Code supporting our paper DataCurator.jl: Efficient, portable, and reproducible validation, curation, and transformation of large heterogeneous datasets using human-readable recipes compiled into machine verifiable templates
Reference Julia implementation of colocalization metrics for 2D and 3D (microscopy) images
Code supporting our paper Log-Paradox: Necessary and sufficient conditions for confounding statistically significant pattern reversal under the log-transform
Code supporting our two papers: MedIA - Guidelines and evaluation of clinical explainable AI in medical image analysis; and AAAI: Evaluating Explainable AI on a Multi-Modal Medical Imaging Task: Can Existing Algorithms Fulfill Clinical Requirements?
Code supporting our paper: Mouth2audio: Intelligible Audio Synthesis from Videos with Distinctive Vowel Articulation
Active Learning from Noisy Teacher
Code supporting our CMIG paper: Active Deep Learning from a Noisy Teacher for Semi-supervised 3D Image Segmentation: Application to COVID-19 Pneumonia Infection in CT
Code supporting our NeurIPS paper MaskTune: Mitigating Spurious Correlations by Forcing to Explore
(Saeid Asgari, Aliasghar Khani)
Learning to Segment from Noisy Annotations
Code supporting our MICCAI MIL3ID paper on Learning to Segment Skin Lesions from Noisy Annotations
Code supporting our MICCAI 2018 paper on star shape prior for CNN based image segmentation
Code supporting our work on SPECHT: Self-tuning Plausibility Based Object Detection Enables Quantification of Conflict in Heterogeneous Multi-scale Microscopy
Code supporting our work on automatic sub-precision membrane contact site detection
Code supporting our ECCV ISIC paper CIRCLe: Color Invariant Representation Learning for Unbiased Classification of Skin Lesions
Code supporting our ECCV ISIC paper, FairDisCo: Fairer AI in Dermatology via Disentanglement Contrastive Learning
Code supporting our ICCV VRMI paper on WhiteNNer-Blind Image Denoising via Noise Whiteness Priors
Code supporting our AAAI paper on Evaluating Explainable AI on Multi-Modal Medical Imaging Task
Code supporting our Medical Image Analysis paper: Skin3D: Detection and Longitudinal Tracking of Pigmented Skin Lesions in 3D Total-Body Textured Meshes
(Mengliu Zhao, Jeremy Kawahara, Kumar Abhishek, and Sajjad Shamanian)
Code supporting our IEEE TMI paper: ERGO: efficient recurrent graph optimized emitter density estimation in single molecule localization microscopy.
Learning-to-Augment Noisy & Denoised Data
Code supporting our paper Computers in Biology and Medicine 2021 paper: Learning-to-Augment Strategy using Noisy and Denoised Data: Improving Generalizability of Deep CNN for the Detection of COVID-19 in X-ray Images
Code supporting our papers CMIG2021, MICCAI2019, MICCAI MLMI 2019 for 2D ImHistNet and 3D ImHistNet - Learnable Image Histogram-based DNN.
Generalizable Feature Learning with Class Imbalance
Code supporting our MICCAI 2019 paper: Generalizable Feature Learning in the Presence of Data Bias and Domain Class Imbalance: Application to Skin Lesion Classification
A loss function for training deep segmentation models based on the Matthews Correlation Coefficient (supporting our ISBI 2021 paper)
Super-Resolution NETwork Analysis, SuperResNET, is an integrated software for the analysis of 3D single molecule localization microscopy (SMLM) point cloud data. It consists of computational modules to read, pre-process, post-process, quantify, and visualize 3D SMLM data. SuperResNET allows the user to extract, visualize, and analyze biological clusters. SuperResNET is largely based on the original Scientific Reports 2018 and Scientific Reports 2019 works.
Scanner Invariant Deep Learning Segmentation
Code for ISBI 2020 paper: Scanner Invariant Multiple Sclerosis Lesion Segmentation from MRI (video)
Code for CVPR 2019 paper: A Kernelized Manifold Mapping to Diminish the Effect of Adversarial Perturbations
(Saeid Asgari and Kumar Abhishek)
Code for CMIG 2019 paper: Combo loss: Handling input and output imbalance in multi-organ segmentation
Missing MRI Pulse Sequence Synthesis using Multi-Modal Generative Adversarial Network
Illumination-based Transformations for Skin Lesion Images
Generating transformations based on illumination information and color imaging of the skin for RGB skin lesion images.
Superresolution visualization of 3D protein localization data from a range of super-resolution microscopes
Code for AECNN: Adversarial and Enhanced CNN for data-efficient segmentation of gastrointestinal polyps from colonoscopy images
Code for ISBI 2018 paper: Generative adversarial networks to segment skin lesions
Super-Resolution via Bilinear Pooling
Code for MICCAI 2019 MLMIR: Image Super Resolution via Bilinear Pooling: Application to Confocal Endomicroscopy
Super-Resolution (Confocal Laser Endomicroscopy)
Code for MICCAI 2018: Can Deep Learning Relax Endomicroscopy Hardware Miniaturization Requirements?
Trained CNN to detect four types of dermoscopic criteria based on our winning entry for Part 2 of the 2017 ISIC Skin Challenge
SMLM PC3 Cav1/CAVIN1 Point Cloud Data
3D point clouds of super resolution molecule localization microscopy (SMLM) of Cav1 protein in prostate cancer (PC3) and CAVIN1/PTRF transfected PC3 (PC3-PTRF) cells
Melanoma Recognition via Visual Attention
Attention-based method for melanoma recognition, with attention map regularization
(Yiqi Yan and Jeremy Kawahara)
Local Synthetic Instances (LSI)
An oversampling method for synthetic data generation or similar tasks. It is designed for the challenging case of high-dimensional, non-gaussian data with low sample size
Recurrent Visual Attention Model for Analyzing Histopathology Whole Slide Images
Highlight vascular structures exhibiting radial distension pulsatile motions
Take a field of diffusion orientation distribution functions (ODF) and map them to a graph representation
Take a diffusion MRI scan and generate a graph representation, combined with Dijkstra’s algorithm to perform minimal path tractography.
MATLAB toolbox to highlight corners, tubular structures, and sheet-like structures in diffusion tensor images.
(Brian Booth & Krishna Nand)
MATLAB code for segmenting cervical cells, based on "A Variational Approach for Overlapping Cell Segmentation"
Image Data Augmentation Tool: Simulate novel images with ground truth segmentations from a single image-segmentation pair
MATLAB viewer for 3D scalar, vector, and tensor-valued medical images
Software library for manipulating multi-region, probabilistic shapes using Aitchison geometry
Simulate 3D images of vascular trees based on oxygen demand maps and physical parameters
Perception-based visualization of manifold-valued medical images using distance preserving dimensionality reduction
ITK and ITK Image IO on Apple iOS
Documentation and code for running ITK on iOS devices (iPod touch, iPhone, iPad)
(Boris Shabash & Zhi Feng Huang)
Intuitive and efficient vessel segmentation tool, finds centre-line and boundaries simultaneously
(Miranda Poon and Ryan Dickie)
n-dimensional scale invariant feature transform (SIFT)(3D SIFT, 4D SIFT, etc.)
Matlab GUI demo for shape representation & (statistics- and operator-based) deformations