Package gov.nih.mipav.view.renderer.WildMagic.ProstateFramework
package gov.nih.mipav.view.renderer.WildMagic.ProstateFramework
-
ClassesClassDescriptionThe class is the driver for the AAM classification.The class is the driver for the AAM classification.This class is the combined Atlas based AAM and SVM model to automatically segment the MRI prostate.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.For Miccai promise 12 prostate data, we apply wp segmentation using HED deep learning model.This class convert the 3D images to 2D slices based atlas.This is the first try to 3D convolution deep learning models. 3D-Unet.This class simply convert the prostate images into isotropic (same x, y resolution ) images.Atfer N4 correction, take those images with binary masks, transform again to 0.5mm x 0.5mm x 1.5 mm images.Atfer N4 correction, take those images with binary masks, transform again to 0.5mm x 0.5mm x 1.5 mm images.After 0.5mm x 0.5mm x 1.5 mm transform, apply the intensity normalization [0, 1000], save image and masks.After 0.5mm x 0.5mm x 1.5 mm transform, apply the intensity normalization [0, 1000], save image and masks.This is the first try to 3D convolution deep learning models. 3D-Unet.This is the first try to 3D convolution deep learning models. 3D-Unet.This is the first try to 3D convolution deep learning models. 3D-Unet.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.This class generates the VOIs from HED prediction maps.This class generates the VOIs from HED prediction maps.This class generates the VOIs from HED prediction maps.This class generates the VOIs from HED prediction maps.This class convert the 3D images to 2D slices based atlas.This class copy ground truth stl file to destination directory.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.For ISBI 2017 paper: Data given: Dr.This class attempts to generate the VOI from the HED predicted edge map (boundary map).For ISBI 2017 paper: ISBI, we only use MR images slices, no CED involved.For ISBI 2017 paper: This class converts the HED generated energy maps into VOI contours.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.This class is the trial-and-error approach for HED deep learning experiment.This class convert the 3D images to 2D slices based atlas.This class is the trial-and-error approach for HED deep learning experiment.This class is the trial-and-error approach for HED deep learning experiment.This class is the trial-and-error approach for HED deep learning experiment.This class is the trial-and-error approach for HED deep learning experiment.This is the trial-and-error test, which applying the NIH data for training, Prostate promise 12 data for testing.This is the first attempt to apply prostate segmentation on Promise 12 data.This is the third attempt to apply prostate segmentation on Promise 12 data.This is the third attempt to apply prostate segmentation on Promise 12 data.This is the third attempt to apply prostate segmentation on Promise 12 data.This is the third attempt to apply prostate segmentation on Promise 12 data.This is the third attempt to apply prostate segmentation on Promise 12 data.This is the second attempt to apply prostate segmentation on Promise 12 data.This class is for SPIE paper: HED prostate MRI segmentation.For JMI 2017 paper: this class uses image processing methods that are different to the SPIE paper. 1) Read the MRI image with VOIs 2) Generate the binary masks 2) Crop the image with the 25% deduction 3) Scale image intensity to range 0 to 1000. 4) Generate the CED image. 5) Extract MRI image and CED image slices with corresponding binary image masks for HED to train the deep learning model.This is for the SPIE paper: HED prostate segmentation.For JMI 2017 paper: this class uses image processing methods that are different to the SPIE paper. 1) Read the MRI image with VOIs 2) Generate the binary masks 2) Crop the image with the 25% deduction 3) Scale image intensity to range 0 to 1000. 4) Generate the CED image. 5) Extract MRI image and CED image slices with corresponding binary image masks for HED to train the deep learning model.For rest of the papers: this class generates the training list from the training fold. 5 fold cross-validation, each training fold contains the png slices from all other folds.The class converts the 2D-volumetric approach axial, sagittal and coronal MRI and CED png slices into a file list.This class simply converts the MRI and CED slices png files into a list file.The class generates the png file list for test cases.The class converts the MRI and CED png slices into a file list.The class reconstructs the 3D surface from the axial, coronal, sagittal VOI contours.For the Miccai promise 12 data, this class simply read the png files and generates the png file list for HED to train the deep learning model.The class reconstruct the 3D prostate surface from the Deep learning generated VOIs.This class convert the 3D images to 2D slices based atlas.The class reads the nii binary masks and generates the comparison list for shell script.This class generates the EvaluateSegmentation shell script to evaluate the binary masks between AAM vs.After the 3D images convert to 2D slices ( 512x512 ), this class picks the 2D slices, extracts the Coherence Enhanced diffusion based features, and saves those features with linear SVM readable file formats.This class exhaustively trains the 2D slices based Active Appearance Model (AAM).This class converts 3D prostate surface into VOIs; saving them for comparison.This class converts 3D prostate surface into nii binary mask images.This class read nii binary mask files, convert them into a shell script list.For testing phase: This class reads the original MRI images and HED deep learning model predicted MRI and CED energy map results, generates the final VOI contours.For the JMI paper, this class converts VOI contours back to binary image masks.This class convert the 3D images to 2D slices based atlas.This class convert the 3D images to 2D slices based atlas.Semi-automatic MR Prostate segmentation - Registration and Fuzzy-C guided segmentation model.Semi-automatic MR Prostate segmentation - Registration guided segmentation model.This class is for ProstateX 2017 MICCAI challenge.This class generates the VOI contours from HED predicted energy maps.For MICCAI ProstateX 2017 challenge testing cases, this file generates the VOI contours from HED predicted MRI+CED energy maps.The MICCAI ProstateX challenge is the first try on MRI image alone.After the MICCAI ProstateX challenge, I try to apply the prostate boundary based training on HED model.The MICCAI ProstateX challenge is rolling over to prostate segmentation, which includes the whole prostate(wp) and central gland(cg).The MICCAI ProstateX challenge is rolling over to prostate segmentation, which includes the whole prostate(wp) and central gland(cg).The MICCAI ProstateX challenge is rolling over to prostate segmentation, which includes the whole prostate(wp) and central gland(cg).The MICCAI ProstateX challenge is rolling over to prostate segmentation, which includes the whole prostate(wp) and central gland(cg).For NIH prostate data, we apply wp and cg segmentation using HED deep learning model.This class simply read the MR, CED images with corresponding binary image masks, converting them into image pairs ( MR slice with binary mask, CED slice with binary mask ).This class simply shuffle the H5 list randomly.Class PlaneRenderWM: renders a single dimension of the ModelImage data as a texture-mapped polygon.Copyright (c) 2000-2014 Chih-Chung Chang and Chih-Jen Lin All rights reserved.Copyright (c) 2000-2014 Chih-Chung Chang and Chih-Jen Lin All rights reserved.Copyright (c) 2000-2014 Chih-Chung Chang and Chih-Jen Lin All rights reserved.Copyright (c) 2000-2014 Chih-Chung Chang and Chih-Jen Lin All rights reserved.