LSKANet : Long Strip Kernel Attention Network for Robotic Surgical Scene Segmentation
Surgical scene segmentation is a critical task in Robotic-assisted surgery. However, the complexity of the surgical scene, which mainly includes local feature similarity (e.g., between different anatomical tissues), intraoperative complex artifacts, and indistinguishable boundaries, poses significant challenges to accurate segmentation. To tackle these problems, we propose the Long Strip Kernel Attention network (LSKANet), including two well-designed modules named Dual-block Large Kernel Attention module (DLKA) and Multiscale Affinity Feature Fusion module (MAFF), which can implement precise segmentation of surgical images. Specifically, by introducing strip convolutions with different topologies (cascaded and parallel) in two blocks and a large kernel design, DLKA can make full use of region- and strip-like surgical features and extract both visual and structural information to reduce the false segmentation caused by local feature similarity. In MAFF, affinity matrices calculated from multiscale feature maps are applied as feature fusion weights, which helps to address the interference of artifacts by suppressing the activations of irrelevant regions. Besides, the hybrid loss with Boundary Guided Head (BGH) is proposed to help the network segment indistinguishable boundaries effectively. We evaluate the proposed LSKANet on three datasets with different surgical scenes. The experimental results show that our method achieves new state-of-the-art results on all three datasets with improvements of 2.6%, 1.4%, and 3.4% mIoU, respectively. Furthermore, our method is compatible with different backbones and can significantly increase their segmentation accuracy. Code is available at https://github.com/YubinHan73/LSKANet.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2024 |
|---|---|
| Erschienen: |
2024 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:43 |
|---|---|
| Enthalten in: |
IEEE transactions on medical imaging - 43(2024), 4 vom: 28. Apr., Seite 1308-1322 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Liu, Min [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
|---|
| Anmerkungen: |
Date Completed 04.04.2024 Date Revised 04.04.2024 published: Print-Electronic Citation Status MEDLINE |
|---|
| doi: |
10.1109/TMI.2023.3335406 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM365068683 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM365068683 | ||
| 003 | DE-627 | ||
| 005 | 20240404234425.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231226s2024 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1109/TMI.2023.3335406 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1364.xml |
| 035 | |a (DE-627)NLM365068683 | ||
| 035 | |a (NLM)38015689 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Liu, Min |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a LSKANet |b Long Strip Kernel Attention Network for Robotic Surgical Scene Segmentation |
| 264 | 1 | |c 2024 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 04.04.2024 | ||
| 500 | |a Date Revised 04.04.2024 | ||
| 500 | |a published: Print-Electronic | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a Surgical scene segmentation is a critical task in Robotic-assisted surgery. However, the complexity of the surgical scene, which mainly includes local feature similarity (e.g., between different anatomical tissues), intraoperative complex artifacts, and indistinguishable boundaries, poses significant challenges to accurate segmentation. To tackle these problems, we propose the Long Strip Kernel Attention network (LSKANet), including two well-designed modules named Dual-block Large Kernel Attention module (DLKA) and Multiscale Affinity Feature Fusion module (MAFF), which can implement precise segmentation of surgical images. Specifically, by introducing strip convolutions with different topologies (cascaded and parallel) in two blocks and a large kernel design, DLKA can make full use of region- and strip-like surgical features and extract both visual and structural information to reduce the false segmentation caused by local feature similarity. In MAFF, affinity matrices calculated from multiscale feature maps are applied as feature fusion weights, which helps to address the interference of artifacts by suppressing the activations of irrelevant regions. Besides, the hybrid loss with Boundary Guided Head (BGH) is proposed to help the network segment indistinguishable boundaries effectively. We evaluate the proposed LSKANet on three datasets with different surgical scenes. The experimental results show that our method achieves new state-of-the-art results on all three datasets with improvements of 2.6%, 1.4%, and 3.4% mIoU, respectively. Furthermore, our method is compatible with different backbones and can significantly increase their segmentation accuracy. Code is available at https://github.com/YubinHan73/LSKANet | ||
| 650 | 4 | |a Journal Article | |
| 700 | 1 | |a Han, Yubin |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Jiazheng |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Can |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Yaonan |e verfasserin |4 aut | |
| 700 | 1 | |a Meijering, Erik |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t IEEE transactions on medical imaging |d 1982 |g 43(2024), 4 vom: 28. Apr., Seite 1308-1322 |w (DE-627)NLM082855269 |x 1558-254X |7 nnns |
| 773 | 1 | 8 | |g volume:43 |g year:2024 |g number:4 |g day:28 |g month:04 |g pages:1308-1322 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1109/TMI.2023.3335406 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 43 |j 2024 |e 4 |b 28 |c 04 |h 1308-1322 | ||