The international society for optics and photonics, along with the American Association of Physicists in Medicine (AAPM) and the National Cancer Institute (NCI), conducts its third Grand Challenge held in conjunction with SPIE Medical Imaging Symposium
BELLINGHAM, Washington, USA and CARDIFF, UK - The Grand Challenge, conducted by SPIE in conjunction with the American Association of Physicists in Medicine (AAPM) and the National Cancer Institute (NCI), focuses on the development of automated methods for analyzing tumor-burden assessment in breast pathology using histology patches extracted from whole slide images. The initial training data was released today, with subsequent releases of validation and test data sets, with and without truth, scheduled prior to the submission deadline of 20 December, 2018. The BreastPathQ Cancer Cellularity Challenge will provide a unique opportunity for participants to compare their algorithms with those of others from academia, industry, and government in a structured, direct way using the same data sets.
The 2019 SPIE Medical Imaging Symposium - set for 16-21 February in San Diego - will showcase the results of this algorithm-scoring challenge in breast pathology. Participants are asked to provide a floating-point score between 0 and 100% for each whole slide image (WSI) patch where the score represents the percentage of cellularity within the tumor bed of the patch. The training/initial validation set consists of 2,579 patches of varying cellularity with final algorithm performance based on scores assigned to an independent test set containing 1,121 patches. Participants are also required to submit a two-page description of their proposed solution describing the basic algorithm structure and training paradigm.
This Challenge highlights the growing application of artificial intelligence (AI) in cancer diagnostics and tumor recognition. Currently, assessment of cellularity in pathology analysis is performed manually; a robust, fully-automated method would mark a critical step forward for the quantitative estimation task, reducing observer variability while addressing issues around reproducibility when it comes to cancer-cellularity assessments in digital pathology.
"This is an exciting new Grand Challenge opportunity for the AI and medical imaging communities to participate in specifically because it is looking at a unique aspect of medical imaging," noted Nicholas Petrick of the U.S. Food and Drug Administration. "In this case, the focus is on developing and accessing cancer cellularity, a quantitative estimation task in digital pathology. This Challenge tackles a difficult estimation task for pathologists and both the algorithms and assessment are quite distinct from binary or multi-classes diagnostic AI tools."
The three top-performing teams of the challenge will receive a registration-fee waiver and invitation to present their methods during a joint computer-aided diagnosis (CAD) and digital pathology session at the SPIE Medical Imaging Symposium in February.
The BreastPathQ Cancer Cellularity Challenge is the third such challenge led by SPIE, AAPM, and NCI, and showcased at SPIE Medical Imaging Symposiums, following 2015's LUNGx Challenge, which tasked participants to classify lung nodules on diagnostic computed tomography (CT) scans as benign or malignant, and the two-part PROSTATEx in 2017, which focused on the development of quantitative image-analysis methods for the diagnostic classification of clinically significant prostate lesions, and quantitative MRI biomarkers used to determine Gleason Grade Group in prostate cancers. Special thanks to Sunnybrook Health Sciences Centre through funding from the Canadian Cancer Society for making the data available for the BreastPathQ Cancer Cellularity challenge.
For more information, access to the training data, and to participate in the BreastPathQ Cancer Cellularity Challenge, please visit: https://spie.org/conferences-and-exhibitions/medical-imaging/challenge-2019
SPIE is the international society for optics and photonics, an educational not-for-profit organization founded in 1955 to advance light-based science, engineering, and technology. The Society serves nearly 264,000 constituents from approximately 166 countries, offering conferences and their published proceedings, continuing education, books, journals, and the SPIE Digital Library. In 2017, SPIE provided more than $4 million in support of education and outreach programs. www.spie.org.
Public Relations Manager
+1 360 685 5478