On the concept of objectivity in digital image analysis in pathology

Aims: The term ‘objective’ connotes a method that is based on facts and not influenced by personal opinions, perception or emotion. One often reads in the biomedical literature claims of objectivity for methods that use digital image analysis applied to histology. Since objective assessment of histology would represent a huge leap forward in scientific measurement and clinical diagnosis, such claims should be substantiated by strong evidence. This paper takes a selective look at the literature on image analysis to assess the definition of objectivity in image analysis and asks whether such a claim is ever justified. Methods: First, a brief background on the basic science of image analysis in histology details some of the controversies and opinions in the field. Then, a literature review of a subset of papers pertaining to image analysis in histology (with claims of objectivity) is conducted to determine what evidence exists for objectivity in these methods. Results: It was found that image analysis may have many benefits (speed, indefatigability, standardisation, etc.). However, algorithms are devised and implemented by human beings who make subjective decisions at each stage of the algorithm design and implementation process. Thus, image analysis methods can be seen as deterministic processes which ‘objectively’ implement the subjective decisions of the programmer. This indicates that ‘inter‐observer’ variation in image analysis is equivalent to ‘inter‐algorithm’ variation (which is rarely studied) and that a single computer algorithm's repeatability is of lesser importance than the repeatability of the image analysis method as a whole (including the block, slide and field selection and the method of tissue processing). Conclusion: Repeatability and automaticity must not be confused with objectivity, but a lack of objectivity does not imply a lack of utility. Unless specific evidence of objectivity is provided, editors should insist that claims of objectivity in image analysis papers be either removed or justified prior to publication.

[1]  Milan Sonka,et al.  Image processing analysis and machine vision [2nd ed.] , 1999 .

[2]  J W Bacus,et al.  Biological grading of breast cancer using antibodies to proliferating cells and other markers. , 1989, The American journal of pathology.

[3]  M. Duffy,et al.  Objective measurement of breast cancer oestrogen receptor status through digital image analysis. , 2003, European journal of surgical oncology : the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology.

[4]  William M Gallagher,et al.  Novel image analysis approach for quantifying expression of nuclear proteins assessed by immunohistochemistry: application to measurement of oestrogen and progesterone receptor levels in breast cancer , 2008, Breast Cancer Research.

[5]  W. Lambert,et al.  Prognostic value of morphometry in papillary thyroid carcinoma , 1989, Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[6]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[7]  A. Ruifrok,et al.  Comparison of Quantification of Histochemical Staining By Hue-Saturation-Intensity (HSI) Transformation and Color-Deconvolution , 2003, Applied immunohistochemistry & molecular morphology : AIMM.

[8]  J. Vegelius,et al.  Computerized nuclear morphometry as an objective method for characterizing human cancer cell populations. , 1978, Cancer research.

[9]  Zenon Kulpa,et al.  Area and perimeter measurement of blobs in discrete binary pictures , 1977 .

[10]  O. Maximova,et al.  Computerized Morphometric Analysis of Pathological Prion Protein Deposition in Scrapie-Infected Hamster Brain , 2006, The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society.

[11]  Herbert Freeman,et al.  On the Encoding of Arbitrary Geometric Configurations , 1961, IRE Trans. Electron. Comput..

[12]  B J Addis,et al.  Inter-observer variation between pathologists in diffuse parenchymal lung disease , 2004, Thorax.

[13]  D. Grube.  Constants and variables in immunohistochemistry. , 2004, Archives of histology and cytology.

[14]  R. Jagoe,et al.  Shape and texture analysis of liver cell nuclei in hepatomas by computer aided microscopy. , 1984, Journal of clinical pathology.

[15]  D. Varberg.  Pick's Theorem Revisited , 1985 .

[16]  Ventzas Dimitrios,et al.  Binarization of pre‐filtered historical manuscripts images , 2009 .

[17]  R. Wootton,et al.  Observer variation in quantification of immunocytochemistry by image analysis , 1991, The Histochemical Journal.

[18]  Josef Kittler,et al.  On threshold selection using clustering criteria , 1985, IEEE Transactions on Systems, Man, and Cybernetics.

[19]  Jinsong Liu,et al.  Assessment of Two Automated Imaging Systems in Evaluating Estrogen Receptor Status in Breast Carcinoma , 2007, Applied immunohistochemistry & molecular morphology : AIMM.

[20]  Milan Sonka,et al.  Image Processing, Analysis and Machine Vision , 1993, Springer US.

[21]  John Meyer,et al.  Grading nuclear pleomorphism on histological micrographs , 2008, 2008 19th International Conference on Pattern Recognition.

[22]  A. Seth,et al.  Computer based receptogram approach: an objective way of assessing immunohistochemistry of androgen receptor staining and its correlation with hormonal response in metastatic carcinoma of prostate , 2004, Journal of Clinical Pathology.

[23]  J. Gil.  Image analysis in anatomic pathology: what are the issues? , 1989, Human pathology.

[24]  Paul Wintz,et al.  Digital image processing (2nd ed.) , 1987 .

[25]  W. Seeger,et al.  Comparison of different detection methods in quantitative microdensitometry. , 2001, The American journal of pathology.

[26]  Philip Sloan,et al.  Why oral histopathology suffers inter-observer variability on grading oral epithelial dysplasia: an attempt to understand the sources of variation. , 2007, Oral oncology.

[27]  R. Becker,et al.  A comparative morphometric and cytophotometric study of endometrial hyperplasia, atypical hyperplasia, and endometrial carcinoma. , 1989, Human pathology.

[28]  G. Ostermeier,et al.  Measurement of bovine sperm nuclear shape using Fourier harmonic amplitudes. , 2001, Journal of andrology.

[29]  P. Aspelin,et al.  Comparison of lesion size estimated by dynamic MR imaging, mammography and histopathology in breast neoplasms , 2003, European Radiology.

[30]  A. Ruifrok,et al.  Quantification of histochemical staining by color deconvolution. , 2001, Analytical and quantitative cytology and histology.

[31]  Joakim Lindblad,et al.  High-Precision Boundary Length Estimation by Utilizing Gray-Level Information , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[32]  Pascal Bamford,et al.  Empirical comparison of cell segmentation algorithms using an annotated dataset , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).