Related Experiment Video
Updated: Jul 30, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.5D MFFAU-Net: a convolutional neural network for kidney segmentation
Peng Sun1, Zengnan Mo2, Fangrong Hu1
1School of Electronic Engineering and Automation, Guilin University of Electronic Technology, Guilin, 541004, Guangxi, China.
Background:
Kidney tumors have become increasingly prevalent among adults and are now considered one of the most common types of tumors. Accurate segmentation of kidney tumors can help physicians assess tumor complexity and aggressiveness before surgery. However, segmenting kidney tumors manually can be difficult because of their heterogeneity.
Methods:
This paper proposes a 2.5D MFFAU-Net (multi-level Feature Fusion Attention U-Net) to segment kidneys, tumors and cysts. First, we propose a 2.5D model for learning to combine and represent a given slice in 2D slices, thereby introducing 3D information to balance memory consumption and model complexity. Then, we propose a ResConv architecture in MFFAU-Net and use the high-level and low-level feature in the model. Finally, we use multi-level information to analyze the spatial features between slices to segment kidneys and tumors.
Results:
The 2.5D MFFAU-Net was evaluated on KiTS19 and KiTS21 kidney datasets and demonstrated an average dice score of 0.924 and 0.875, respectively, and an average Surface dice (SD) score of 0.794 in KiTS21.
Conclusion:
The 2.5D MFFAU-Net model can effectively segment kidney tumors, and the results are comparable to those obtained with high-performance 3D CNN models, and have the potential to serve as a point of reference in clinical practice.
Insights
A novel 2.5D multi-level Feature Fusion Attention U-Net (MFFAU-Net) effectively segments kidney tumors. This deep learning approach achieves high accuracy, offering a valuable tool for clinical practice.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Kidney tumors are increasingly prevalent, posing diagnostic challenges due to their heterogeneity.
- Accurate segmentation is crucial for surgical planning and assessing tumor aggressiveness.
- Manual segmentation is time-consuming and difficult.
Purpose of the Study:
- To develop an automated method for segmenting kidneys, tumors, and cysts.
- To introduce a novel 2.5D deep learning model balancing performance and computational cost.
- To improve the accuracy of kidney tumor segmentation in medical imaging.
Main Methods:
- Proposed a 2.5D multi-level Feature Fusion Attention U-Net (MFFAU-Net) architecture.
- Integrated high-level and low-level features using a ResConv architecture.
- Utilized multi-level feature fusion for spatial analysis between slices.
Main Results:
- Achieved an average Dice score of 0.924 on the KiTS19 dataset.
- Obtained an average Dice score of 0.875 and Surface Dice score of 0.794 on the KiTS21 dataset.
- Demonstrated comparable performance to high-performance 3D Convolutional Neural Network (CNN) models.
Conclusions:
- The 2.5D MFFAU-Net model effectively segments kidney tumors.
- The model shows potential as a clinical reference tool for kidney tumor analysis.
- This approach offers an efficient alternative to complex 3D models for segmentation.
Related Concept Videos
01:28Imaging Studies I: Kidney, Ureter, and Bladder Studies
01:45Kidney Structure
01:20Renal Corpuscle
Glomerulus: Structure and Function
The glomerulus is a tiny, intricate network of capillaries located at the beginning of the nephron. It's enveloped by the Bowman's capsule and receives its blood supply from an afferent arteriole, which divides into numerous...
01:12Internal Anatomy of the Kidney
Anatomical Position and Dimensions
The kidneys are retroperitoneal organs positioned against the posterior abdominal wall on either side of the spine, roughly between the twelfth thoracic and third lumbar vertebrae. Each kidney is typically 10-12 cm long, 5-6 cm wide, and 3-4 cm thick, weighing about 150 grams.
Renal Cortex
The outermost region of the kidney is the...
01:21External Anatomy of the Kidney
The kidneys are located in the retroperitoneal space on either side of the vertebral column, protected posteriorly by the 11th and 12th ribs. The right kidney sits slightly lower than the left owing to the presence of the liver...
01:18Blood and Nerve Supply to the Kidney
Bloody Supply to the Kidneys:
The kidneys receive their blood supply from the renal arteries, which branch off from the abdominal aorta—the main artery supplying the abdomen and lower body. The renal arteries enter the kidneys at the hilum, a notch on the medial side of...