교수자 소개
-
박상현 교수
- DGIST 로봇공학전공 조교수 (2017.02~)
- 스탠포드연구센터 (SRI International) 포닥 (2016.03~2017.02) 
- 노스캐롤라이나 대학 (UNC at Chapel Hill) 포닥 (2014.03~2016.02) 
- 서울대학교 전기컴퓨터공학부 박사 (2008.03~2014.02)
- https://sites.google.com/view/mispl/members/professor
강의계획
강의목록
-
1. Introduction to medical image analysis
- 1.Overview
- 2.Introduction to medical image analysis 1
- 3.Introduction to medical image analysis 2
- 4.PACS/DICOM/Visualization
- 5.Image acquisition
- 6.X-ray / CT / PET
- 7.Magnetic Resonance Imaging (MRI)
- Quiz 1
-
2. Medical image classification(1)
- 1.Introduction to medical image classification
- 2.Linear Regression
- 3.Logistic Regression
- 4.Neural Network
- 5.Image Classification
- 6.Medical image classification
- 7.Classification with demographic scores
- Quiz 2
-
3. Medical image classification(2)
- 1.Property of Deep Neural Network
- 2.Convolution
- 3.Convolutional Neural Network (CNN)
- 4.Advanced CNNs (LeNet, AlexNet, VGG)
- 5.Advanced CNNs (ResNet, InceptionNet, DenseNet)
- 6.3D CNN with demographic scores
- Quiz 3
-
4. Medical image classification(3)
- 1.Overall procedure
- 2.Validation
- 3.Overfitting / Regularization
- 4.Transfer Learning
- 5.Data Augmentation
- 6.Evaluation of classification model
- 7.Evaluation of classification model (Multi-label)
- Quiz 4
-
5. Medical image classification(4)
- 1.Feature selection using L1 regularization
- 2.Feature selection using Entropy / Mutual information
- 3.Feature extraction using Deep Learning
- 4.Class Activation Map
- 5.Weekly supervised learning
- 6.Multiple instance learning
- Quiz 5
-
6. Medical image segmentation(1)
- 1.Introduction to medical image segmentation
- 2.Otsu thresholding
- 3.Morphological processing
- 4.Region growing / Watershed algorithm
- 5.Segmentation using graph model
- 6.Graph cut optimization
- Quiz 6
-
7. Medical image segmentation(2)
- 1.Active Contour Model
- 2.Atlas based method / Label fusion
- 3.Segmentation via learning based method
- 4.Principle Component Analysis (PCA)
- 5.Active shape model
- 6.Segmentation using classifier
- Quiz 7
-
8. Medical image segmentation(3)
- 1.Fully Convolution Network(FCN)
- 2.U-net
- 3.Dilated Convolution
- 4.DeepLab V3+
- 5.Segmentation using 3D CNN
- 6.Loss Function
- 7.Segmentation Metric
- Quiz 8
-
9. Medical image Enhancement(1)
- 1.Introduction to medical image enhancement
- 2.Intensity normalization
- 3.Histogram equalization
- 4.Histogram Matching
- 5.Spatial Filtering
- 6.Anisotropic diffusion filtering
- 7.Vessel enhancement filtering
- Quiz 9
-
10. Medical image Enhancement(2)
- 1.Filtering in frequency domain
- 2.Filtering in 2D frequency domain
- 3.Spatial domain vs Frequency domain
- 4.Non-Local Mean denoising
- 5.Denoising with Dictionary
- 6.Dictionary Learning
- 7.Super-resolution via dictionary learning
- Quiz 10
-
11. Medical image Enhancement(3)
- 1.SRCNN
- 2.Upsampling strategy
- 3.Deep networks for super resolution
- 4.Generative Adversarial Network(GAN)
- 5.SRGAN
- 6.CNN for medical image enhancement
- 7.Enhancement metric
- Quiz 11
-
12. Medical image registration(1)
- 1.Introduction to medical image registration
- 2.Overview
- 3.Transformation Matrix in 2D
- 4.Transformation Matrix in 3D
- 5.Backward warping
- 6.Interpolation
- 7.Similarity measure – SSD, SAD, NCC
- 8.Similarity measure – Mutual information
- Quiz 12
-
13. Medical image registration(2)
- 1.Registration types
- 2.Registration using main axis
- 3.Iterative Closest Point (ICP)
- 4.Nonrigid registration via ICP
- 5.Nonrigid registration via B-spline
- 6.Nonrigid registration via deformable model
- Quiz 13
-
14. Medical image registration(3)
- 1.Optical flow / FlowNet
- 2.Data augmentation for optical flow
- 3.3D image registration via CNN
- 4.Spatial Transformer Network
- 5.3D image registration via unsupervised learning
- 6.Registration metric
- Quiz 14
-
강의평가
- 강의평가