 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832592573567336448 |
|---|---|
| author | Jianguo Ju Qian Zhang Pengfei Xu Tiange Liu Cheng Li Ziyu Guan |
| author_facet | Jianguo Ju Qian Zhang Pengfei Xu Tiange Liu Cheng Li Ziyu Guan |
| author_sort | Jianguo Ju |
| collection | DOAJ |
| description | Transrectal ultrasound (TRUS) videos offer valuable histopathologic information about the prostate. Accurate prostate zonal segmentation in TRUS videos is vital for diagnosing prostate cancer and guiding surgery. However, TRUS videos are manually recorded by urologists, resulting in no standardized coordinate system, which limits direct prostate zonal segmentation in these videos. To overcome the limitation, a novel Prostate Zonal Segmentation Network (PZS‐Net), based on U‐Net, which learns critical cross‐frame information and multi‐scale features from sequential frames, is proposed. First, a sequential frame cross‐attention (SFCA) module is designed to capture remote information from sequential frames to enhance the feature representation of the current frame. The SFCA module is embedded at each skip connection layer to extract crucial cross‐frame information. Then, a multi‐scale fusion (MSF) module that utilizes three parallel branches with different atrous convolutions is designed. The MSF module is placed at the bottleneck layer to dynamically fuse multi‐scale context information from high‐level features. Extensive experiments on TRUS image datasets show that the PZS‐Net achieves higher accuracy in both the transitional zone (dice coefficient [Dice]: 68.90% ± 1.73%, mean intersection over union [mIoU]: 59.19% ± 2.09%, 95% Hausdorff distance [HD95]: 5.02 ± 0.83 mm) and the peripheral zone (Dice: 63.99% ± 3.16%, mIoU: 54.60% ± 3.35%, HD95: 5.28 ± 1.12 mm) and demonstrates the effectiveness and competitiveness of its key components via comprehensive ablation studies. |
| format | Article |
| id | doaj-art-73ae46f217a6483ba3cfcc1a7d7ce0e4 |
| institution | Kabale University |
| issn | 2640-4567 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Advanced Intelligent Systems |
| spelling | doaj-art-73ae46f217a6483ba3cfcc1a7d7ce0e42025-01-21T07:26:27ZengWileyAdvanced Intelligent Systems2640-45672025-01-0171n/an/a10.1002/aisy.202400302PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal UltrasoundJianguo Ju0Qian Zhang1Pengfei Xu2Tiange Liu3Cheng Li4Ziyu Guan5School of Information Science and Technology Northwest University Xi'an Shaanxi 710127 ChinaSchool of Information Science and Technology Northwest University Xi'an Shaanxi 710127 ChinaSchool of Information Science and Technology Northwest University Xi'an Shaanxi 710127 ChinaSchool of Intelligent Science and Technology University of Science and Technology Beijing Beijing 100083 ChinaDepartment of Ultrasound Ruijin Hospital Shanghai Jiao Tong University School of Medicine Shanghai 200025 ChinaSchool of Information Science and Technology Northwest University Xi'an Shaanxi 710127 ChinaTransrectal ultrasound (TRUS) videos offer valuable histopathologic information about the prostate. Accurate prostate zonal segmentation in TRUS videos is vital for diagnosing prostate cancer and guiding surgery. However, TRUS videos are manually recorded by urologists, resulting in no standardized coordinate system, which limits direct prostate zonal segmentation in these videos. To overcome the limitation, a novel Prostate Zonal Segmentation Network (PZS‐Net), based on U‐Net, which learns critical cross‐frame information and multi‐scale features from sequential frames, is proposed. First, a sequential frame cross‐attention (SFCA) module is designed to capture remote information from sequential frames to enhance the feature representation of the current frame. The SFCA module is embedded at each skip connection layer to extract crucial cross‐frame information. Then, a multi‐scale fusion (MSF) module that utilizes three parallel branches with different atrous convolutions is designed. The MSF module is placed at the bottleneck layer to dynamically fuse multi‐scale context information from high‐level features. Extensive experiments on TRUS image datasets show that the PZS‐Net achieves higher accuracy in both the transitional zone (dice coefficient [Dice]: 68.90% ± 1.73%, mean intersection over union [mIoU]: 59.19% ± 2.09%, 95% Hausdorff distance [HD95]: 5.02 ± 0.83 mm) and the peripheral zone (Dice: 63.99% ± 3.16%, mIoU: 54.60% ± 3.35%, HD95: 5.28 ± 1.12 mm) and demonstrates the effectiveness and competitiveness of its key components via comprehensive ablation studies.https://doi.org/10.1002/aisy.202400302cross‐attentionsmulti‐scalesprostate zonal segmentationssequential framestransrectal ultrasounds |
| spellingShingle | Jianguo Ju Qian Zhang Pengfei Xu Tiange Liu Cheng Li Ziyu Guan PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound Advanced Intelligent Systems cross‐attentions multi‐scales prostate zonal segmentations sequential frames transrectal ultrasounds |
| title | PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound |
| title_full | PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound |
| title_fullStr | PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound |
| title_full_unstemmed | PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound |
| title_short | PZS‐Net: Incorporating of Frame Sequence and Multi‐Scale Priors for Prostate Zonal Segmentation in Transrectal Ultrasound |
| title_sort | pzs net incorporating of frame sequence and multi scale priors for prostate zonal segmentation in transrectal ultrasound |
| topic | cross‐attentions multi‐scales prostate zonal segmentations sequential frames transrectal ultrasounds |
| url | https://doi.org/10.1002/aisy.202400302 |
| work_keys_str_mv | AT jianguoju pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound AT qianzhang pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound AT pengfeixu pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound AT tiangeliu pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound AT chengli pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound AT ziyuguan pzsnetincorporatingofframesequenceandmultiscalepriorsforprostatezonalsegmentationintransrectalultrasound |