fitushar/WeaklySupervised-3D-Classification-of-Chest-CT-using-Aggregated-MultiResolution-Segmentation-Feature

Stars
21
Rank 1,078,803 (Top 22 %)
Language
Python
Created over 4 years ago
Updated over 4 years ago

fitushar/WeaklySupervised-3D-Classification-of-Chest-CT-using-Aggregated-MultiResolution-Segmentation-Feature

fitushar

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

This Repo contains the updated implementation of our paper "Weakly supervised 3D classification of chest CT using aggregated multi-resolution deep segmentation features", Proc. SPIE 11314, Medical Imaging 2020: Computer-Aided Diagnosis, 1131408 (16 March 2020)

3D-Medical-Imaging-Preprocessing-All-you-need

This Repo Will contain the Preprocessing Code for 3D Medical Imaging

Jupyter Notebook

3D-GuidedGradCAM-for-Medical-Imaging

This Repo containes the implemnetation of generating Guided-GradCAM for 3D medical Imaging using Nifti file in tensorflow 2.0. Different input files can be used in that case need to edit the input to the Guided-gradCAM model.

3D-Grad-CAM

This repo contains Grad-CAM for 3D volumes.

Jupyter Notebook

3DUnet_tensorflow2.0

This Repo is for implementation of 3D unet in Tensorflow 2.0v

Brain-Tissue-Segmentation-Using-Deep-Learning-Pipeline-NeuroNet

This Repository is for the MISA Course final project which was Brain tissue segmentation. we adopt NeuroNet which is a comprehensive brain image segmentation tool based on a novel multi-output CNN architecture which has been trained and tuned using IBSR18 dataset

Jupyter Notebook

Skin-lesion-Segmentation-using-grabcut

Skin lesion segmentation is one of the first steps towards automatic Computer-Aided Diagnosis of skin cancer. Vast variety in the appearance of the skin lesion makes this task very challenging. The contribution of this paper is to apply a power foreground extraction technique called GrabCut for automatic skin lesion segmentation in HSV color space with minimal human interaction. Preprocessing was performed for removing the outer black border. Jaccard Index was measured to evaluate the performance of the segmentation method. On average, 0.71 Jaccard Index was achieved on 1000 images from ISIC challenge 2017 Training Dataset.

Study-of-Low-dose-to-High-dose-CT-using-Supervised-Learning-with-GAN-and-Virtual-Imaging-Trials

Computed tomography (CT) is one of the most widely used radiography exams worldwide for different diagnostic applications. However, CT scans involve ioniz- ing radiational exposure, which raises health concerns. Counter-intuitively, low- ering the adequate CT dose level introduces noise and reduces the image quality, which may impact clinical diagnosis. This study analyzed the feasibility of using a conditional generative adversarial network (cGAN) called pix2pix to learn the mapping from low dose to high dose CT images under different conditions. This study included 270 three-dimensional (3D) CT scan images (85,050 slices) from 90 unique patients imaged virtually using virtual imaging trials platform for model development and testing. Performance was reported as peak signal-to-noise ra- tio (PSNR) and structural similarity index measure (SSIM). Experimental results demonstrated that mapping a single low-dose CT to high-dose CT and weighted two low-dose CTs to high-dose CT have comparable performances using pix2pix CGAN and applicability of using VITs

3DCNNs_TF2Modelhub

Almost all the deeplearning libraries provide ready to use 2D models with/without imagenet weights, But In the case of 3D, CNN models are not as available. This repo will contain commonly used 2D CNNs 3D implementations.

multi-label-weakly-supervised-classification-of-body-ct

A rule-based algorithm enabled the automatic extraction of disease labels from tens of thousands of radiology reports. These weak labels were used to create deep learning models to classify multiple diseases for three different organ systems in body CT.

3D-Attention-in-tf2--Position-Channel-attention

This repo contains the 3D implementation of the commonly used attention mechanism for imaging.

Quantification-of-Trabeculae-Inside-the-Heart-from-MRI-Using-Fractal-Analysis

In this project, we demonstrate the fractal dimension (FD) and manual segmentation of the Magnetic Resonance Imaging (MRI) of the heart that quantify amount of trabeculae inside the heart.

multi-label-annotation-text-reports-body-CT

There is progress to be made in building artificially intelligent systems to detect abnormalities that are not only accurate but can handle the true breadth of findings that radiologists encounter in body (chest, abdomen, and pelvis) Computed Tomography (CT). Currently, the major bottleneck for developing multi-disease classifiers is a lack of manually annotated data. The purpose of this work was to develop high throughput multi-label annotators for body CT reports that can be applied across a variety of abnormalities, organs, and disease states thereby mitigating the need for human annotation.

SURGERY-ASSISTED-BY-COMPUTER-AND-MEDICAL-ROBOTICS-

This Repository contains all the work done for MAIA 3rd Semester Coursework of SURGERY ASSISTED BY COMPUTER AND MEDICAL ROBOTICS

Registration-as-Data-Augumentation-for-CT--DATA

Main idea of this project was making some new data using simple registration technique from existing CT data. This code be a possible alternative of arbitrary augmentation such as flip, rotation, zoom etc.

Cyclical-Learning-Rates-for-Training-Neural-Networks-With-Unbalanced-Data-Sets

As the learning rate is one of the most important hyper-parameters to tune for training convolutional neural networks. In this paper, a powerful technique to select a range of learning rates for a neural network that named cyclical learning rate was implemented with two different skewness degrees. It is an approach to adjust where the value is cycled between a lower bound and upper bound. CLR policies are computationally simpler and can avoid the computational expense of fine tuning with fixed learning rate. It is clearly shown that changing the learning rate during the training phase provides by far better results than fixed values with similar or even smaller number of epochs.

Jupyter Notebook

Automatic_Breast_Region_Extraction_using_python

This Repo will contain Python Implementation of an Automatic approach to Extraction Breast Region from Memogram

Jupyter Notebook

DenseVNet3D_Chest_Abdomen_Pelvis_Segmentation_tf2

This Repo containes the implemnetation of DenseVent in tensorflow 2.0 for chest-abdomen-pelvis (CAP) Segmentation

Image_pixel_recovery_with_lesso_regression

The objective of this mini-project is to Recover a full image from a small number of sampled pixels (compressed sensing). Although the primary goal of this project is to understand and explore the application of regularized. In the process of recovering image pixel using regularized regression, we will explore different concepts and their understanding as following: Understanding how regression can be applied in 2D image analysis domain. Understanding of the discrete cosine transforms (DCT) to define an image in a frequency domain. Explore the importance and application of cross validation in model tunning and hyper-parameter selections. Understanding the impact of applying filtering approach such as median filter on reconstructed image Finally, quantitively evaluating the quality of removed image.

Luna16_Monai_Model_XAI_Project

Brain-Tumour-CAD-Lecture-Activity

Student Lecture Activity for CAD lecture

Design-an-inverse-kinematic-controller-to-move-end-effector-

Design an inverse kinematic controller to move end-effector. This work is done as a coursework for the 2nd semester of MAIA in Introduction to Robotics course.

Brain-Tissue-Segmentation-Using-Expectation-Maximization

Medical Image Segmentation and Applications (MISA) LAB task.

Classification-of-chest-CT-using-caselevel-weak-supervision

Classification of chest CT using caselevel weak supervision

SeamCarving_Content-Aware-Image-Resizing

The availability of sophisticated source attribution techniques raises new concerns about privacy and anonymity of photographers, activists, and human right defenders who need to stay anonymous while spreading their images and videos. An image can be considered to be a combination of both significant (foreground) objects and some less significant (background) objects. Content aware image resizing (CAIR) algorithm uses the different edge detection methods to segregate the useful objects from the background. When applied to an image, CAIR can resize the image to a very different aspect ratio without destroying the aspect ratio of the useful objects in the image. In this project, we simply implement a content aware image resizing (CAIR) in MATLAB environment. The main idea to implement CAIR is to remove or insert the vertical or horizontal seams (paths of pixel) having the lowest energy. After implanted the Seam Carving Algorithm for Content aware image resizing (CAIR), analysis shows that the implemented seam carving for CAIR can generate more desirable resized images than cropping, resampling, and conventional seam carving techniques.