Journal of Pathology Informatics Journal of Pathology Informatics
Contact us | Home | Login   |  Users Online: 2263  Print this pageEmail this pageSmall font sizeDefault font sizeIncrease font size 

Table of Contents    
J Pathol Inform 2011,  2:25

Computer aided diagnostic tools aim to empower rather than replace pathologists: Lessons learned from computational chess

1 Department of Pathology, University of Michigan, 4233A Medical Science I, 1301 Catherine, Ann Arbor, Michigan, USA
2 Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA
3 Department of Biomedical Engineering, Rutgers The State University of New Jersey, 599 Taylor Road, Piscataway, NJ, USA
4 MGH Pathology Imaging and Communication, Technology (PICT) Center, 101 Merrimac Street, Boston, MA 02114-4719, USA
5 National Institutes of Health, National Cancer Institute, Laboratory of Pathology, Advanced Technology Center, 8717 Grovemont Circle, Gaithersburg, MD 20877, USA
6 Esoterix Genetic Laboratories, LLC, Successor to Genzyme Genetics, Pathology Services, 5300 McConnell Ave, Los Angeles, CA 90066, USA
7 Massachusetts General Hospital, Harvard Medical School, 16th Street Building 114, Charlestown, MA 02129, USA
8 Department of Pathology, Massachusetts General Hospital, Harvard Medical School, 1 Longfellow Place, Suite 216, Boston, MA 02114, USA

Date of Submission17-Mar-2011
Date of Acceptance02-May-2011
Date of Web Publication14-Jun-2011

Correspondence Address:
Ulysses J Balis
Department of Pathology, University of Michigan, 4233A Medical Science I, 1301 Catherine, Ann Arbor, Michigan
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/2153-3539.82050

Rights and Permissions

How to cite this article:
Hipp J, Flotte T, Monaco J, Cheng J, Madabhushi A, Yagi Y, Rodriguez-Canales J, Emmert-Buck M, Dugan MC, Hewitt S, Toner M, Tompkins RG, Lucas D, Gilbertson JR, Balis UJ. Computer aided diagnostic tools aim to empower rather than replace pathologists: Lessons learned from computational chess. J Pathol Inform 2011;2:25

How to cite this URL:
Hipp J, Flotte T, Monaco J, Cheng J, Madabhushi A, Yagi Y, Rodriguez-Canales J, Emmert-Buck M, Dugan MC, Hewitt S, Toner M, Tompkins RG, Lucas D, Gilbertson JR, Balis UJ. Computer aided diagnostic tools aim to empower rather than replace pathologists: Lessons learned from computational chess. J Pathol Inform [serial online] 2011 [cited 2022 Jul 2];2:25. Available from:

   Computer Aided Diagnosis : New Horizons in Diagnostic Pathology Top

The recent availability of digital whole slide imaging (WSI) data sets from glass slides creates new opportunities for possible deployment of computer aided diagnostic (CAD) technologies. [1],[2],[3],[4],[5],[6],[7] Use of CAD has the potential to improve the practice of pathology in various ways by helping the pathologist: in the screening of slides, the provisioning for real time clinical decision support tools, the instantiation of additional automated layers of quality assurance and diagnostic consistency, and lastly, imparting a quantitative component to the practice of diagnostic pathology. [4]

As observed by Thomas Kuhn, scientific revolutions - such as the computational and informatics revolution that characterize our times - can generate fears and resistance from a conservative worldview in the same way the contemporary circumstance of proposed adoption of smart digital pathology technologies, such as CAD, may be seen as a threat that "will replace the human," and in this case, the pathologist. Instead, we would rather believe that properly leveraging the strength of informatics and computation can be of valuable assistance, making pathologists better and more efficient. Akin to Polymerase Chain Reaction (PCR) for the molecular biologist, and immunohistochemistry for the pathologist, CAD technology provides many examples of how widespread adoption will require proper use and clinical interpretation by pathologists. Accordingly, lessons learned from the computational chess community can illustrate how CAD may be incorporated successfully into the practice of pathology.

   'How CAD can be Integrated into the Practice of Pathology' Top

A classic example of how CAD has been applied to pathology practice and be seen with automated screening of  Pap smear More Detailss. CAD technology has been shown to improve efficiency, accuracy and consistency, while freeing skilled pathologists and cytotechnologists from the monotonous and therefore error-prone screening of large numbers of smears. [8] In a similar vein, the evaluation of multiple prostate biopsies for prostate cancer is hampered by & larger number of slides/cores/sections, (potentially) small foci of cancer, and the need for consistency in diagnosis and grading. [4] Like the automated pap screening devices, CAD approaches hold the potential of greatly assisting the reviewing pathologist by pre-screening prostate tissue sections to highlight those areas that are likely to contain cancer and perhaps, as an added metric of performance, contributing towards the goal of attaining greater consistency in grading. [4],[7],[9]

There are other important examples from dermatopathology, infectious disease, and consultative practice. Mitotic figures are often important but not diagnostic of melanoma in biopsies of atypical melanocytic lesions. In those that are less than 1 mm in thickness, the proper mitotic count may signify the difference between requiring and not requiring follow on sentinel lymph node biopsy. For the diagnosis of infectious diseases, we may eventually use microchips that can identify every known pathogen, but in the interim, we remain dependent upon special stains, where potential use of CAD could assist with the rapid identification of rare small organisms, such as mycobacteria or spirochetes. For rare and unusual tumors, CAD could assist those pathologists who have not yet been amassed sufficient experience with certain diagnostic entities, to be able to identify them with confidence. CAD could also assist in the screening of tissue sections, such as those encountered with sentinel lymph nodes, where there is a need to identify micrometastases - a laborious task that is prone to error.

Many challenges in applying CAD to an all digital workflow will be in the transport, storage, retrieval, and analysis of digital slides. The majority of these operational issues can be addressed through the effective use of Grid technologies. [10] Grid computing enables the plurality of connected users to simultaneously access a single web presence, even though the overall application may be implemented as a geographically diverse collection of separate data centers, thus confering both enhanced reliability and response times. [10] Integration of this emerging type of software hosting model will allow those users of ordinary desktop/laptop PCs to perform tasks normally reserved for high-throughput computation settings (enabled by the reality that the actual computation will be taking place within a high-performance data center or centers). Consequently, we envision that the expanding use of automated feature identification applications in digital pathology will usher in another large revolution in the practice of pathology similar to the advent of immunohistochemistry or the use of light microscopy. Such technology will not replace, but rather, assist pathologists with their rendering of the best possible decisions regarding difficult cases.

   Lessons Learned from Chess Top

In his work on "the Cyborg Advantage", Clive Thompson described the different advantages possessed by humans and computers (summarized below). [11] A computer's advantage lies in its ability to perform high-speed calculations, covering millions of possible moves in search of an optimal one, while a human relies on strategy and intuition perfected by years of experience and study. Computational approaches in this regard are thus tactically superior, whereas, human chess performance is strategically superior (

In 1997, IBM's Deep Blue supercomputer beat the international grand champion Garry Kasparov. Afterward, Kasparov created the field known as "advanced chess" - a game format in which players are assisted by off-the-shelf computer chess programs, which serve in the role of offering a selection of possible moves that are based upon the current position of game pieces.

In 2005, Kasparov created an online tournament comprised of grandmasters, supercomputers, and computer-assisted humans. Ironically, the overall winner was a pair of amateur players (less than 30 years old!) who participated in the competition with nothing more than an average personal computer running an inexpensive chess program. Their advantage stemmed from their ability to know when to effectively leverage the computer's assistance, and when to consult and/or ignore the computer's advice. [11] Kasparov later said that a weak human with a machine can outperform a strong human with a machine if the weak human has a better cognitive process. [11] Therefore, the most brilliant chess players, worldwide, are neither high-end machines nor high-end humans but surprisingly, those average-playing-ability chess players who are effective with the blending of their experience with the exhaustive thoroughness as made possible by computational decision support. [11]

Can "advanced chess" offer a new paradigm for the integration of computers into pathology? The decision support approach could in theory effectively serve as potential decision support tools for pathologists upon completion of appropriate optimization and validation. The improving quantitative algorithmic performance of CAD algorithms reported to date gives hope that there will be continued evolutionary improvement in WSI algorithmic performance, leading to highly accurate tools. The greater challenge for the specialty lies in our ability to integrate such tools into our core workflow and practice approach. We believe the experiences of computational chess, as accrued to date, are encouraging for the plausible deployment in our field of what might one day be termed "advanced pathology."

   Disclosure/Conflict of Interest Top

AM and JM are majority stockholders in Ibris Inc.

   Acknowledgments Top

This work was made possible via grants from the National Cancer Institute (Grant Nos. R01CA136535-01, R01CA140772­01, and R03CA143991-01); and the Cancer Institute of New Jersey; and at the University of Michigan by Clinical Translational Science Award (CTSA) 5ULRR02498603 PI Ken Pienta.

   References Top

1.Madabhushi A. Digital Pathology Image Analysis: Opportunities and Challenges. Imaging Med 2009;1:4.  Back to cited text no. 1
2.Doyle S, Feldman M, Tomaszewski J, Madabhushi A. A Boosted Bayesian Multi-Resolution Classifier for Prostate Cancer Detection from Digitized Needle Biopsies. IEEE Trans Biomed Eng 2010 [In Press].  Back to cited text no. 2
3.Fatakdawala H, Xu J, Basavanhally A, Bhanot G, Ganesan S, Feldman M, et al. Expectation-maximization-driven geodesic active contour with overlap resolution (EMaGACOR): application to lymphocyte segmentation on breast cancer histopathology. IEEE Trans Biomed Eng 2010;57:1676-89.  Back to cited text no. 3
4.Basavanhally AN, Ganesan S, Agner S, Monaco JP, Feldman MD, Tomaszewski JE, et al. Computerized image-based detection and grading of lymphocytic infiltration in HER2+ breast cancer histopathology. IEEE Trans Biomed Eng 2010;57:642-53.  Back to cited text no. 4
5.Lexe G, Monaco J, Doyle S, Basavanhally A, Reddy A, Seiler M, et al. Towards improved cancer diagnosis and prognosis using analysis of gene expression data and computer aided imaging. Exp Biol Med (Maywood) 2009;234:860-79.  Back to cited text no. 5
6.Monaco JP, Tomaszewski JE, Feldman MD, Hagemann I, Moradi M, Mousavi P, et al. High-throughput detection of prostate cancer in histological sections using probabilistic pairwise Markov models. Med Image Anal 2010;14:617-29.  Back to cited text no. 6
7.Gurcan MN, Boucheron L, Can A, Madabhushi A, Rajpoot N, Yener B. Histopathological Image Analysis: A Review. IEEE Rev Biomed Eng 2009;2:147-71.  Back to cited text no. 7
8.Bishop JW, Marshall CJ, Bentz JS. New technologies in gynecologic cytology. J Reprod Med 2000;45:701-19.  Back to cited text no. 8
9.Madabhushi A. Digital pathology image analysis: Opportunities and challenges. Imaging Med 2009;1:7-10.  Back to cited text no. 9
10.Kayser K, Gortler J, Borkenfeld S, Kayser G. Grid computing in image analysis. Diagn Pathol 2011;6 Suppl 1:S12.  Back to cited text no. 10
11.Thompson C. Clive Thompson on the Cyborg Advantage. Wired Mag 2010.  Back to cited text no. 11

This article has been cited by
1 Clinical use of a machine learning histopathological image signature in diagnosis and survival prediction of clear cell renal cell carcinoma
Siteng Chen, Ning Zhang, Liren Jiang, Feng Gao, Jialiang Shao, Tao Wang, Encheng Zhang, Hong Yu, Xiang Wang, Junhua Zheng
International Journal of Cancer. 2021; 148(3): 780
[Pubmed] | [DOI]
2 Development of a Histopathology Informatics Pipeline for Classification and Prediction of Clinical Outcomes in Subtypes of Renal Cell Carcinoma
Eliana Marostica, Rebecca Barber, Thomas Denize, Isaac S. Kohane, Sabina Signoretti, Jeffrey A. Golden, Kun-Hsing Yu
Clinical Cancer Research. 2021; 27(10): 2868
[Pubmed] | [DOI]
3 Deep Convolutional Neural Network-Based Lymph Node Metastasis Prediction for Colon Cancer Using Histopathological Images
Min Seob Kwak, Hun Hee Lee, Jae Min Yang, Jae Myung Cha, Jung Won Jeon, Jin Young Yoon, Ha Il Kim
Frontiers in Oncology. 2021; 10
[Pubmed] | [DOI]
4 Integrative Analysis of Histopathological Images and Genomic Data in Colon Adenocarcinoma
Hui Li, Linyan Chen, Hao Zeng, Qimeng Liao, Jianrui Ji, Xuelei Ma
Frontiers in Oncology. 2021; 11
[Pubmed] | [DOI]
Caiyun Huang, Changhua Yin
Biomedical Engineering: Applications, Basis and Communications. 2021; 33(04): 2150031
[Pubmed] | [DOI]
6 Standardized Method for Defining a 1-mm2 Region of Interest for Calculation of Mitotic Rate on Melanoma Whole Slide Images
Minhua Wang, Phyu P. Aung, Victor G. Prieto
Archives of Pathology & Laboratory Medicine. 2021; 145(10): 1255
[Pubmed] | [DOI]
7 Clinical Decision Support for Ovarian Carcinoma Subtype Classification: A Pilot Observer Study With Pathology Trainees
Marios A. Gavrielides, Meghan Miller, Ian S. Hagemann, Heba Abdelal, Zahra Alipour, Jie-Fu Chen, Behzad Salari, Lulu Sun, Huifang Zhou, Jeffrey D Seidman
Archives of Pathology & Laboratory Medicine. 2020; 144(7): 869
[Pubmed] | [DOI]
8 Histopathological distinction of non-invasive and invasive bladder cancers using machine learning approaches
Peng-Nien Yin, Kishan KC, Shishi Wei, Qi Yu, Rui Li, Anne R. Haake, Hiroshi Miyamoto, Feng Cui
BMC Medical Informatics and Decision Making. 2020; 20(1)
[Pubmed] | [DOI]
9 Classification and Prognosis Prediction from Histopathological Images of Hepatocellular Carcinoma by a Fully Automated Pipeline Based on Machine Learning
Haotian Liao, Tianyuan Xiong, Jiajie Peng, Lin Xu, Mingheng Liao, Zhen Zhang, Zhenru Wu, Kefei Yuan, Yong Zeng
Annals of Surgical Oncology. 2020; 27(7): 2359
[Pubmed] | [DOI]
10 Report on computational assessment of Tumor Infiltrating Lymphocytes from the International Immuno-Oncology Biomarker Working Group
Mohamed Amgad, Elisabeth Specht Stovgaard, Eva Balslev, Jeppe Thagaard, Weijie Chen, Sarah Dudgeon, Ashish Sharma, Jennifer K. Kerner, Carsten Denkert, Yinyin Yuan, Khalid AbdulJabbar, Stephan Wienert, Peter Savas, Leonie Voorwerk, Andrew H. Beck, Anant Madabhushi, Johan Hartman, Manu M. Sebastian, Hugo M. Horlings, Jan Hudecek, Francesco Ciompi, David A. Moore, Rajendra Singh, Elvire Roblin, Marcelo Luiz Balancin, Marie-Christine Mathieu, Jochen K. Lennerz, Pawan Kirtani, I-Chun Chen, Jeremy P. Braybrooke, Giancarlo Pruneri, Sandra Demaria, Sylvia Adams, Stuart J. Schnitt, Sunil R. Lakhani, Federico Rojo, Laura Comerma, Sunil S. Badve, Mehrnoush Khojasteh, W. Fraser Symmans, Christos Sotiriou, Paula Gonzalez-Ericsson, Katherine L. Pogue-Geile, Rim S. Kim, David L. Rimm, Giuseppe Viale, Stephen M. Hewitt, John M. S. Bartlett, Frédérique Penault-Llorca, Shom Goel, Huang-Chun Lien, Sibylle Loibl, Zuzana Kos, Sherene Loi, Matthew G. Hanna, Stefan Michiels, Marleen Kok, Torsten O. Nielsen,
npj Breast Cancer. 2020; 6(1)
[Pubmed] | [DOI]
11 Classifying non-small cell lung cancer types and transcriptomic subtypes using convolutional neural networks
Kun-Hsing Yu, Feiran Wang, Gerald J Berry, Christopher Ré, Russ B Altman, Michael Snyder, Isaac S Kohane
Journal of the American Medical Informatics Association. 2020; 27(5): 757
[Pubmed] | [DOI]
12 Derivation of a nuclear heterogeneity image index to grade DCIS
Mary-Kate Hayward, J. Louise Jones, Allison Hall, Lorraine King, Alastair J. Ironside, Andrew C. Nelson, E. Shelley Hwang, Valerie M. Weaver
Computational and Structural Biotechnology Journal. 2020; 18: 4063
[Pubmed] | [DOI]
13 Automated Osteosclerosis Grading of Clinical Biopsies Using Infrared Spectroscopic Imaging
Rupali Mankar, Carlos E. Bueso-Ramos, C. Cameron Yin, Juliana Elisa Hidalgo-Lopez, Sebastian Berisha, Mustafa Kansiz, David Mayerich
Analytical Chemistry. 2020; 92(1): 749
[Pubmed] | [DOI]
14 A structured latent model for ovarian carcinoma subtyping from histopathology slides
Aïcha BenTaieb,Hector Li-Chang,David Huntsman,Ghassan Hamarneh
Medical Image Analysis. 2017; 39: 194
[Pubmed] | [DOI]
15 Integrative Analysis of Histopathological Images and Genomic Data Predicts Clear Cell Renal Cell Carcinoma Prognosis
Jun Cheng, Jie Zhang, Yatong Han, Xusheng Wang, Xiufen Ye, Yuebo Meng, Anil Parwani, Zhi Han, Qianjin Feng, Kun Huang
Cancer Research. 2017; 77(21): e91
[Pubmed] | [DOI]
16 Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image features
Kun-Hsing Yu,Ce Zhang,Gerald J. Berry,Russ B. Altman,Christopher Ré,Daniel L. Rubin,Michael Snyder
Nature Communications. 2016; 7: 12474
[Pubmed] | [DOI]
17 A survey on automated cancer diagnosis from histopathology images
J. Angel Arul Jothi,V. Mary Anita Rajam
Artificial Intelligence Review. 2016;
[Pubmed] | [DOI]
18 A robust nonlinear tissue-component discrimination method for computational pathology
Jacob S Sarnecki,Kathleen H Burns,Laura D Wood,Kevin M Waters,Ralph H Hruban,Denis Wirtz,Pei-Hsun Wu
Laboratory Investigation. 2016;
[Pubmed] | [DOI]
19 Selective invocation of shape priors for deformable segmentation and morphologic classification of prostate cancer tissue microarrays
Sahirzeeshan Ali,Robert Veltri,Jonathan I. Epstein,Christhunesa Christudass,Anant Madabhushi
Computerized Medical Imaging and Graphics. 2014;
[Pubmed] | [DOI]
20 Validating Whole Slide Digital Morphometric Analysis as a Microscopy Tool
Robert B. Diller,Robert S. Kellar
Microscopy and Microanalysis. 2014; : 1
[Pubmed] | [DOI]
21 Wide-field computational imaging of pathology slides using lens-free on-chip microscopy
A. Greenbaum,Y. Zhang,A. Feizi,P.-L. Chung,W. Luo,S. R. Kandukuri,A. Ozcan
Science Translational Medicine. 2014; 6(267): 267ra175
[Pubmed] | [DOI]
22 Cascaded discrimination of normal, abnormal, and confounder classes in histopathology: Gleason grading of prostate cancer
Scott Doyle,Michael D Feldman,Natalie Shih,John Tomaszewski,Anant Madabhushi
BMC Bioinformatics. 2012; 13(1): 282
[Pubmed] | [DOI]
23 Prognostic value of automatically extracted nuclear morphometric features in whole slide images of male breast cancer
Mitko Veta,Robert Kornegoor,André Huisman,Anoek H J Verschuur-Maes,Max A Viergever,Josien P W Pluim,Paul J van Diest
Modern Pathology. 2012; 25(12): 1559
[Pubmed] | [DOI]
24 Review and Cross-Validation of Gene Expression Signatures and Melanoma Prognosis
Sarah-Jane Schramm,Anna E Campain,Richard A Scolyer,Yee Hwa Yang,Graham J Mann
Journal of Investigative Dermatology. 2012; 132(2): 274
[Pubmed] | [DOI]




   Browse articles
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
    Computer Aided D...
    'How CAD can be ...
    Lessons Learned ...

 Article Access Statistics
    PDF Downloaded1237    
    Comments [Add]    
    Cited by others 24    

Recommend this journal