Automated detection of arterial input function in DSC perfusion MRI in a stroke rat model

M. Y. Yeh, T. H. Lee, S. T. Yang, H. H. Kuo, T. K. Chyi, H. L. Liu

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Quantitative cerebral blood flow (CBF) estimation requires deconvolution of the tissue concentration time curves with an arterial input function (AIF). However, image-based determination of AIF in rodent is challenged due to limited spatial resolution. We evaluated the feasibility of quantitative analysis using automated AIF detection and compared the results with commonly applied semi-quantitative analysis. Permanent occlusion of bilateral or unilateral common carotid artery was used to induce cerebral ischemia in rats. The image using dynamic susceptibility contrast method was performed on a 3-T magnetic resonance scanner with a spin-echo echo-planar-image sequence (TR/TE = 700/80 ms, FOV = 41 mm, matrix = 64, 3 slices, SW = 2 mm), starting from 7 s prior to contrast injection (1.2 ml/kg) at four different time points. For quantitative analysis, CBF was calculated by the AIF which was obtained from 10 voxels with greatest contrast enhancement after deconvolution. For semi-quantitative analysis, relative CBF was estimated by the integral divided by the first moment of the relaxivity time curves. We observed if the AIFs obtained in the three different ROIs (whole brain, hemisphere without lesion and hemisphere with lesion) were similar, the CBF ratios (lesion/normal) between quantitative and semi-quantitative analyses might have a similar trend at different operative time points. If the AIFs were different, the CBF ratios might be different. We concluded that using local maximum one can define proper AIF without knowing the anatomical location of arteries in a stroke rat model.

Original languageEnglish
Article numberP05021
JournalJournal of Instrumentation
Volume4
Issue number5
DOIs
Publication statusPublished - 2009
Externally publishedYes

Fingerprint

blood flow
Blood Flow
strokes
Stroke
Magnetic resonance imaging
rats
Rats
Quantitative Analysis
Blood
quantitative analysis
lesions
Hemisphere
Deconvolution
Arteries
arteries
hemispheres
Chemical analysis
echoes
ischemia
Contrast Enhancement

Keywords

  • Data processing methods
  • MRI (whole body, cardiovascular, breast, others), MRangiography (MRA)

ASJC Scopus subject areas

  • Instrumentation
  • Mathematical Physics

Cite this

Automated detection of arterial input function in DSC perfusion MRI in a stroke rat model. / Yeh, M. Y.; Lee, T. H.; Yang, S. T.; Kuo, H. H.; Chyi, T. K.; Liu, H. L.

In: Journal of Instrumentation, Vol. 4, No. 5, P05021, 2009.

Research output: Contribution to journalArticle

Yeh, M. Y. ; Lee, T. H. ; Yang, S. T. ; Kuo, H. H. ; Chyi, T. K. ; Liu, H. L. / Automated detection of arterial input function in DSC perfusion MRI in a stroke rat model. In: Journal of Instrumentation. 2009 ; Vol. 4, No. 5.
@article{40546bb768fe4bdf9967c11db0d2c070,
title = "Automated detection of arterial input function in DSC perfusion MRI in a stroke rat model",
abstract = "Quantitative cerebral blood flow (CBF) estimation requires deconvolution of the tissue concentration time curves with an arterial input function (AIF). However, image-based determination of AIF in rodent is challenged due to limited spatial resolution. We evaluated the feasibility of quantitative analysis using automated AIF detection and compared the results with commonly applied semi-quantitative analysis. Permanent occlusion of bilateral or unilateral common carotid artery was used to induce cerebral ischemia in rats. The image using dynamic susceptibility contrast method was performed on a 3-T magnetic resonance scanner with a spin-echo echo-planar-image sequence (TR/TE = 700/80 ms, FOV = 41 mm, matrix = 64, 3 slices, SW = 2 mm), starting from 7 s prior to contrast injection (1.2 ml/kg) at four different time points. For quantitative analysis, CBF was calculated by the AIF which was obtained from 10 voxels with greatest contrast enhancement after deconvolution. For semi-quantitative analysis, relative CBF was estimated by the integral divided by the first moment of the relaxivity time curves. We observed if the AIFs obtained in the three different ROIs (whole brain, hemisphere without lesion and hemisphere with lesion) were similar, the CBF ratios (lesion/normal) between quantitative and semi-quantitative analyses might have a similar trend at different operative time points. If the AIFs were different, the CBF ratios might be different. We concluded that using local maximum one can define proper AIF without knowing the anatomical location of arteries in a stroke rat model.",
keywords = "Data processing methods, MRI (whole body, cardiovascular, breast, others), MRangiography (MRA)",
author = "Yeh, {M. Y.} and Lee, {T. H.} and Yang, {S. T.} and Kuo, {H. H.} and Chyi, {T. K.} and Liu, {H. L.}",
year = "2009",
doi = "10.1088/1748-0221/4/05/P05021",
language = "English",
volume = "4",
journal = "Journal of Instrumentation",
issn = "1748-0221",
publisher = "IOP Publishing Ltd.",
number = "5",

}

TY - JOUR

T1 - Automated detection of arterial input function in DSC perfusion MRI in a stroke rat model

AU - Yeh, M. Y.

AU - Lee, T. H.

AU - Yang, S. T.

AU - Kuo, H. H.

AU - Chyi, T. K.

AU - Liu, H. L.

PY - 2009

Y1 - 2009

N2 - Quantitative cerebral blood flow (CBF) estimation requires deconvolution of the tissue concentration time curves with an arterial input function (AIF). However, image-based determination of AIF in rodent is challenged due to limited spatial resolution. We evaluated the feasibility of quantitative analysis using automated AIF detection and compared the results with commonly applied semi-quantitative analysis. Permanent occlusion of bilateral or unilateral common carotid artery was used to induce cerebral ischemia in rats. The image using dynamic susceptibility contrast method was performed on a 3-T magnetic resonance scanner with a spin-echo echo-planar-image sequence (TR/TE = 700/80 ms, FOV = 41 mm, matrix = 64, 3 slices, SW = 2 mm), starting from 7 s prior to contrast injection (1.2 ml/kg) at four different time points. For quantitative analysis, CBF was calculated by the AIF which was obtained from 10 voxels with greatest contrast enhancement after deconvolution. For semi-quantitative analysis, relative CBF was estimated by the integral divided by the first moment of the relaxivity time curves. We observed if the AIFs obtained in the three different ROIs (whole brain, hemisphere without lesion and hemisphere with lesion) were similar, the CBF ratios (lesion/normal) between quantitative and semi-quantitative analyses might have a similar trend at different operative time points. If the AIFs were different, the CBF ratios might be different. We concluded that using local maximum one can define proper AIF without knowing the anatomical location of arteries in a stroke rat model.

AB - Quantitative cerebral blood flow (CBF) estimation requires deconvolution of the tissue concentration time curves with an arterial input function (AIF). However, image-based determination of AIF in rodent is challenged due to limited spatial resolution. We evaluated the feasibility of quantitative analysis using automated AIF detection and compared the results with commonly applied semi-quantitative analysis. Permanent occlusion of bilateral or unilateral common carotid artery was used to induce cerebral ischemia in rats. The image using dynamic susceptibility contrast method was performed on a 3-T magnetic resonance scanner with a spin-echo echo-planar-image sequence (TR/TE = 700/80 ms, FOV = 41 mm, matrix = 64, 3 slices, SW = 2 mm), starting from 7 s prior to contrast injection (1.2 ml/kg) at four different time points. For quantitative analysis, CBF was calculated by the AIF which was obtained from 10 voxels with greatest contrast enhancement after deconvolution. For semi-quantitative analysis, relative CBF was estimated by the integral divided by the first moment of the relaxivity time curves. We observed if the AIFs obtained in the three different ROIs (whole brain, hemisphere without lesion and hemisphere with lesion) were similar, the CBF ratios (lesion/normal) between quantitative and semi-quantitative analyses might have a similar trend at different operative time points. If the AIFs were different, the CBF ratios might be different. We concluded that using local maximum one can define proper AIF without knowing the anatomical location of arteries in a stroke rat model.

KW - Data processing methods

KW - MRI (whole body, cardiovascular, breast, others), MRangiography (MRA)

UR - http://www.scopus.com/inward/record.url?scp=71049124766&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=71049124766&partnerID=8YFLogxK

U2 - 10.1088/1748-0221/4/05/P05021

DO - 10.1088/1748-0221/4/05/P05021

M3 - Article

AN - SCOPUS:71049124766

VL - 4

JO - Journal of Instrumentation

JF - Journal of Instrumentation

SN - 1748-0221

IS - 5

M1 - P05021

ER -