Segmentation problems in medical images
5 likes1,853 views
The document discusses three research articles on medical image segmentation using deep learning techniques, focusing on spinal cord gray matter, liver tumors, and myocardial segmentation in MRI. Each study presents unique methods and challenges related to segmentation accuracy and data variability, highlighting the significance of automation in medical imaging for improved diagnostics. The proposed methods offer advancements in segmentation accuracy while addressing challenges such as tissue contrast variability and image artifacts.
Segmentation problems in medical images
- 1. 2 4 O c t . 2 0 1 7 Segmentation Problems in Medical Images Jimin Lee Radiological Physics Laboratory, Seoul National University SNU TF 스터디 모임
- 2. Contents 2 1. Spinal cord gray matter segmentation using deep dilated convolutions ( https://arxiv.org/abs/1710.01269 ) 2. H-DenseUNet: Hybrid Densely Connected UNet for Liver and Liver Tumor Segmentation from CT Volumes ( https://arxiv.org/abs/1709.07330 ) 3. Automatic Myocardial Segmentation by Using A Deep Learning Network in Cardiac MRI ( https://arxiv.org/abs/1708.07452 )
- 3. Spinal cord gray matter segmentation using deep dilated convolutions October 4, 2017 * Article under submission to Nature Scientific Reports.
- 4. 1. Introduction 4 Purpose : To devise an end-to-end fully automated human spinal cord gray matter segmentation method using Deep Learning Gray matter (GM) tissue changes in spinal cord (SC) • It is linked to neurological disorders. • SCGM atrophy (위축) is a relevant biomarker for predicting disability in amyotrophic lateral sclerosis (루게릭병). The fully-automated segmentation is very useful for longitudinal studies. • The delineation of gray matter is very time-consuming.
- 5. 1. Introduction 5 Accurate segmentation of the GM is still a remaining challenge. • Inconsistent surrounding tissue intensities • Pathology-induced changes in the image contrast • Differences in MRI acquisition protocols • Lack of standardized data sets • Different pixel sizes • Image artifacts ⋮ MRI samples from different centers → Variability
- 6. 2. Methods and Materials 6 Dilated convolution https://github.com/vdumoulin/conv_arithmetic 3x3 Kenel with dilation rate 2
- 7. 2. Methods and Materials 7 Proposed method 2 layers each
- 8. 2. Methods and Materials 8 Loss : DSC (Dice Similarity Coefficient) Data augmentation : rotation, shifting, scaling, flipping, noise, elastic deformation p : Predictions, r : Gold standard * 𝜖 term is used to ensure the loss stability by avoiding the numerical issues.
- 9. 2. Methods and Materials 9 Data sets (1) Spinal Cord Gray Matter Challenge • 80 healthy subjects (20 subjects from 4 centers) • 3 different MRI systems (Philips Achieva, Siemens Trio, Siemens Skyra) • Training set : 40 subjects, Test set : 40 subjects
- 10. 2. Methods and Materials 10 Data sets (2) Ex vivo high-resolution spinal cord • An entire human spinal cord • 7T horizontal-bore small animal MRI system • 4676 axial slices with 100 𝜇𝑚 resolution • (120 hours to take..!)
- 11. 2. Methods and Materials 11 Training Protocol (Spinal Cord Gray Matter Challenge dataset) • Resampling and cropping : 0.25 x 0.25 mm (resampled) / 200 x 200 pixels (cropped) • Normalization : mean centering, SD normalization • Train/validation set split : 32 subjects / 8 subjects • Batch size = 11 (samples), 1000 epochs (32 batches at each epoch) • Optimization : Adam optimizer (Learning rate : 0.001) • Batch Normalization : momentum ∅ = 0.01 • Dropout : rate of 0.4 • Learning rate scheduling (𝜂 𝑡0 : initial learning rate, N : the number of epochs, p = 0.9)
- 12. 3. Results 12 Spinal Cord Gray Matter Challenge
- 13. 3. Results 13 Spinal Cord Gray Matter Challenge
- 14. 3. Results 14 Ex vivo high-resolution spinal cord U-Net (비교용) • 14 layers • 3x3 2D convolution filters • Same training protocol & loss
- 15. 3. Results 15 Ex vivo high-resolution spinal cord
- 16. 4. Discussion 16 They devised a simple, but efficient and end-to-end method. Compared to U-Net, proposed method provides better results in many metrics. A major parameter reduction (more than 6 times) IU : Intersection over Union
- 17. H-DenseUNet: Hybrid Densely Connected UNet for Liver and Liver Tumor Segmentation from CT Volumes September 21, 2017
- 18. 1. Introduction 18 Purpose : Liver and liver tumor segmentation in contrast-enhanced 3D abdominal CT scans Liver and liver tumor segmentation can assist doctors in making accurate hepatocellular carcinoma evaluation and treatment planning. Automatic segmentation is a very challenging task. • Tumor : Various size, shape, location and numbers within one patient • Not clear boundaries • CT voxel spacing ranges from 0.45mm to 6.0mm
- 19. 2. Method 19 Proposed method : A novel hybrid densely connected UNet (H-DenseUNet) U-Net U-Net: Convolutional Networks for Biomedical Image Segmentation ( https://arxiv.org/abs/1505.04597 ) To aggregate semantic information To recover the spatial information with the help of shortcut connections
- 20. 2. Method 20 Proposed method : A novel hybrid densely connected UNet (H-DenseUNet) 2D DenseUNet
- 21. 2. Method 21 Proposed method : A novel hybrid densely connected UNet (H-DenseUNet) DenseUNet
- 22. 2. Method 22 Proposed method : A novel hybrid densely connected UNet (H-DenseUNet) H-DenseUNet : 2D DenseUNet + 3D DenseUNet
- 23. 3. Experiment and Results 23 Data set • MICCAI 2017 LiTS (Liver Tumor Segmentation) Challenge • Contrast-enhanced 3D abdominal CT scans • 131 for training, 70 for testing Evaluation Metrics • Dice per case score : An average Dice per volume score • Dice global score : Dice score evaluated by combining all datasets into one
- 24. 3. Experiment and Results 24 Results
- 25. 3. Experiment and Results 25 Results
- 26. Automatic Myocardial Segmentation by Using A Deep Learning Network in Cardiac MRI August 24, 2017
- 27. 1. U-Net 27 Purpose : Myocardial (심근) segmentation in Cardiac MRI Myocardial motion is useful in the evaluation of regional cardiac functions.
- 28. 2. Results 28
- 29. 2. Results 29