Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Meningioma microstructure assessed by diffusion MRI : An investigation of the source of mean diffusivity and fractional anisotropy by quantitative histology

Brabec, Jan LU ; Friedjungová, Magda ; Vašata, Daniel ; Englund, Elisabet LU orcid ; Bengzon, Johan LU ; Knutsson, Linda LU orcid ; Szczepankiewicz, Filip LU orcid ; van Westen, Danielle LU orcid ; Sundgren, Pia C LU orcid and Nilsson, Markus LU (2023) In NeuroImage: Clinical 37.
Abstract

BACKGROUND: Mean diffusivity (MD) and fractional anisotropy (FA) from diffusion MRI (dMRI) have been associated with cell density and tissue anisotropy across tumors, but it is unknown whether these associations persist at the microscopic level.

PURPOSE: To quantify the degree to which cell density and anisotropy, as determined from histology, account for the intra-tumor variability of MD and FA in meningioma tumors. Furthermore, to clarify whether other histological features account for additional intra-tumor variability of dMRI parameters.

MATERIALS AND METHODS: We performed ex-vivo dMRI at 200 μm isotropic resolution and histological imaging of 16 excised meningioma tumor samples. Diffusion tensor imaging (DTI) was used... (More)

BACKGROUND: Mean diffusivity (MD) and fractional anisotropy (FA) from diffusion MRI (dMRI) have been associated with cell density and tissue anisotropy across tumors, but it is unknown whether these associations persist at the microscopic level.

PURPOSE: To quantify the degree to which cell density and anisotropy, as determined from histology, account for the intra-tumor variability of MD and FA in meningioma tumors. Furthermore, to clarify whether other histological features account for additional intra-tumor variability of dMRI parameters.

MATERIALS AND METHODS: We performed ex-vivo dMRI at 200 μm isotropic resolution and histological imaging of 16 excised meningioma tumor samples. Diffusion tensor imaging (DTI) was used to map MD and FA, as well as the in-plane FA (FA
IP). Histology images were analyzed in terms of cell nuclei density (CD) and structure anisotropy (SA; obtained from structure tensor analysis) and were used separately in a regression analysis to predict MD and FA
IP, respectively. A convolutional neural network (CNN) was also trained to predict the dMRI parameters from histology patches. The association between MRI and histology was analyzed in terms of out-of-sample (R
2
OS) on the intra-tumor level and within-sample R
2 across tumors. Regions where the dMRI parameters were poorly predicted from histology were analyzed to identify features apart from CD and SA that could influence MD and FA
IP, respectively.

RESULTS: Cell density assessed by histology poorly explained intra-tumor variability of MD at the mesoscopic level (200 μm), as median R
2
OS = 0.04 (interquartile range 0.01-0.26). Structure anisotropy explained more of the variation in FA
IP (median R
2
OS = 0.31, 0.20-0.42). Samples with low R
2
OS for FA
IP exhibited low variations throughout the samples and thus low explainable variability, however, this was not the case for MD. Across tumors, CD and SA were clearly associated with MD (R
2 = 0.60) and FA
IP (R
2 = 0.81), respectively. In 37% of the samples (6 out of 16), cell density did not explain intra-tumor variability of MD when compared to the degree explained by the CNN. Tumor vascularization, psammoma bodies, microcysts, and tissue cohesivity were associated with bias in MD prediction based solely on CD. Our results support that FA
IP is high in the presence of elongated and aligned cell structures, but low otherwise.

CONCLUSION: Cell density and structure anisotropy account for variability in MD and FA
IP across tumors but cell density does not explain MD variations within the tumor, which means that low or high values of MD locally may not always reflect high or low tumor cell density. Features beyond cell density need to be considered when interpreting MD.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
NeuroImage: Clinical
volume
37
article number
103365
publisher
Elsevier
external identifiers
  • scopus:85149703543
  • pmid:36898293
ISSN
2213-1582
DOI
10.1016/j.nicl.2023.103365
language
English
LU publication?
yes
additional info
Published by Elsevier Inc.
id
12b7c08f-8fdd-48c8-aadc-4b48c296e058
date added to LUP
2023-03-11 17:56:54
date last changed
2024-11-01 04:37:13
@article{12b7c08f-8fdd-48c8-aadc-4b48c296e058,
  abstract     = {{<p>BACKGROUND: Mean diffusivity (MD) and fractional anisotropy (FA) from diffusion MRI (dMRI) have been associated with cell density and tissue anisotropy across tumors, but it is unknown whether these associations persist at the microscopic level.</p><p>PURPOSE: To quantify the degree to which cell density and anisotropy, as determined from histology, account for the intra-tumor variability of MD and FA in meningioma tumors. Furthermore, to clarify whether other histological features account for additional intra-tumor variability of dMRI parameters.</p><p>MATERIALS AND METHODS: We performed ex-vivo dMRI at 200 μm isotropic resolution and histological imaging of 16 excised meningioma tumor samples. Diffusion tensor imaging (DTI) was used to map MD and FA, as well as the in-plane FA (FA<br>
 IP). Histology images were analyzed in terms of cell nuclei density (CD) and structure anisotropy (SA; obtained from structure tensor analysis) and were used separately in a regression analysis to predict MD and FA<br>
 IP, respectively. A convolutional neural network (CNN) was also trained to predict the dMRI parameters from histology patches. The association between MRI and histology was analyzed in terms of out-of-sample (R<br>
 2<br>
 OS) on the intra-tumor level and within-sample R<br>
 2 across tumors. Regions where the dMRI parameters were poorly predicted from histology were analyzed to identify features apart from CD and SA that could influence MD and FA <br>
 IP, respectively.<br>
 </p><p>RESULTS: Cell density assessed by histology poorly explained intra-tumor variability of MD at the mesoscopic level (200 μm), as median R<br>
 2<br>
 OS = 0.04 (interquartile range 0.01-0.26). Structure anisotropy explained more of the variation in FA<br>
 IP (median R<br>
 2<br>
 OS = 0.31, 0.20-0.42). Samples with low R<br>
 2<br>
 OS for FA<br>
 IP exhibited low variations throughout the samples and thus low explainable variability, however, this was not the case for MD. Across tumors, CD and SA were clearly associated with MD (R<br>
 2 = 0.60) and FA<br>
 IP (R<br>
 2 = 0.81), respectively. In 37% of the samples (6 out of 16), cell density did not explain intra-tumor variability of MD when compared to the degree explained by the CNN. Tumor vascularization, psammoma bodies, microcysts, and tissue cohesivity were associated with bias in MD prediction based solely on CD. Our results support that FA<br>
 IP is high in the presence of elongated and aligned cell structures, but low otherwise.<br>
 </p><p>CONCLUSION: Cell density and structure anisotropy account for variability in MD and FA<br>
 IP across tumors but cell density does not explain MD variations within the tumor, which means that low or high values of MD locally may not always reflect high or low tumor cell density. Features beyond cell density need to be considered when interpreting MD.<br>
 </p>}},
  author       = {{Brabec, Jan and Friedjungová, Magda and Vašata, Daniel and Englund, Elisabet and Bengzon, Johan and Knutsson, Linda and Szczepankiewicz, Filip and van Westen, Danielle and Sundgren, Pia C and Nilsson, Markus}},
  issn         = {{2213-1582}},
  language     = {{eng}},
  month        = {{03}},
  publisher    = {{Elsevier}},
  series       = {{NeuroImage: Clinical}},
  title        = {{Meningioma microstructure assessed by diffusion MRI : An investigation of the source of mean diffusivity and fractional anisotropy by quantitative histology}},
  url          = {{http://dx.doi.org/10.1016/j.nicl.2023.103365}},
  doi          = {{10.1016/j.nicl.2023.103365}},
  volume       = {{37}},
  year         = {{2023}},
}