Journal Article
. 2020 Mar;4().
doi: 10.1200/CCI.19.00068.

Quantitative Assessment of the Effects of Compression on Deep Learning in Digital Pathology Image Analysis

Yijiang Chen 1 Andrew Janowczyk 1 Anant Madabhushi 1 
  • PMID: 32155093
  •     16 References


Purpose: Deep learning (DL), a class of approaches involving self-learned discriminative features, is increasingly being applied to digital pathology (DP) images for tasks such as disease identification and segmentation of tissue primitives (eg, nuclei, glands, lymphocytes). One application of DP is in telepathology, which involves digitally transmitting DP slides over the Internet for secondary diagnosis by an expert at a remote location. Unfortunately, the places benefiting most from telepathology often have poor Internet quality, resulting in prohibitive transmission times of DP images. Image compression may help, but the degree to which image compression affects performance of DL algorithms has been largely unexplored.

Methods: We investigated the effects of image compression on the performance of DL strategies in the context of 3 representative use cases involving segmentation of nuclei (n = 137), segmentation of lymph node metastasis (n = 380), and lymphocyte detection (n = 100). For each use case, test images at various levels of compression (JPEG compression quality score ranging from 1-100 and JPEG2000 compression peak signal-to-noise ratio ranging from 18-100 dB) were evaluated by a DL classifier. Performance metrics including F1 score and area under the receiver operating characteristic curve were computed at the various compression levels.

Results: Our results suggest that DP images can be compressed by 85% while still maintaining the performance of the DL algorithms at 95% of what is achievable without any compression. Interestingly, the maximum compression level sustainable by DL algorithms is similar to where pathologists also reported difficulties in providing accurate interpretations.

Conclusion: Our findings seem to suggest that in low-resource settings, DP images can be significantly compressed before transmission for DL-based telepathology applications.

Telepathology Impacts and Implementation Challenges: A Scoping Review.
Julien Meyer, Guy Paré.
Arch Pathol Lab Med, 2015 Dec 01; 139(12). PMID: 26619028
Voronoi-Based Curvature and Feature Estimation from Point Clouds.
Quentin Mérigot, Maks Ovsjanikov, Leonidas Guibas.
IEEE Trans Vis Comput Graph, 2010 Dec 15; 17(6). PMID: 21149883
A Practical Guide to Whole Slide Imaging: A White Paper From the Digital Pathology Association.
Mark D Zarella, Douglas Bowman, +6 authors, Douglas J Hartman.
Arch Pathol Lab Med, 2018 Oct 12; 143(2). PMID: 30307746
Highly Cited. Review.
Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer.
Babak Ehteshami Bejnordi, Mitko Veta, +67 authors, Rui Venâncio.
JAMA, 2017 Dec 14; 318(22). PMID: 29234806    Free PMC article.
Highly Cited.
Metastasis detection from whole slide images using local features and random forests.
Mira Valkonen, Kimmo Kartasalo, +3 authors, Pekka Ruusuvuori.
Cytometry A, 2017 Apr 21; 91(6). PMID: 28426134
Mosaic-Based Color-Transform Optimization for Lossy and Lossy-to-Lossless Compression of Pathology Whole-Slide Images.
Miguel Hernandez-Cabronero, Victor Sanchez, +3 authors, Joan Serra-Sagrista.
IEEE Trans Med Imaging, 2018 Jul 12; 38(1). PMID: 29994394
Quantitative nuclear histomorphometry predicts oncotype DX risk categories for early stage ER+ breast cancer.
Jon Whitney, German Corredor, +6 authors, Anant Madabhushi.
BMC Cancer, 2018 Jun 01; 18(1). PMID: 29848291    Free PMC article.
Overview of telepathology, virtual microscopy, and whole slide imaging: prospects for the future.
Ronald S Weinstein, Anna R Graham, +7 authors, John R Gilbertson.
Hum Pathol, 2009 Jun 26; 40(8). PMID: 19552937
Deep learning for digital pathology image analysis: A comprehensive tutorial with selected use cases.
Andrew Janowczyk, Anant Madabhushi.
J Pathol Inform, 2016 Aug 27; 7. PMID: 27563488    Free PMC article.
Highly Cited.
Impact of JPEG 2000 compression on deep convolutional neural networks for metastatic cancer detection in histopathological images.
Farhad Ghazvinian Zanjani, Svitlana Zinger, +3 authors, Peter H N de With.
J Med Imaging (Bellingham), 2019 May 01; 6(2). PMID: 31037247    Free PMC article.
Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning.
Hoo-Chang Shin, Holger R Roth, +6 authors, Ronald M Summers.
IEEE Trans Med Imaging, 2016 Feb 18; 35(5). PMID: 26886976    Free PMC article.
Highly Cited.
Optimized JPEG 2000 Compression for Efficient Storage of Histopathological Whole-Slide Images.
Henrik Helin, Teemu Tolonen, +3 authors, Jorma Isola.
J Pathol Inform, 2018 Jun 19; 9. PMID: 29910969    Free PMC article.
A Deep Convolutional Neural Network for segmenting and classifying epithelial and stromal regions in histopathological images.
Jun Xu, Xiaofei Luo, +2 authors, Anant Madabhushi.
Neurocomputing, 2017 Feb 06; 191. PMID: 28154470    Free PMC article.
Highly Cited.
Image analysis and machine learning in digital pathology: Challenges and opportunities.
Anant Madabhushi, George Lee.
Med Image Anal, 2016 Jul 18; 33. PMID: 27423409    Free PMC article.
Highly Cited. Review.
Robotic surgical telepathology between the Iron Mountain and Milwaukee Department of Veterans Affairs Medical Centers: a 12-year experience.
Bruce E Dunn, Hongyung Choi, +2 authors, Benjamin L Wagenman.
Hum Pathol, 2009 Jun 26; 40(8). PMID: 19552935
Mitosis detection in breast cancer pathology images by combining handcrafted and convolutional neural network features.
Haibo Wang, Angel Cruz-Roa, +6 authors, Anant Madabhushi.
J Med Imaging (Bellingham), 2015 Jul 15; 1(3). PMID: 26158062    Free PMC article.