Data and Code – Kimia Lab

Data Sets and Source Codes

A Cytology work: “ROI detection” and “Cell Detection and Classification” in bone marrow cytology. July 2021
Yottixel: In this study, we report the results from searching the largest public repository (The Cancer Genome Atlas, TCGA) of whole-slide images. We successfully indexed and searched almost 30,000 high-resolution digitized slides. The proposed method achieved high accuracy values +90% for many cancer types. March 2021
KIMIA Path24C Image Dataset (24 scans, almost 30,000 images of size 1000×1000). February 2021
KimiaNet is a histopathology deep feature extractor trained from scratch with more than 240,000 image patches of 1000×1000 pixels. These images were acquired at 20× magnification through our proposed “high-cellularity mosaic” approach to enable the usage of weak labels of 7,126 FFPE whole slide images, spanning 30 primary diagnoses, obtained from the TCGA repository. January 2021
CLAW Stain NormalizationThis study introduces a novel approach for stain normalization based on learning a mixture of multivariate skew-normal distributions for stain clustering and parameter estimation alongside a stain transformation technique. November 2020
Representation Learning of Histopathology Images using Graph Neural Networks Representation learning for Whole Slide Images (WSIs) is pivotal in developing image-based systems to achieve higher precision in diagnostic pathology. April 2020
A Comparative Study of U-Net Topologies for Background Removal in Histopathology Images This paper compares the performance of different U-Net architectures for tissue segmentation of histopathology whole slide images by variating their backbones. June 2020
ELP Image Descriptor This code is an approximation for the ELP histogram implemented by Python. Using the Kernel idea has decreased the CPU time considerably. October 2018
KIMIA Path960 Image Dataset (960 patches of size 308×168 from histopathology scans belong to 20 different classes; 48 instances per class). November 2017
Consensus Contouring (500 synthetic prostate images and their ground truth images; every image has 20 contours simulating 20 different users; ideal for experimenting with inter-observer variability). September 2017
Consensus Contouring (Prostate MR images of 15 patients contoured by 5 oncologists; for experimenting with inter-observer variability). September 2017
Radon Barcodes This Matlab function extracts a Radon barcode from an image. September 2015