Digital Pathology -- Classification of whole slide tissue images

Organized by cpm.organizing.committee - Current server time: Sept. 26, 2017, 8:55 a.m. UTC

First phase

Training
June 1, 2017, midnight UTC

End

Competition Ends
Aug. 7, 2017, 11:59 p.m. UTC

Overview

Please note the test phase will start on July 20 and will end on August 7 (extended from July 31).

Grading and diagnosis of tumors in cancer patients have traditionally been done by examination of tissue specimens under a powerful microscope by expert pathologists. While this process continues to be widely applied in clinical settings, it is not scalable to translational and clinical research studies involving hundreds or thousands of tissue specimens. State-of-the-art digitizing microscopy instruments are capable of capturing high-resolution images of whole slide tissue specimens rapidly. Computer aided segmentation and classification has the potential to improve the tumor diagnosis and grading process as well as to enable quantitative studies of the mechanisms underlying disease onset and progression.

The objective of the Digital Pathology Challenge is to evaluate and compare classification and segmentation algorithms and to encourage the biomedical imaging community to design and implement more accurate and efficient algorithms. This year’s challenge targets algorithms for whole slide tissue images obtained from patients with non small cell lung cancer (NSCLC), head and neck squamous cell carcinoma (HNSCC), glioblastoma multiforme (GBM), and lower grade glioma (LGG) tumors. These cancer types are complex and deadly diseases accounting for a large number of diagnostic patient deaths in spite of application of various treatment strategies.

The goal of this sub-challenge is to evaluate the performance of automated classification algorithms. There are two cohorts: one from NSCLC patients and one from HNSCC patients. Participants are asked to distinguish between NSCLC adeno and NSCLS squamous cell patients. For the HNSCC cohort, participants will be asked to classify patients into HPV +/- and K17 (mRNA) +/- molecular subtypes.

The whole slide tissue images are stored in Aperio SVS format. There are open source tools and libraries that can read these images: OpenSlide (http://openslide.org) and Bio-Formats (https://www.openmicroscopy.org/site/support/bio-formats5.1/about/index.html).

Evaluation

Please note the test phase will start on July 20 and will end on August 7 (extended from July 31).

Each participant is required to submit an extended abstract (up to 4 pages) that describes their algorithm by the end of the training phase (update: Please email your abstracts by Monday, July 24). You may submit your abstract to tahsin.kurc@stonybrook.edu .

The scoring for this sub-challenge will be computed as the number of correctly classified cases divided by the total number of cases.

Each participant will be required at the end of the test phase to submit a zip file containing tab-delimited text files. Each row of the text file corresponds to a case and stores (case ID, classification label) tuples.

The HNSC classification file must be named hnsc_classification.csv.

The NSCLC classification file must be named lung_classification.csv.

The format of the file for the HNSC classification will be as follows. The file should be a CSV file with each row corresponding to classification of an image. Each row should be formatted as follows:

 
image file name, 1 or 0, 1 or 0
image file name, 1 or 0, 1 or 0
...

1 or 0 in the second column indicates K17 High or K17 Low, respectively:
  
K17 High = 1
K17 Low  = 0

1 or 0 in the third column indicates HPV Positive or HPV Negative, respectively:

HPV Positive = 1
HPV Negative = 0

For example, the classification file in the training set contains the following: 

image1.svs,1,1
image2.svs,1,0
image3.svs,1,0
image4.svs,1,1
image5.svs,1,1
image6.svs,0,0
image7.svs,1,1
image8.svs,1,0
image9.svs,0,0
image10.svs,0,0
image11.svs,0,1
image12.svs,0,0
image13.svs,0,1
image14.svs,1,0
image15.svs,0,1
image16.svs,0,1

The format of the file for classification of Lung images will be a CSV file with each row containing the image file name and the classification label:

image file name, 0 or 1
image file name, 0 or 1
...

0 or 1 in the second column indicates the image is NSCLC adeno or NSCLS squamous cell, respectively. 

For example, the classification file in the training set contains the following: 

image1.svs,0
image2.svs,0
image3.svs,0
image4.svs,0
image5.svs,0
image6.svs,0
image7.svs,0
image8.svs,0
image9.svs,0
image10.svs,0
image11.svs,0
image12.svs,0
image13.svs,0
image14.svs,0
image15.svs,0
image16.svs,0
image17.svs,1
image18.svs,1
image19.svs,1
image20.svs,1
image21.svs,1
image22.svs,1
image23.svs,1
image24.svs,1
image25.svs,1
image26.svs,1
image27.svs,1
image28.svs,1
image29.svs,1
image30.svs,1
image31.svs,1
image32.svs,1

Terms and Conditions

Please note the test phase will start on July 20 and will end on August 7 (extended from July 31).

This page enumerated the terms and conditions of the competition.

No submissions have been made public!

Training

Start: June 1, 2017, midnight

Test

Start: July 20, 2017, midnight

You must be logged in to participate in competitions.

Sign In
No submissions have been made public!
No submissions have been made public!