Multiparametric ISODATA analysis of embolic stroke and rt-PA intervention in rat

https://doi.org/10.1016/j.jns.2004.05.017Get rights and content

Abstract

To increase the sensitivity of MRI parameters to detect tissue damage of ischemic stroke, an unsupervised analysis method, Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA), was applied to analyze the temporal evolution of ischemic damage in a focal embolic cerebral ischemia model in rat with and without recombinant tissue plasminogen activator (rt-PA) treatment.

Male Wistar rats subjected to embolic stroke were investigated using a 7-T MRI system. Rats were randomized into control (n=9) and treated (n=9) groups. The treated rats received rt-PA via a femoral vein at 4 h after onset of embolic ischemia. ISODATA analysis employed parametric maps or weighted images (T1, T2, and diffusion). ISODATA results with parametric maps are superior to ISODATA with weighted images, and both of them were highly correlated with the infarction size measured from the corresponding histological section. At 24 h after embolic stroke, the average map ISODATA lesion sizes were 37.7±7.0 and 39.2±5.6 mm2 for the treated and the control group, respectively. Average histological infarction areas were 37.9±7.4 mm2 for treated rats and 39.4±6.1 mm2 for controls. The R2 values of the linear correlation between map ISODATA and histological data were 0.98 and 0.96 for treated and control rats, respectively. Both histological and map ISODATA data suggest that there is no significant difference in infarction area between non-treated and rt-PA-treated rats when treatment was administered 4 h after the onset of embolic stroke. The ISODATA lesion size analysis was also sensitive to changes of lesion size during acute and subacute stages of stroke.

Our data demonstrate that the multiparameter map ISODATA approach provides a more sensitive quantitation of the ischemic lesion at all time points than image ISODATA and single MRI parametric analysis using T1, T2 or ADCw.

Introduction

Stroke is a major cause of disability and death [1]. Currently, the only successful acute treatment of stroke is thrombolytic therapy using recombinant tissue plasminogen activator (rt-PA) [2], [3]. Diagnostic tools to predict and characterize the efficacy and safety of thrombolysis will greatly improve the management of stroke patients. However, methodologies to evaluate efficacy and risk of thrombolytic treatment have not been fully developed. Since brain tissue undergoes time-dependent heterogeneous histopathological changes after stroke, it is unlikely that a single magnetic resonance imaging (MRI) parameter can characterize the complexity of cerebral tissue at all acute time points [4]. Finding a method to measure ischemic damage after stroke accurately and reproducibly is an important task in diagnostic research.

In the field of MRI, diffusion-, T1- and T2-weighted images (DWI, T1WI, T2WI) can provide complementary information about the status of the cerebral tissue. DWI appears useful for the early detection of stroke [5], [6], [7], [8], [9], [10].

T1, T2 and the apparent diffusion coefficient of water (ADCw) exhibit different temporal profiles in ischemic brain. ADCw declines acutely after onset of ischemia while T2 exhibits a delayed increase compared to ADCw [11], [12]. Conventional T2-weighted MRI provides information on tissue edema [13]. Elevated T2 may identify vasogenic edema and thereby may provide complementary information along with ADCw to distinguish histopathological changes of damaged tissue. By combining information from multiple MRI parameters, it may be possible to determine the histopathological stage in stroke more accurately than using any single MRI parameter.

The multiparametric (multispectral) nature of MRI data for tissue characterization has been reported [14], and various MRI segmentation techniques reviewed [15]. We have recently focused our efforts towards an unsupervised segmentation technique to minimize the need for human interaction and bias. In this method, a vector tissue signature model is used along with a modified version of the Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA). ISODATA incorporates multiparametric MRI images in an iterative, multistep process that assigns the input data into a set of clusters. ISODATA is an objective unsupervised computer segmentation algorithm [16], [17].

Previous studies using ISODATA have employed image-based ISODATA, and not map-based ISODATA. The map ISODATA is more objective and robust than the image ISODATA in the experiment with surface coil due to the maps are free from the surface correction, which has to be done as pre-processing in image ISODATA. In addition, previous preclinical studies used a mechanical versus embolic occlusion of the MCA. In this study, we applied an unsupervised ISODATA method to incorporate three types of MRI parametric maps or weighted images of embolic stroke rats with and without the treatment of rt-PA at 4 h after the onset of ischemia. For the first time, we tested whether the ISODATA method can identify ischemic lesion in the embolic stroke model in rat with or without rt-PA treatment at 4 h after stroke onset, as a negative control, and compared map versus image ISODATA in the analysis of ischemic lesion volume up to 24 h after stroke.

Section snippets

Materials and methods

All studies were performed in accordance with institutional guidelines for animal research under a protocol approved by the Institutional Animal Care and Use Committee (IACUC).

Results

Physiologic parameters of rats after embolism were maintained within the normal range of values.

Fig. 1 presents a typical set of images from a representative rat obtained at 24 h after stroke onset without rt-PA treatment. The top row contains three MRI images: diffusion- (Fig. 1a), T1- (Fig. 1b) and T2- (Fig. 1c) weighted images. MRI parameter maps of ADCw (Fig. 1d), T1 (Fig. 1e) and T2 (Fig. 1f) are shown in the middle row. The bottom row shows the results of map and image ISODATA (Fig. 1g,h)

Discussion

Our data indicate that map ISODATA provides a more accurate measurement of lesion size at 2, 6 and 24 h after embolic stroke in the rat with or without rt-PA treatment than image ISODATA and select MRI parametric maps of T1, T2 and ADCw. The ISODATA both map and image, provides a time independent (for all time points up to 24 h after the onset of embolic stroke) highly accurate, as measured histologically, and objective analysis method to detect the evolution of ischemic damage.

Most previous

Acknowledgements

This work was supported by NINDS grants PO1 NS23393, RO1 NS38292 and HL64766.

References (52)

  • D. LeBihan et al.

    MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders

    Radiology

    (1986)
  • M.E. Moseley et al.

    Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy

    Magn. Reson. Med

    (1990)
  • R.J. Sevick et al.

    Diffusion-weighted MR imaging and T2-weighted MR imaging in acute cerebral ischemia: comparison and correlation with histopathology

    Acta Neurochir., Suppl

    (1990)
  • H. Shimizu et al.

    Early detection of cerebral ischemia by diffusion-weighted MRI after middle cerebral artery occlusion and reperfusion in rats

    J. Neurosurgery

    (1990)
  • C.T.W. Moonen et al.

    Restricted and anisotropic displacement of water in healthy cat brain and in stroke studied by NMR diffusion imaging

    Magn. Reson. Med

    (1991)
  • A.L. Busza et al.

    Diffusion-weighted imaging studies of cerebral ischemia in gerbils: potential relevance to energy failure

    Stroke

    (1992)
  • R.A. Knight et al.

    Temporal evaluation of ischemic damage in rat brain measured by proton nuclear magnetic resonance imaging

    Stroke

    (1991)
  • M. Hoehn-Berlage et al.

    Changes of relaxation times (T1, T2) and apparent diffusion coefficient after permanent middle cerebral artery occlusion in the rat: temporal evolution, regional extent, and comparison with histology

    Magn. Reson. Med

    (1995)
  • M.W. Vannier et al.

    Multispectral analysis of magnetic resonance images

    Radiology

    (1985)
  • M.A. Jacobs et al.

    A model for multiparametric MRI tissue characterization in experimental cerebral ischemia with histological validation in rat: Part 1

    Stroke

    (2001)
  • H. Soltanian-Zadeh et al.

    Semi-supervised segmentation of MRI stroke studies

    Proc. SPIE Medical Imaging: Image Processing Conference, Newport Beach, CA

    (1997 (Feb.))
  • Q. Jiang et al.

    Diffusion-, T2-, and perfusion-weighted nuclear magnetic resonance imaging of middle cerebral artery embolic stroke and recombinant tissue plasminogen activator intervention in the rat

    J. Cereb. Blood Flow Metab

    (1998)
  • J. Frahm et al.

    Rapid NMR imaging of dynamic processes using the FLASH technique

    Magn. Reson. Med

    (1986)
  • D.C. Look et al.

    Time saving in measurement of NMR and EPR relaxation times

    Rev. Sci. Instrum

    (1970)
  • A.P. Crawley et al.

    A comparison of one-shot and recovery methods in T1 imaging

    Magn. Reson. Med

    (1988)
  • N. Gelman et al.

    Interregional variation of longitudinal relaxation rates in human brain at 3.0 T: relation to estimated iron and water contents

    Magn. Reson. Med

    (2000)
  • Cited by (28)

    • Mean microvessel segment length and radius after embolic stroke: Comparison of magnetic resonance imaging (MRI) and laser scanning confocal microscopy (LSCM)

      2011, Brain Research
      Citation Excerpt :

      T2-weighted images were taken as reference images for purposes of co-registration, warping, and determination of ischemic lesion size. For this, we considered all lesions on the post-stroke T2 maps with values above the mean ± 2SD of the contralateral measurements (Bosomtwi et al., 2008; Ding et al., 2004). The mean absolute value of T2 in the normal tissue in the contralateral hemisphere is 55.7 + 10.2 ms (mean + 2SD).

    • Detection of BBB disruption and hemorrhage by Gd-DTPA enhanced MRI after embolic stroke in rat

      2006, Brain Research
      Citation Excerpt :

      It seems difficult to predict and identify all subtypes of HT with MRI using only contrast agent related MRI parameters. Further improvement of prediction and identification of HT may require multiple MRI parameter analysis, such as ISODATA or N-dimension cluster plot analysis (Okada et al., 1994; Ding et al., 2004; Li et al., 2005). Different MRI parameters have different sensitivity in detecting HT.

    • Characterization of cerebral tissue by MRI map ISODATA in embolic stroke in rat

      2006, Brain Research
      Citation Excerpt :

      At 48 h after embolism, combination-treated animals showed a significant reduction of lesion area (P < 0.002), mean signature (P < 0.03) and total lesion severity (P < 0.001) compared with the control group. The significant reduction of volume of cerebral infarction using combination rtPA and 7E3 F(ab′)2 treatment at 4 h after embolic stroke compared to rtPA alone and saline groups has been previously reported (Ding et al., 2004, 2005; Zhang et al., 2003). Our data indicate that map ISODATA provides the signatures to characterize the status of cerebral tissue after ischemia which are well correlated with the histological measurements (R ∼ 0.80, P < 0.0001) for all animals treated with saline, rtPA alone and adjuvant treatment with rtPA and 7E3 F(ab′)2 in both kinds of cell counting, i.e., on damaged cells (H&E) at 6 or 24 h after stroke and on normal neurons (MAP2) at 48 h after stroke.

    View all citing articles on Scopus
    View full text