Model-free classification of X-ray scattering signals applied to image segmentation

Viviane Lutz-Bueno, C. Arboleda, Leon Leu, Martin J. Blunt, Andreas Busch, A. Georgiadis, Pieter Bertier, J. Schmatz, Z. Varga, P. Villanueva-Perez, Z. Wang, M. Lebugle, C. David, M. Stampanoni, A. Diaz, M. Guizar-Sicairos, A. Menzel

Research output: Contribution to journalArticle

Abstract

In most cases, the analysis of small-angle and wide-angle X-ray scattering, SAXS and WAXS, respectively, requires a theoretical model to describe the sample's scattering, complicating the interpretation of the scattering from complex heterogeneous samples. This is the reason why in most cases, the analysis of the large number of scattering patterns, such as is generated by time-resolved and scanning methods, remains challenging. Here we introduce and demonstrate a model-free classification method to separate SAXS/WAXS signals based on their inflection points. We focus on the segmentation of scanning SAXS/WAXS maps, for which each pixel corresponds to an azimuthally integrated scattering curve. In such a way, the sample composition distribution can be segmented through signal classification without applying a model or previous sample knowledge. We employ dimensionality reduction and clustering algorithms to classify SAXS/WAXS signals according to their similarity. The number of clusters, i.e. the main sample regions detected by SAXS/WAXS signal similarity, is automatically estimated. From each cluster, a main representative SAXS/WAXS signal is extracted to uncover the spatial distribution of the mixtures of phases that form the sample. As examples of application, a mudrock sample and two breast tissue lesions are segmented.
LanguageEnglish
Pages1378-1386
Number of pages9
JournalJournal of Applied Crystallography
Volume51
Issue number5
DOIs
Publication statusPublished - Oct 2018

Fingerprint

segmentation
scattering
lesion
mudstone
pixel
spatial distribution
method
analysis

Cite this

Lutz-Bueno, V., Arboleda, C., Leu, L., Blunt, M. J., Busch, A., Georgiadis, A., ... Menzel, A. (2018). Model-free classification of X-ray scattering signals applied to image segmentation. Journal of Applied Crystallography, 51(5), 1378-1386. https://doi.org/10.1107/S1600576718011032
Lutz-Bueno, Viviane ; Arboleda, C. ; Leu, Leon ; Blunt, Martin J. ; Busch, Andreas ; Georgiadis, A. ; Bertier, Pieter ; Schmatz, J. ; Varga, Z. ; Villanueva-Perez, P. ; Wang, Z. ; Lebugle, M. ; David, C. ; Stampanoni, M. ; Diaz, A. ; Guizar-Sicairos, M. ; Menzel, A. / Model-free classification of X-ray scattering signals applied to image segmentation. In: Journal of Applied Crystallography. 2018 ; Vol. 51, No. 5. pp. 1378-1386.
@article{ffaf7687e06d4b2396ee5c4e440340a9,
title = "Model-free classification of X-ray scattering signals applied to image segmentation",
abstract = "In most cases, the analysis of small-angle and wide-angle X-ray scattering, SAXS and WAXS, respectively, requires a theoretical model to describe the sample's scattering, complicating the interpretation of the scattering from complex heterogeneous samples. This is the reason why in most cases, the analysis of the large number of scattering patterns, such as is generated by time-resolved and scanning methods, remains challenging. Here we introduce and demonstrate a model-free classification method to separate SAXS/WAXS signals based on their inflection points. We focus on the segmentation of scanning SAXS/WAXS maps, for which each pixel corresponds to an azimuthally integrated scattering curve. In such a way, the sample composition distribution can be segmented through signal classification without applying a model or previous sample knowledge. We employ dimensionality reduction and clustering algorithms to classify SAXS/WAXS signals according to their similarity. The number of clusters, i.e. the main sample regions detected by SAXS/WAXS signal similarity, is automatically estimated. From each cluster, a main representative SAXS/WAXS signal is extracted to uncover the spatial distribution of the mixtures of phases that form the sample. As examples of application, a mudrock sample and two breast tissue lesions are segmented.",
author = "Viviane Lutz-Bueno and C. Arboleda and Leon Leu and Blunt, {Martin J.} and Andreas Busch and A. Georgiadis and Pieter Bertier and J. Schmatz and Z. Varga and P. Villanueva-Perez and Z. Wang and M. Lebugle and C. David and M. Stampanoni and A. Diaz and M. Guizar-Sicairos and A. Menzel",
year = "2018",
month = "10",
doi = "10.1107/S1600576718011032",
language = "English",
volume = "51",
pages = "1378--1386",
journal = "Journal of Applied Crystallography",
issn = "0021-8898",
publisher = "International Union of Crystallography",
number = "5",

}

Lutz-Bueno, V, Arboleda, C, Leu, L, Blunt, MJ, Busch, A, Georgiadis, A, Bertier, P, Schmatz, J, Varga, Z, Villanueva-Perez, P, Wang, Z, Lebugle, M, David, C, Stampanoni, M, Diaz, A, Guizar-Sicairos, M & Menzel, A 2018, 'Model-free classification of X-ray scattering signals applied to image segmentation', Journal of Applied Crystallography, vol. 51, no. 5, pp. 1378-1386. https://doi.org/10.1107/S1600576718011032

Model-free classification of X-ray scattering signals applied to image segmentation. / Lutz-Bueno, Viviane; Arboleda, C.; Leu, Leon; Blunt, Martin J.; Busch, Andreas; Georgiadis, A.; Bertier, Pieter; Schmatz, J.; Varga, Z.; Villanueva-Perez, P.; Wang, Z.; Lebugle, M.; David, C.; Stampanoni, M.; Diaz, A.; Guizar-Sicairos, M.; Menzel, A.

In: Journal of Applied Crystallography, Vol. 51, No. 5, 10.2018, p. 1378-1386.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Model-free classification of X-ray scattering signals applied to image segmentation

AU - Lutz-Bueno, Viviane

AU - Arboleda, C.

AU - Leu, Leon

AU - Blunt, Martin J.

AU - Busch, Andreas

AU - Georgiadis, A.

AU - Bertier, Pieter

AU - Schmatz, J.

AU - Varga, Z.

AU - Villanueva-Perez, P.

AU - Wang, Z.

AU - Lebugle, M.

AU - David, C.

AU - Stampanoni, M.

AU - Diaz, A.

AU - Guizar-Sicairos, M.

AU - Menzel, A.

PY - 2018/10

Y1 - 2018/10

N2 - In most cases, the analysis of small-angle and wide-angle X-ray scattering, SAXS and WAXS, respectively, requires a theoretical model to describe the sample's scattering, complicating the interpretation of the scattering from complex heterogeneous samples. This is the reason why in most cases, the analysis of the large number of scattering patterns, such as is generated by time-resolved and scanning methods, remains challenging. Here we introduce and demonstrate a model-free classification method to separate SAXS/WAXS signals based on their inflection points. We focus on the segmentation of scanning SAXS/WAXS maps, for which each pixel corresponds to an azimuthally integrated scattering curve. In such a way, the sample composition distribution can be segmented through signal classification without applying a model or previous sample knowledge. We employ dimensionality reduction and clustering algorithms to classify SAXS/WAXS signals according to their similarity. The number of clusters, i.e. the main sample regions detected by SAXS/WAXS signal similarity, is automatically estimated. From each cluster, a main representative SAXS/WAXS signal is extracted to uncover the spatial distribution of the mixtures of phases that form the sample. As examples of application, a mudrock sample and two breast tissue lesions are segmented.

AB - In most cases, the analysis of small-angle and wide-angle X-ray scattering, SAXS and WAXS, respectively, requires a theoretical model to describe the sample's scattering, complicating the interpretation of the scattering from complex heterogeneous samples. This is the reason why in most cases, the analysis of the large number of scattering patterns, such as is generated by time-resolved and scanning methods, remains challenging. Here we introduce and demonstrate a model-free classification method to separate SAXS/WAXS signals based on their inflection points. We focus on the segmentation of scanning SAXS/WAXS maps, for which each pixel corresponds to an azimuthally integrated scattering curve. In such a way, the sample composition distribution can be segmented through signal classification without applying a model or previous sample knowledge. We employ dimensionality reduction and clustering algorithms to classify SAXS/WAXS signals according to their similarity. The number of clusters, i.e. the main sample regions detected by SAXS/WAXS signal similarity, is automatically estimated. From each cluster, a main representative SAXS/WAXS signal is extracted to uncover the spatial distribution of the mixtures of phases that form the sample. As examples of application, a mudrock sample and two breast tissue lesions are segmented.

U2 - 10.1107/S1600576718011032

DO - 10.1107/S1600576718011032

M3 - Article

VL - 51

SP - 1378

EP - 1386

JO - Journal of Applied Crystallography

T2 - Journal of Applied Crystallography

JF - Journal of Applied Crystallography

SN - 0021-8898

IS - 5

ER -