[Perfusion measurement using the T2* contrast media dynamics in neuro-oncology. Physical basics and clinical applications]

MRI perfusion analysis using freeware, standard imaging software

Background

Perfusion-weighted imaging is only scarcely used in veterinary medicine. The exact reasons are unclear. One reason might be the typically high costs of the software packages for image analysis. In addition, a great variability concerning available programs makes it hard to compare results between different studies. Moreover, these algorithms are tuned for their usage in human medicine and often difficult to adapt to veterinary studies.

In order to address these issues, our aim is to deliver a free open source package for calculating quantitative perfusion parameters. We develop an “R package” calculating mean transit time, cerebral blood flow and cerebral blood volume from data obtained with freely imaging software (OsiriX Light®). We hope that the free availability, in combination with the fact that the underlying algorithm is open and adaptable, makes it easier for scientists in veterinary medicine to use, compare and adapt perfusion-weighted imaging analysis.

In order to demonstrate the usage of our software package, we reviewed previously acquired perfusion-weighted images from a group of eight purpose-breed healthy beagle dogs and twelve client-owned dogs with idiopathic epilepsy. In order to obtain the data needed for our algorithm, the following steps were performed: First, regions of interest (ROI) were drawn around different, previously reported, brain regions and the middle cerebral artery. Second, a ROI enhancement curve was generated for each ROI using a freely available PlugIn. Third, the signal intensity curves were exported as a comma-separated-value file. These files constitute the input to our software package, which then calculates the PWI parameters.

Results

We used our software package to re-assess perfusion weighted images from two previous studies. The clinical results were similar, showing a significant increase in the mean transit time and a significant decrease in cerebral blood flow for diseased dogs.

Conclusion

We provide an “R package” for computing the main perfusion parameters from measurements taken with standard imaging software and describe in detail how to obtain these measurements. We hope that our contribution enables users in veterinary medicine to easily obtain perfusion parameters using standard Open Source software in a standard, adaptable and comparable way.

Validation of vessel size imaging (VSI) in high-grade human gliomas using magnetic resonance imaging, image-guided biopsies, and quantitative immunohistochemistry

To evaluate the association between a vessel size index (VSIMRI) derived from dynamic susceptibility contrast (DSC) perfusion imaging using a custom spin-and-gradient echo echoplanar imaging (SAGE-EPI) sequence and quantitative estimates of vessel morphometry based on immunohistochemistry from image-guided biopsy samples. The current study evaluated both relative cerebral blood volume (rCBV) and VSIMRI in eleven patients with high-grade glioma (7 WHO grade III and 4 WHO grade IV). Following 26 MRI-guided glioma biopsies in these 11 patients, we evaluated tissue morphometry, including vessel density and average radius, using an automated procedure based on the endothelial cell marker CD31 to highlight tumor vasculature. Measures of rCBV and VSIMRI were then compared to histological measures. We demonstrate good agreement between VSI measured by MRI and histology; VSIMRI = 13.67 μm and VSIHistology = 12.60 μm, with slight overestimation of VSIMRI in grade III patients compared to histology. rCBV showed a moderate but significant correlation with vessel density (r = 0.42, p = 0.03), and a correlation was also observed between VSIMRI and VSIHistology (r = 0.49, p = 0.01). The current study supports the hypothesis that vessel size measures using MRI accurately reflect vessel caliber within high-grade gliomas, while traditional measures of rCBV are correlated with vessel density and not vessel caliber.

SVM-based glioma grading: Optimization by feature reduction analysis

We investigated the predictive power of feature reduction analysis approaches in support vector machine (SVM)-based classification of glioma grade. In 101 untreated glioma patients, three analytic approaches were evaluated to derive an optimal reduction in features; (i) Pearson's correlation coefficients (PCC), (ii) principal component analysis (PCA) and (iii) independent component analysis (ICA). Tumor grading was performed using a previously reported SVM approach including whole-tumor cerebral blood volume (CBV) histograms and patient age. Best classification accuracy was found using PCA at 85% (sensitivity=89%, specificity=84%) when reducing the feature vector from 101 (100-bins rCBV histogram+age) to 3 principal components. In comparison, classification accuracy by PCC was 82% (89%, 77%, 2 dimensions) and 79% by ICA (87%, 75%, 9 dimensions). For improved speed (up to 30%) and simplicity, feature reduction by all three methods provided similar classification accuracy to literature values (∼87%) while reducing the number of features by up to 98%.

Support vector machines in DSC-based glioma imaging: suggestions for optimal characterization

Dynamic susceptibility contrast magnetic resonance perfusion imaging (DSC-MRI) is a useful method to characterize gliomas. Recently, support vector machines (SVMs) have been introduced as means to prospectively characterize new patients based on information from previous patients. Based on features derived from automatically segmented tumor volumes from 101 DSC-MR examinations, four different SVM models were compared. All SVM models achieved high prediction accuracies (>82%) after rebalancing the training data sets to equal amounts of samples per class. Best discrimination was obtained using a SVM model with a radial basis function kernel. A correct prediction of low-grade glioma was obtained at 83% (true positive rate) and for high-grade glioma at 91% (true negative rate) on the independent test data set. In conclusion, the combination of automated tumor segmentation followed by SVM classification is feasible. Thereby, a powerful tool is available to characterize glioma presurgically in patients.

Post-processing correction of magnetization transfer effects in FENSI perfusion MRI data

Magnetization transfer effects induced by repetitive saturation pulses employed in flow enhancement of signal intensity imaging sequences currently prevent quantitative, in vivo, cerebral perfusion studies. This study investigates the magnitude of these effects and introduces a post-processing correction protocol. The study shows that the magnetization transfer effect is consistent across individuals, which enables the derivation of a correction factor to be applied in post-acquisition. Our results, obtained for cerebral flux in white and gray matter in rodent brains, are in agreement with cerebral blood flow measurements previously reported in the literature.

Intraprocedural diffusion-weighted PROPELLER MRI to guide percutaneous biopsy needle placement within rabbit VX2 liver tumors

PURPOSE

To test the hypothesis that diffusion-weighted (DW)-PROPELLER (periodically rotated overlapping parallel lines with enhanced reconstruction) magnetic resonance imaging (MRI) can be used to guide biopsy needle placement during percutaneous interventional procedures to selectively target viable and necrotic tissues within VX2 rabbit liver tumors.

MATERIALS AND METHODS

Our institutional Animal Care and Use Committee approved all experiments. In six rabbits implanted with 15 VX2 liver tumors, baseline DW-PROPELLER images acquired prior to the interventional procedure were used for apparent diffusion coefficient (ADC) measurements. Next, intraprocedural DW-PROPELLER scans were performed with needle position iteratively adjusted to target viable, necrotic, or intermediate border tissue regions. DW-PROPELLER ADC measurements at the selected needle tip locations were compared with the percentage of tumor necrosis qualitatively assessed at histopathology.

RESULTS

DW-PROPELLER images demonstrated intratumoral tissue heterogeneity and clearly depicted the needle tip position within viable and necrotic tumor tissues. Mean ADC measurements within the region-of-interest encompassing the needle tip were highly correlated with histopathologic tumor necrotic tissue assessments.

CONCLUSION

DW-PROPELLER is an effective method to selectively position the biopsy needle tip within viable and necrotic tumor tissues. The DW-PROPELLER method may offer an important complementary tool for functional guidance during MR-guided percutaneous procedures.

Funktionelle Bildgebung bei Hirntumoren (Perfusion, DTI, MR-Spektroskopie)

This contribution considers the possibilities involved with using functional methods in magnetic resonance imaging (MRI) diagnostics for brain tumors. Of the functional methods available, we discuss perfusion MRI (PWI), diffusion MRI (DWI and DTI) and MR spectroscopy (H-MRS). In cases of brain tumor, PWI aids in grading and better differentiation in diagnostics as well as for pre-therapeutic planning. In addition, the course of treatment, both after chemo- as well as radiotherapy in combination with surgical treatment, can be optimized. PWI allows better estimates of biological activity and aggressiveness in low grade brain tumors, and in the case of WHO grade II astrocytoma showing anaplasically transformed tumor areas, allows more rapid visu-alization and a better prediction of the course of the disease than conventional MRI diagnostics. Diffusion MRI, due to the directional dependence of the diffusion, can illustrate the course and direction of the nerve fibers, as well as reconstructing the nerve tracts in the cerebrum, pons and cerebellum 3-dimensionally. Diffusion imaging can be used for describing brain tumors, for evaluating contralateral involvement and the course of the nerve fibers near the tumor. Due to its operator dependence, DTI based fiber tracking for defining risk structures is controversial. DWI can also not differentiate accurately between cystic and necrotic brain tumors, or between metastases and brain abscesses. H-MRS provides information on cell membrane metabolism, neuronal integrity and the function of neuronal structures, energy metabolism and the formation of tumors and brain tissue necroses. Diagnostic problems such as the differentiation between neoplastic and non-neoplastic lesions, grading cerebral glioma and distinguishing between primary brain tumors and metastases can be resolved. An additional contribution will discuss the control of the course of glial tumors after radiotherapy.

Non-invasive assessment of vessel morphology and function in tumors by magnetic resonance imaging

The switch to an angiogenic phenotype is an important precondition for tumor growth, invasion and spread. Since newly formed vessels are characterized by structural, functional and molecular abnormalities, they offer promising targets for tumor diagnosis and therapy. Previous studies indicate that MRI is valuable to assess vessel morphology and function. It can be used to distinguish between benign and malignant lesions and to improve delineation of proliferating areas within heterogeneous tumors. In addition, tracer kinetic analysis of contrast-enhanced image series allows the estimation of well-defined physiological parameters such as blood volume, blood flow and vessel permeability. Frequently, changes of these parameters during cytostatic, anti-angiogenic and radiation therapy precede tumor volume reduction. Moreover, target-specific MRI techniques can be used to elucidate the expression of angiogenic markers at the molecular level. This review summarizes strategies for non-invasive characterization of tumor vascularization by functional and molecular MRI, hereby introducing representative preclinical and clinical applications.

MR-Perfusions- und spektroskopische Bildgebung bei WHO-Grad-II-Astrozytomen

BACKGROUND

This study evaluates whether MR perfusion imaging and spectroscopic imaging (MRSI) can depict anaplastic areas in WHO grade II astrocytomas, whether these areas are co-localized, and whether the prognosis can be better predicted.

MATERIAL AND METHODS

Fifteen patients (nine female, six male, aged 42+/-14 years) with WHO grade II astrocytomas but without preceding radio- or chemotherapy were examined every 3 months with MR perfusion imaging and MRSI (mean follow-up 18 months). Using a region of interest analysis, the regional relative cerebral blood volume (rrCBV) and blood flow (rrCBF) were measured in tumor tissue. In the same areas, choline/creatine (Cho/Cr) and choline/N-acetyl-aspartate (Cho/NAA) ratios were quantified.

RESULTS

During follow-up, nine patients had stable disease. In six patients, the tumor showed progression and contrast-enhancement. The progressing tumors had already had higher perfusion (rrCBF 2.1+/-1.4; rrCBV 1.9+/-1.1) parameters than the stable astrocytomas (rrCBF 1.2+/-0.6, p=0.01; rrCBV 1.4+/-0.8, p=0.05) at first examination. However, the Cho/NAA and Cho/Cr ratios only tended to be higher than in stable astrocytomas (Cho/NAA 2.4+/-1.0 vs. 2.0+/-1.5, p=0.23; Cho/Cr 1.7+/-0.6 vs. 1.4+/-0.5, p=0.06). In all six progressing tumors, areas of maximum perfusion and maximum Cho/NAA and Cho/Cr ratio were co-localized. During follow-up, contrast-enhancement was observed in these areas.

CONCLUSIONS

MR perfusion imaging can depict anaplastic areas in WHO grade II astrocytomas earlier than conventional MRI and thus enables a better prediction of prognosis.

Clinical relevance of diffusion and perfusion magnetic resonance imaging in assessing intra-axial brain tumors

Advanced magnetic resonance (MR) imaging techniques provide physiologic information that complements the anatomic information available from conventional MR imaging. We evaluated the roles of diffusion and perfusion imaging for the assessment of grade and type of histologically proven intraaxial brain tumors. A total of 28 patients with intraaxial brain tumors underwent conventional MR imaging (T2- and T1-weighted sequences after gadobenate dimeglumine injection), diffusion imaging and T2*-weighted echo-planar perfusion imaging. Examinations were performed on 19 patients during initial diagnosis and on nine patients during follow-up therapy. Determinations of relative cerebral blood volume (rCBV) and apparent diffusion coefficient (ADC) were performed in the solid parts of each tumor, peritumoral region and contralateral white matter. For gliomas, rCBV values were greater in high-grade than in low-grade tumors (3.87+/-1.94 versus 1.30+/-0.42) at the time of initial diagnosis. rCBV values were increased in all recurrent tumors, except in one patient who presented with a combination of recurrent glioblastoma and massive radionecrosis on histology. Low-grade gliomas had low rCBV even in the presence of contrast medium enhancement. Differentiation between high- and low-grade gliomas was not possible using diffusion-weighted images and ADC values alone. In the peritumoral areas of untreated high-grade gliomas and metastases, the mean rCBV values were higher for high-grade gliomas (1.7+/-0.37) than for metastases (0.54+/-0.18) while the mean ADC values were higher for metastases. The rCBV values of four lymphomas were low and the signal intensity-time curves revealed a significant increase in signal intensity after the first pass of gadobenate dimeglumine. Diffusion and perfusion imaging, even with relatively short imaging and data processing times, provide important information for lesion characterization.