ANHIR: Automatic Non-Rigid Histological Image Registration Challenge

Borovec, Jiří; Kybic, Jan; Arganda‐Carreras, Ignacio; Sorokin, Dmitry V.; Bueno, Gloria; Khvostikov, Alexander; Bakas, Spyridon; Chang, Eric; Heldmann, Stefan; Kartasalo, Kimmo; Latonen, Leena; Lotz, Johannes; Noga, Michelle; Pati, Sarthak; Punithakumar, Kumaradevan; Ruusuvuori, Pekka; Skalski, Andrzej; Tahmasebi, Nazanin; Valkonen, Masi; Venet, Ludovic; Wang, Yizhe; Weiss, Nick; Wodziński, Marek; Yu, Xi; Xu, Yan; Yan, Yan; Yushkevich, Paul A.; Zhao, Shengyu; Muñoz-Barrutia, Arrate

doi:10.1109/tmi.2020.2986331

Cited by 113 publications

(125 citation statements)

References 67 publications

Supporting

Mentioning

124

Contrasting

Order By: Relevance

“…1 a). Distinct repositories exist for various medical fields, e.g., radiology 2 – 9 , pathology 10 , and genomics 11 . We refer to this approach as collaborative data sharing ( CDS ).…”

Section: Introductionmentioning

confidence: 99%

Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data

Sheller

Edwards

Reina

et al. 2020

Sci Rep

Self Cite

739

371

View full text Add to dashboard Cite

Several studies underscore the potential of deep learning in identifying complex patterns, leading to diagnostic and prognostic biomarkers. Identifying sufficiently large and diverse datasets, required for training, is a significant challenge in medicine and can rarely be found in individual institutions. Multi-institutional collaborations based on centrally-shared patient data face privacy and ownership challenges. federated learning is a novel paradigm for data-private multi-institutional collaborations, where model-learning leverages all available data without sharing data between institutions, by distributing the model-training to the data-owners and aggregating their results. We show that federated learning among 10 institutions results in models reaching 99% of the model quality achieved with centralized data, and evaluate generalizability on data from institutions outside the federation. We further investigate the effects of data distribution across collaborating institutions on model quality and learning patterns, indicating that increased access to data through data private multi-institutional collaborations can benefit model quality more than the errors introduced by the collaborative method. finally, we compare with other collaborative-learning approaches demonstrating the superiority of federated learning, and discuss practical implementation considerations. clinical adoption of federated learning is expected to lead to models trained on datasets of unprecedented size, hence have a catalytic impact towards precision/personalized medicine. Abbreviations CDS Collaborative data sharing FL Federated learning IIL Institutional incremental learning CIIL Cyclic institutional incremental learning IID Independent and identically distributed BraTS Brain tumor segmentation

show abstract

“…1 a). Distinct repositories exist for various medical fields, e.g., radiology 2 – 9 , pathology 10 , and genomics 11 . We refer to this approach as collaborative data sharing ( CDS ).…”

Section: Introductionmentioning

confidence: 99%

Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data

Sheller

Edwards

Reina

et al. 2020

Sci Rep

Self Cite

739

371

View full text Add to dashboard Cite

show abstract

“…Registering histological stains with little or no grey-white matter contrast is beyond the scope of the current work, and is therefore not readily supported by the current pipeline. However, the results of a recent grand challenge competition (ANHIR) [66] might be used in the future to register histological sections with different stains in advance, and the ones with appropriate grey-white matter contrast to the MRI. Alternatively, these images could be registered linearly by matching outer contours or non-linearly by manually defined landmarks.…”

Section: Discussionmentioning

confidence: 99%

Tensor Image Registration Library: Automated Non-Linear Registration of Sparsely Sampled Histological Specimens to Post-Mortem MRI of the Whole Human Brain

Huszár

Pallebage-Gamarallage

Foxley

et al. 2019

Preprint

View full text Add to dashboard Cite

Highlights 29 30• TIRL: new framework for prototyping bespoke image registration pipelines 31• Pipeline for automated registration of small-slide histology to whole-brain MRI 32• Slice-to-volume registration accounting for through-plane deformations 33• No need for serial histological sampling 34 35Abstract 36 37 There is a need to understand the histopathological basis of MRI signal characteristics in 38 complex biological matter. Microstructural imaging holds promise for sensitive and specific 39 indicators of the early stages of human neurodegeneration but requires validation against 40 traditional histological markers before it can be reliably applied in the clinical setting. 41Validation relies on a precise and preferably automatic method to align MRI and histological 42 images of the same tissue, which poses unique challenges compared to more conventional 43 MRI-to-MRI registration. 44 45A customisable open-source platform, Tensor Image Registration Library (TIRL) is presented. 46Based on TIRL, a fully automated pipeline was implemented to align small stained histological 47

show abstract

“…For this work, we used publicly available data from the Automatic Non-rigid Histological Image Registration (ANHIR) challenge 2019 [ 5 , 24 , 25 ]. This challenge was set up to compare the performance of different registration algorithms on a varied set of microscopy histology images.…”

Section: Methodsmentioning

confidence: 99%

Nonlinear Image Registration and Pixel Classification Pipeline for the Study of Tumor Heterogeneity Maps

Nicolás-Sáenz

Guerrero-Aspizúa

Pascau

et al. 2020

Entropy

View full text Add to dashboard Cite

We present a novel method to assess the variations in protein expression and spatial heterogeneity of tumor biopsies with application in computational pathology. This was done using different antigen stains for each tissue section and proceeding with a complex image registration followed by a final step of color segmentation to detect the exact location of the proteins of interest. For proper assessment, the registration needs to be highly accurate for the careful study of the antigen patterns. However, accurate registration of histopathological images comes with three main problems: the high amount of artifacts due to the complex biopsy preparation, the size of the images, and the complexity of the local morphology. Our method manages to achieve an accurate registration of the tissue cuts and segmentation of the positive antigen areas.

show abstract

ANHIR: Automatic Non-Rigid Histological Image Registration Challenge

Cited by 113 publications

References 67 publications

Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data

Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data

Tensor Image Registration Library: Automated Non-Linear Registration of Sparsely Sampled Histological Specimens to Post-Mortem MRI of the Whole Human Brain

Nonlinear Image Registration and Pixel Classification Pipeline for the Study of Tumor Heterogeneity Maps

Contact Info

Product

Resources

About