New AI model improves pancreatic tumor detection with high accuracy and efficiency
TRIUNE-Net: Harmonizing Scale, Shape, and Efficiency in Pancreatic Tumor Segmentation
Computer Vision and Pattern Recognition
Summary
Detecting pancreatic tumors in 3D medical scans is difficult because of how much the size and shape of tumors can change. The authors created a new lightweight AI system called TRIUNE-Net that looks at the scans in multiple ways to better understand these differences. It uses special techniques to keep important details and adapt to unusual tumor shapes, all while running efficiently. Their approach beats previous models in accuracy without needing extra training on other data. This means it can better find and outline tumors, which could help doctors make decisions.
pancreatic tumor segmentation3D CT volumesmulti-scale context aggregationdilated convolutionsdeformable convolutionsattention mechanismdownsamplingDice scoreF1 scoresensitivity and precision
Authors
Amir Hossein Saleknia, Alireza Kheyrkhah, Sanaz Karimijafarbigloo, Reza Azad, Sina Houshmand, Ulas Bagci, Dorit Merhof, Alaa Sulaiman
Abstract
Pancreatic tumor segmentation in 3D CT volumes is challenged by extreme scale variability across both the pancreas and tumor, and highly irregular tumor morphology. While recent advances have pushed segmentation performance, existing methods do not explicitly address these challenges and come at the cost of excessive computational complexity, limiting their practicality in resource-constrained clinical environments. We propose TRIUNE-Net, a lightweight unified architecture that harmonizes scale, shape, and efficiency through three synergistic innovations. A multi-scale context aggregation module with stage-adaptive dilated convolutions enables the model to reason across the broad range of anatomical scales present in both organs. A serial linear-deformable attention mechanism combines large effective receptive fields with shapeadaptive deformable convolutions to capture irregular, non-convex tumor morphologies. Finally, an information-preserving downsampling module replaces conventional max pooling entirely, retaining all spatial information while adding negligible parameters, preventing small tumors from being discarded before they can be recognized. On both the MSD Pancreas and NVD Pancreas datasets, TRIUNE-Net achieves state-of-theart results with only 5.86 M parameters and no external pre-training, outperforming all baselines across all key tumor metrics. Specifically, it surpasses the next-best model by 0.45% in tumor Dice, 6.0 points in F1 score, 6.6 points in sensitivity, and 3.4 points in precision, simultaneously reflecting its ability to suppress both missed tumors and false alarms in clinically realistic conditions. Our code is available at: https://github.com/abdora-ai/TRIUNE-Net