IMAU-Net: A Hybrid Multi-Scale Deep Learning Framework for Liver Segmentation from Laparoscopic Images
Sensors · 27.04.2026 · 26 (9) · 2695
DOI: 10.3390/s26092695 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 27.04.2026. Verfasst von Syeda Sitara Waseem und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science & IT, The Government Sadiq College Women University, Bahawalpur 63100, Pakistan. Hauptthemen laut MEDLINE: Deep Learning, Laparoscopy, Liver, Image Processing, Computer-Assisted. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Norwegian University of Science and Technology, NTNU Norwegian University of Science and Technology.
Abstract (Originalsprache)
Accurate liver segmentation in laparoscopic surgery is critical but remains challenging due to low contrast, occlusion, and irregular organ boundaries. While deep learning has advanced medical image segmentation, existing models often trade off between accuracy, computational efficiency, and boundary precision. We propose IMAU-Net, a hybrid architecture integrating a pre-trained InceptionV3 encoder with a novel bottleneck combining Multi-Core Pooling (MCP) and enhanced Atrous Spatial Pyramid Pooling (ASPP). The MCP module captures fine-to-medium spatial details through parallel multi-kernel pooling, while ASPP extracts multi-scale contextual information via dilated convolutions. Evaluated on the M2CAI dataset with 5-fold cross-validation, IMAU-Net achieves a mean Dice coefficient of 0.9179 ± 0.012 and IoU of 0.8483 ± 0.015. Furthermore, external validation on the independent CholecSeg8K dataset (250 test samples) demonstrates generalizability across different laparoscopic procedures, achieving a Dice coefficient of 0.8745 ± 0.0312 and AUC of 0.9542, with a performance degradation of only 4.3% despite domain shift between liver surgery and cholecystectomy. Comparative analysis with state of the art methods demonstrates superior performance, with computational efficiency suitable for real-time applications (45 FPS, 42.3 M parameters). The proposed architecture provides an optimal balance between accuracy and efficiency for intraoperative guidance systems. While evaluated on retrospective laparoscopic image datasets rather than real-time intraoperative workflows, the model demonstrates potential for integration into surgical guidance systems pending prospective validation.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Science & IT, The Government Sadiq College Women University, Bahawalpur 63100, Pakistan.
- Department of Information Security and Communication Technology (IIK), Norwegian University of Science and Technology (NTNU), Teknologivegen 22, 2815 Gjøvik, Norway.
- Department of Computer Engineering, Gachon University, Seongnam-si 13120, Republic of Korea.
ORCID-Profile: Syeda Sitara Waseem, Sarang Shaikh, Syed Rizwan Hassan
Förderung
- Norwegian University of Science and Technology
- NTNU Norwegian University of Science and Technology
Zitieren
APA
Waseem, S. S., Shaikh, S., & Hassan, S. R. (2026). IMAU-Net: A Hybrid Multi-Scale Deep Learning Framework for Liver Segmentation from Laparoscopic Images. Sensors, 26(9), 2695. https://doi.org/10.3390/s26092695
Vancouver
Waseem SS, Shaikh S, Hassan SR. IMAU-Net: A Hybrid Multi-Scale Deep Learning Framework for Liver Segmentation from Laparoscopic Images. Sensors. 2026;26(9):2695. doi:10.3390/s26092695
BibTeX
@article{Waseem2026,
author = {Waseem, Syeda Sitara and Shaikh, Sarang and Hassan, Syed Rizwan},
title = {IMAU-Net: A Hybrid Multi-Scale Deep Learning Framework for Liver Segmentation from Laparoscopic Images},
journal = {Sensors},
year = {2026},
volume = {26},
number = {9},
pages = {2695},
doi = {10.3390/s26092695}
}
EUVIMED