Assessment of algorithms for mitosis detection in breast cancer histopathology images

Veta, Mitko; Diest, Paul J. van; Willems, Stefan M.; Wang, Haibo; Madabhushi, Anant; Cruz-Roa, Ángel; González, Fabio A.; Larsen, Allan; Vestergaard, Jacob S.; Dahl, Anders Bjorholm; Cireşan, Dan; Schmidhuber, Juergen; Giusti, Alessandro; Gambardella, Luca Maria; Tek, F. Boray; Walter, Thomas; Wang, Ching-Wei; Kondo, Satoshi; Matuszewski, Bogdan J.; Precioso, Fŕed́eric; Snell, Violet; Kittler, Josef; Campos, Teófilo de; Khan, Asifullah; Rajpoot, Nasir M.; Arkoumani, Evdokia; Laclé, Miangela M.; Viergever, Max A.; Pluim, Josien P. W.

doi:10.1016/j.media.2014.11.010

Cited by 403 publications

(312 citation statements)

References 31 publications

Supporting

Mentioning

292

Contrasting

Unclassified

Order By: Relevance

“…Analyses based on WSI images shows a higher degree of agreement when measuring expression of HER2 / Neu with the results of FISH gene amplification (fluorescence in situ hybridization) in breast cancer compared with the traditional assessment by microscopy [39]. Nevertheless, digital analysis is not fully effective for all its applications; for instance when measuring mitotic index , using the most advanced image analysis software, only a 80-85% agreement is reached when compared to conventional microscopy [40]. This aforementioned digital analysis is a very useful IT tool, but immunohistochemistry measurements represents only a small proportion of the pathologist's diagnostic workload.…”

Section: Digital Diagnosticsmentioning

confidence: 98%

Future perspectives of digital pathology

Prochorec‐Sobieszek

2016

Nowotwory. Journal of Oncology

View full text Add to dashboard Cite

Technological advances have enabled innovative solutions to be achieved in pathology based on digital imaging, now superseding those of conventional microscopy. Digital pathology has been defined as 'virtual microscopy' and depends on computer-generated digital imaging of microscope slides (WSI -whole slide imaging) which are in turn created, reviewed, managed, shared, analysed and interpreted. Such WSI systems and digital consulting platforms are now used for teaching, scientific research, telepathology / teleconsultation and diagnostics. They also permit easy and interactive sharing of WSI that can be integrated into other medical information systems. The software for automated image analysis and computer aided diagnosis can thereby make highly accurate diagnoses and help standardise study findings. Despite the technique's many advantages, its noted drawbacks include high equipment and software costs, image quality issues of standardisation and most importantly, that pathologists are reluctant to use it routinely for making diagnoses. NOWOTWORY J Oncol 2016; 66, 4: 277-285

show abstract

Section: Digital Diagnosticsmentioning

confidence: 98%

Future perspectives of digital pathology

Prochorec‐Sobieszek

2016

Nowotwory. Journal of Oncology

View full text Add to dashboard Cite

show abstract

“…We compared our approach with seven state-of-the-art approaches using the same MITOS dataset, developed by the research teams who participated in the ICPR 2012 [58] or/and MICCAI 2013 [60,64] competitions on mitosis detection and ranked as the top 7 best performing methods. The first approach (IDSIA) developed a pixel-based deep neural network framework to identify mitosis [20], which won the first place in both ICPR 2012 and MIC-CAI 2013 contests.…”

Section: Comparative Strategiesmentioning

confidence: 99%

Automated mitosis detection in histopathology based on non-gaussian modeling of complex wavelet coefficients

et al. 2017

View full text Add to dashboard Cite

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. AbstractTo diagnose breast cancer, the number of mitotic cells present in histology sections is an important indicator for examining and grading biopsy specimen. This study aims at improving the accuracy of automated mitosis detection by characterizing mitotic cells in wavelet based multi-resolution representations via a non-Gaussian modeling method. The potential mitosis candidates were decomposed into multi-scale forms by an undecimated dual-tree complex wavelet transform. Two non-Gaussian models (the generalized Gaussian distribution (GGD) and the symmetric alpha-stable (SαS) distributions) were used to accurately model the heavy-tailed behavior of wavelet marginal distributions. The method was evaluated on two independent data cohorts, including the benchmark dataset (MITOS), via a support vector machine classifier.

show abstract

“…The inconsistency of stain condition makes the appearances of H&E stained histology drastically different, so the classify performance was degraded [13] [14]. For instance, many false mitosis may arise when the histopathology slide is over-stained.…”

Section: Stain-normalization and Training Dataset Buildingmentioning

confidence: 99%

Multi-Scale Deep Neural Network for Mitosis Detection in Breast Cancer Histological Images

Wu¹,

Fan²,

Qiao³

et al. 2017

Preprint

View full text Add to dashboard Cite

Abstract. Accurate assessment of the breast cancer deterioration degree plays a crucial role in making medical plan, and the important basis for degree assessment is the number of mitoses in a given area of the pathological image. We utilized deep multi-scale fused fully convolutional neural network (MFF-CNN) combing with conditional random felid (CRF) to detect mitoses in hematoxylin and eosin stained histology image. Analyze the characteristics of mitotic detection ----scale invariance and sparsity, as well as the difficulties ----small amount of data , inconsistent image staining and sample class unbalanced. Based on this, mitotic detection model is designed. In this paper, a tissue-based staining equalization method is used, and to establish an effective training sample set, we select training samples by using CNN. A mitotic detection model fusing multi-level and multi-scale features and context information was designed, and the corresponding training strategy was made to reduce over-fitting. As preliminarily validated on the public 2014 ICPR MITOSIS data, our method achieves a better performance in term of detection accuracy than ever recorded for this dataset.

show abstract

Assessment of algorithms for mitosis detection in breast cancer histopathology images

Cited by 403 publications

References 31 publications

Future perspectives of digital pathology

Future perspectives of digital pathology

Automated mitosis detection in histopathology based on non-gaussian modeling of complex wavelet coefficients

Multi-Scale Deep Neural Network for Mitosis Detection in Breast Cancer Histological Images

Contact Info

Product

Resources

About