Development of a decision support system for diagnosis and grading of brain tumours using in vivo magnetic resonance single voxel spectra

Manifold Learning in MR spectroscopy using nonlinear dimensionality reduction and unsupervised clustering

PURPOSE

To investigate whether nonlinear dimensionality reduction improves unsupervised classification of (1) H MRS brain tumor data compared with a linear method.

METHODS

In vivo single-voxel (1) H magnetic resonance spectroscopy (55 patients) and (1) H magnetic resonance spectroscopy imaging (MRSI) (29 patients) data were acquired from histopathologically diagnosed gliomas. Data reduction using Laplacian eigenmaps (LE) or independent component analysis (ICA) was followed by k-means clustering or agglomerative hierarchical clustering (AHC) for unsupervised learning to assess tumor grade and for tissue type segmentation of MRSI data.

RESULTS

An accuracy of 93% in classification of glioma grade II and grade IV, with 100% accuracy in distinguishing tumor and normal spectra, was obtained by LE with unsupervised clustering, but not with the combination of k-means and ICA. With (1) H MRSI data, LE provided a more linear distribution of data for cluster analysis and better cluster stability than ICA. LE combined with k-means or AHC provided 91% accuracy for classifying tumor grade and 100% accuracy for identifying normal tissue voxels. Color-coded visualization of normal brain, tumor core, and infiltration regions was achieved with LE combined with AHC.

CONCLUSION

The LE method is promising for unsupervised clustering to separate brain and tumor tissue with automated color-coding for visualization of (1) H MRSI data after cluster analysis.

Multicentre evaluation of the INTERPRET decision support system 2.0 for brain tumour classification

In a previous study, we have shown the added value of (1) H MRS for the neuroradiological characterisation of adult human brain tumours. In that study, several methods of MRS analysis were used, and a software program, the International Network for Pattern Recognition of Tumours Using Magnetic Resonance Decision Support System 1.0 (INTERPRET DSS 1.0), with a short-TE classifier, provided the best results. Since then, the DSS evolved into a version 2.0 that contains an additional long-TE classifier. This study has two objectives. First, to determine whether clinicians with no experience of spectroscopy are comparable with spectroscopists in the use of the system, when only minimum training in the use of the system was given. Second, to assess whether or not a version with another TE is better than the initial version. We undertook a second study with the same cases and nine evaluators to assess whether the diagnostic accuracy of DSS 2.0 was comparable with the values obtained with DSS 1.0. In the second study, the analysis protocol was flexible in comparison with the first one to mimic a clinical environment. In the present study, on average, each case required 5.4 min by neuroradiologists and 9 min by spectroscopists for evaluation. Most classes and superclasses of tumours gave the same results as with DSS 1.0, except for astrocytomas of World Health Organization (WHO) grade III, in which performance measured as the area under the curve (AUC) decreased: AUC = 0.87 (0.72-1.02) with DSS 1.0 and AUC = 0.62 (0.55-0.70) with DSS 2.0. When analysing the performance of radiologists and spectroscopists with respect to DSS 1.0, the results were the same for most classes. Having data with two TEs instead of one did not affect the results of the evaluation.

Fast spectroscopic multiple analysis (FASMA) for brain tumor classification: a clinical decision support system utilizing multi-parametric 3T MR data

INTRODUCTION

A clinical decision support system (CDSS) for brain tumor classification can be used to assist in the diagnosis and grading of brain tumors. A Fast Spectroscopic Multiple Analysis (FASMA) system that uses combinations of multiparametric MRI data sets was developed as a CDSS for brain tumor classification.

METHODS

MRI metabolic ratios and spectra, from long and short TE, respectively, as well as diffusion and perfusion data were acquired from the intratumoral and peritumoral area of 126 patients with untreated intracranial tumors. These data were categorized based on the pathology, and different machine learning methods were evaluated regarding their classification performance for glioma grading and differentiation of infiltrating versus non-infiltrating lesions. Additional databases were embedded to the system, including updated literature values of the related MR parameters and typical tumor characteristics (imaging and histological), for further comparisons. Custom Graphical User Interface (GUI) layouts were developed to facilitate classification of the unknown cases based on the user's available MR data.

RESULTS

The highest classification performance was achieved with a support vector machine (SVM) using the combination of all MR features. FASMA correctly classified 89 and 79% in the intratumoral and peritumoral area, respectively, for cases from an independent test set. FASMA produced the correct diagnosis, even in the misclassified cases, since discrimination between infiltrative versus non-infiltrative cases was possible.

CONCLUSIONS

FASMA is a prototype CDSS, which integrates complex quantitative MR data for brain tumor characterization. FASMA was developed as a diagnostic assistant that provides fast analysis, representation and classification for a set of MR parameters. This software may serve as a teaching tool on advanced MRI techniques, as it incorporates additional information regarding typical tumor characteristics derived from the literature.

(1) H 2D MRSI tissue type analysis of gliomas

PURPOSE

To decompose 1H MR spectra of glioma patients into normal and abnormal tissue proportions for tumor classification and delineation.

METHODS

Anatomical imaging and 1H magnetic resonance spectroscopic imaging data have been acquired from 11 grade II and 13 grade IV glioma patients. LCModel was used to decompose the magnetic resonance spectroscopic imaging data into normal brain, grade II, and grade IV tissue proportions using a tissue type basis set. Simulations were conducted to evaluate the accuracy of the methodology. Results were visualized using colormaps and abnormality contours showing tumor grade and extent.

RESULTS

Simulations suggest that infiltrative tumor proportions as low as 20% can be identified at the typical 1H magnetic resonance spectroscopy signal-to-noise found in vivo. Tumor grading according to the highest estimated tumor grade within a lesion gave a classification accuracy of 86% discriminating between grade II and grade IV glioma. Voxels with significant proportions of tumor type spectra were found beyond the margins of contrast enhancement for most grade IV cases consistent with infiltration whereas the abnormality contours show that some tumors are confined within the hyperintensities shown by both post contrast T1 weighted and T2 weighted imaging.

CONCLUSION

LCModel can be used to decompose 1H MR spectra into proportions of normal and abnormal tissue to identify tumor extent, infiltration, and overall grade.

Clinical decision support systems for brain tumor characterization using advanced magnetic resonance imaging techniques

In recent years, advanced magnetic resonance imaging (MRI) techniques, such as magnetic resonance spectroscopy, diffusion weighted imaging, diffusion tensor imaging and perfusion weighted imaging have been used in order to resolve demanding diagnostic problems such as brain tumor characterization and grading, as these techniques offer a more detailed and non-invasive evaluation of the area under study. In the last decade a great effort has been made to import and utilize intelligent systems in the so-called clinical decision support systems (CDSS) for automatic processing, classification, evaluation and representation of MRI data in order for advanced MRI techniques to become a part of the clinical routine, since the amount of data from the aforementioned techniques has gradually increased. Hence, the purpose of the current review article is two-fold. The first is to review and evaluate the progress that has been made towards the utilization of CDSS based on data from advanced MRI techniques. The second is to analyze and propose the future work that has to be done, based on the existing problems and challenges, especially taking into account the new imaging techniques and parameters that can be introduced into intelligent systems to significantly improve their diagnostic specificity and clinical application.

Clinical proton MR spectroscopy in central nervous system disorders

A large body of published work shows that proton (hydrogen 1 [(1)H]) magnetic resonance (MR) spectroscopy has evolved from a research tool into a clinical neuroimaging modality. Herein, the authors present a summary of brain disorders in which MR spectroscopy has an impact on patient management, together with a critical consideration of common data acquisition and processing procedures. The article documents the impact of (1)H MR spectroscopy in the clinical evaluation of disorders of the central nervous system. The clinical usefulness of (1)H MR spectroscopy has been established for brain neoplasms, neonatal and pediatric disorders (hypoxia-ischemia, inherited metabolic diseases, and traumatic brain injury), demyelinating disorders, and infectious brain lesions. The growing list of disorders for which (1)H MR spectroscopy may contribute to patient management extends to neurodegenerative diseases, epilepsy, and stroke. To facilitate expanded clinical acceptance and standardization of MR spectroscopy methodology, guidelines are provided for data acquisition and analysis, quality assessment, and interpretation. Finally, the authors offer recommendations to expedite the use of robust MR spectroscopy methodology in the clinical setting, including incorporation of technical advances on clinical units.

A novel semi-supervised methodology for extracting tumor type-specific MRS sources in human brain data

Background

The clinical investigation of human brain tumors often starts with a non-invasive imaging study, providing information about the tumor extent and location, but little insight into the biochemistry of the analyzed tissue. Magnetic Resonance Spectroscopy can complement imaging by supplying a metabolic fingerprint of the tissue. This study analyzes single-voxel magnetic resonance spectra, which represent signal information in the frequency domain. Given that a single voxel may contain a heterogeneous mix of tissues, signal source identification is a relevant challenge for the problem of tumor type classification from the spectroscopic signal.

Methodology/Principal Findings

Non-negative matrix factorization techniques have recently shown their potential for the identification of meaningful sources from brain tissue spectroscopy data. In this study, we use a convex variant of these methods that is capable of handling negatively-valued data and generating sources that can be interpreted as tumor class prototypes. A novel approach to convex non-negative matrix factorization is proposed, in which prior knowledge about class information is utilized in model optimization. Class-specific information is integrated into this semi-supervised process by setting the metric of a latent variable space where the matrix factorization is carried out. The reported experimental study comprises 196 cases from different tumor types drawn from two international, multi-center databases. The results indicate that the proposed approach outperforms a purely unsupervised process by achieving near perfect correlation of the extracted sources with the mean spectra of the tumor types. It also improves tissue type classification.

Conclusions/Significance

We show that source extraction by unsupervised matrix factorization benefits from the integration of the available class information, so operating in a semi-supervised learning manner, for discriminative source identification and brain tumor labeling from single-voxel spectroscopy data. We are confident that the proposed methodology has wider applicability for biomedical signal processing.

Diagnosing relapse in children's brain tumors using metabolite profiles

BACKGROUND

Malignant brain tumors in children generally have a very poor prognosis when they relapse and improvements are required in their management. It can be difficult to accurately diagnose abnormalities detected during tumor surveillance, and new techniques are required to aid this process. This study investigates how metabolite profiles measured noninvasively by (1)H magnetic resonance spectroscopy (MRS) at relapse reflect those at diagnosis and may be used in this monitoring process.

METHODS

Single-voxel MRS (1.5 T, point-resolved spectroscopy, echo time 30 ms, repetition time 1500 ms was performed on 19 children with grades II-IV brain tumors during routine MRI scans prior to treatment for a suspected brain tumor and at suspected first relapse. MRS was analyzed using TARQUIN software to provide metabolite concentrations. Paired Student's t-tests were performed between metabolite profiles at diagnosis and at first relapse.

RESULTS

There was no significant difference (P > .05) in the level of any metabolite, lipid, or macromolecule from tumors prior to treatment and at first relapse. This was true for the whole group (n = 19), those with a local relapse (n = 12), and those with a distant relapse (n = 7). Lipids at 1.3 ppm were close to significance when comparing the level at diagnosis with that at distant first relapse (P = .07, 6.5 vs 12.9). In 5 cases the MRS indicative of tumor preceded a formal diagnosis of relapse.

CONCLUSIONS

Tumor metabolite profiles, measured by MRS, do not change greatly from diagnosis to first relapse, and this can aid the confirmation of the presence of tumor.

Glutamate and glutamine: a review of in vivo MRS in the human brain

Our understanding of the roles that the amino acids glutamate (Glu) and glutamine (Gln) play in the mammalian central nervous system has increased rapidly in recent times. Many conditions are known to exhibit a disturbance in Glu-Gln equilibrium, and the exact relationships between these changed conditions and these amino acids are not fully understood. This has led to increased interest in Glu/Gln quantitation in the human brain in an array of conditions (e.g. mental illness, tumor, neuro-degeneration) as well as in normal brain function. Accordingly, this review has been undertaken to describe the increasing number of in vivo techniques available to study Glu and Gln separately, or pooled as 'Glx'. The present MRS methods used to assess Glu and Gln vary in approach, complexity, and outcome, thus the focus of this review is on a description of MRS acquisition approaches, and an indication of relative utility of each technique rather than brain pathologies associated with Glu and/or Gln perturbation. Consequently, this review focuses particularly on (1) one-dimensional (1)H MRS, (2) two-dimensional (1)H MRS, and (3) one-dimensional (13)C MRS techniques.

Randomized pilot study and qualitative evaluation of a clinical decision support system for brain tumour diagnosis based on SV ¹H MRS: evaluation as an additional information procedure for novice radiologists

The results of a randomized pilot study and qualitative evaluation of the clinical decision support system Curiam BT are reported. We evaluated the system's feasibility and potential value as a radiological information procedure complementary to magnetic resonance (MR) imaging to assist novice radiologists in diagnosing brain tumours using MR spectroscopy (1.5 and 3.0T). Fifty-five cases were analysed at three hospitals according to four non-exclusive diagnostic questions. Our results show that Curiam BT improved the diagnostic accuracy in all the four questions. Additionally, we discuss the findings of the users' feedback about the system, and the further work to optimize it for real environments and to conduct a large clinical trial.

Multivoxel ¹H MR spectroscopy is superior to contrast-enhanced MRI for response assessment after anti-angiogenic treatment of orthotopic human glioma xenografts and provides handles for metabolic targeting

BACKGROUND

Anti-angiogenic treatment of glioblastoma characteristically results in therapy resistance and tumor progression via diffuse infiltration. Monitoring tumor progression in these patients is thwarted because therapy results in tumor invisibility in contrast-enhanced (CE) MRI. To address this problem, we examined whether tumor progression could be monitored by metabolic mapping using (1)H MR spectroscopic imaging (MRSI).

METHODS

We treated groups of BALB/c nu/nu mice carrying different orthotopic diffuse-infiltrative glioblastoma xenografts with bevacizumab (anti-vascular endothelial growth factor [VEGF] antibody, n = 13), cabozantinib (combined VEGF receptor 2/c-Met tyrosine kinase inhibitor, n = 11), or placebo (n = 15) and compared CE-MRI with MRS-derived metabolic maps before, during, and after treatment. Metabolic maps and CE-MRIs were subsequently correlated to histology and immunohistochemistry.

RESULTS

In vivo imaging of choline/n-acetyl aspartate ratios via multivoxel MRS is better able to evaluate response to therapy than CE-MRI. Lactate imaging revealed that diffuse infiltrative areas in glioblastoma xenografts did not present with excessive glycolysis. In contrast, glycolysis was observed in hypoxic areas in angiogenesis-dependent compact regions of glioma only, especially after anti-angiogenic treatment.

CONCLUSION

Our data present MRSI as a powerful and feasible approach that is superior to CE-MRI and may provide handles for optimizing treatment of glioma. Furthermore, we show that glycolysis is more prominent in hypoxic areas than in areas of diffuse infiltrative growth. The Warburg hypothesis of persisting glycolysis in tumors under normoxic conditions may thus not be valid for diffuse glioma.

Classification methods for the differentiation of atypical meningiomas using diffusion and perfusion techniques at 3-T MRI

The purpose was to investigate the contribution of machine learning algorithms using diffusion and perfusion techniques in the differentiation of atypical meningiomas from glioblastomas and metastases. Apparent diffusion coefficient, fractional anisotropy, and relative cerebral blood volume were measured in different tumor regions. Naive Bayes, k-Nearest Neighbor, and Support Vector Machine classifiers were used in the classification procedure. The application of classification methods adds incremental differential diagnostic value. Differentiation is mainly achieved using diffusion metrics, while perfusion measurements may provide significant information for the peritumoral regions.

Lipid and macromolecules quantitation in differentiating glioblastoma from solitary metastasis: a short-echo time single-voxel magnetic resonance spectroscopy study at 3 T

OBJECTIVE

The differentiation between solitary metastasis (MET) and glioblastoma (GBM) is difficult using only magnetic resonance imaging techniques. Magnetic resonance spectroscopy (MRS) lipid signal indicates cellular necrosis both in GBMs and METs. The purpose of this prospective study was to determine whether a class of lipids and/or macromolecules (MMs), able to efficiently discriminate between these two types of lesions, exists.

METHODS

Forty-one patients with solitary brain tumor (23 GBMs and 18 METs) underwent magnetic resonance imaging and single-voxel MRS. Short-echo time point resolved spectroscopy sequence acquisition with water suppression technique was used. Spectra were analyzed using LCModel. Absolute quantification was performed with "water-scaling" procedure. The analysis was focused on sums of lipid and macromolecular (LM) components at 0.9 and 1.3 ppm.

RESULTS

The LM13 absolute concentration was statistically different (P < 0.0001) between GBMs and METs. With a cutoff of 81 mM in LM13 absolute concentration, METs and GBMs can be distinguished with a 78% of specificity and an 81% of sensitivity. The presence of the MM12 peak, related to the fucose II complex, in tumors harboring a K-ras gene mutation has been investigated.

CONCLUSIONS

We exploited the performance of a clinically easily implementable method, such as short-echo time single-voxel MRS, for the differentiation between brain metastasis and primary brain tumors. The study showed that MRS absolute lipid and macromolecular signals could be helpful in differentiating GBM from metastasis. LM13 class was found to be a discriminant parameter with an accuracy of 85%. Detection of the MM12-fucose peak may also have a role in understanding molecular biology of brain metastasis and should be further investigated to address specific metabolic phenotypes.

Extracting MRS discriminant functional features of brain tumors

The current challenge in automatic brain tumor classification based on MRS is the improvement of the robustness of the classification models that explicitly account for the probable breach of the independent and identically distributed conditions in the MRS data points. To contribute to this purpose, a new algorithm for the extraction of discriminant MRS features of brain tumors based on a functional approach is presented. Functional data analysis based on region segmentation (RSFDA) is based on the functional data analysis formalism using nonuniformly distributed B splines according to spectral regions that are highly correlated. An exhaustive characterization of the method is presented in this work using controlled and real scenarios. The performance of RSFDA was compared with other widely used feature extraction methods. In all simulated conditions, RSFDA was proven to be stable with respect to the number of variables selected and with respect to the classification performance against noise and baseline artifacts. Furthermore, with real multicenter datasets classification, RSFDA and peak integration (PI) obtained better performance than the other feature extraction methods used for comparison. Other advantages of the method proposed are its usefulness in selecting the optimal number of features for classification and its simplified functional representation of the spectra, which contributes to highlight the discriminative regions of the MR spectrum for each classification task.

Preoperative magnetic resonance spectroscopy improves diagnostic accuracy in a series of neurosurgical dilemmas

OBJECT

The purpose of this study was to evaluate the usefulness of preoperative magnetic resonance spectroscopy (MRS) in neurosurgical patients with diagnostically challenging intracranial lesions.

METHODS

Included in this study are twenty-three consecutive patients presenting to the neurosurgery service with diagnostically challenging intracranial lesions and who were investigated by conventional MR imaging and proton ((1)H) MRS, followed by surgery with subsequent histopathological diagnosis. An experienced neuroradiologist (RJ) blinded to the final histopathology evaluated the imaging studies retrospectively. Provisional diagnoses based on preoperative clinical and conventional MR data versus preoperative MRS data were compared with definitive histopathological diagnoses.

RESULTS

Compared with preoperative clinical and conventional MR data, (1)H MRS improved the accuracy of MR imaging from 60.9% to 83%. We found (1)H MRS reliably distinguished between abscess and high-grade tumour, and between high-grade glioma and low-grade glioma, but was not able to reliably distinguish between recurrent glioma and radiation necrosis. In 12/23 cases (52%) the (1)H MRS findings positively altered our clinical management. Two representative cases are presented.

CONCLUSIONS

Our study supports a beneficial role for (1)H MRS in certain diagnostic intracranial dilemmas presenting to neurosurgeons. The information gleaned from preoperative (1)H MRS can be a useful adjunct to clinical and conventional MR imaging data in guiding the management of patients with intracranial pathologies, particularly high-grade tumour versus abscess, and high-grade versus low-grade glioma. Further larger prospective studies are needed to clearly define the utility of (1)H MRS in diagnostically challenging intracranial lesions in neurosurgery.

Dimethyl sulfoxide (DMSO) as a potential contrast agent for brain tumors

Dimethyl sulfoxide (DMSO) is commonly used in preclinical studies of animal models of high-grade glioma as a solvent for chemotherapeutic agents. A strong DMSO signal was detected by single-voxel MRS in the brain of three C57BL/6 control mice during a pilot study of DMSO tolerance after intragastric administration. This led us to investigate the accumulation and wash-out kinetics of DMSO in both normal brain parenchyma (n=3 control mice) by single-voxel MRS, and in 12 GL261 glioblastomas (GBMs) by single-voxel MRS (n=3) and MRSI (n=9). DMSO accumulated differently in each tissue type, reaching its highest concentration in tumors: 6.18 ± 0.85 µmol/g water, 1.5-fold higher than in control mouse brain (p<0.05). A faster wash-out was detected in normal brain parenchyma with respect to GBM tissue: half-lives of 2.06 ± 0.58 and 4.57 ± 1.15 h, respectively. MRSI maps of time-course DMSO changes revealed clear hotspots of differential spatial accumulation in GL261 tumors. Additional MRSI studies with four mice bearing oligodendrogliomas (ODs) revealed similar results as in GBM tumors. The lack of T(1) contrast enhancement post-gadolinium (gadopentetate dimeglumine, Gd-DTPA) in control mouse brain and mice with ODs suggested that DMSO was fully able to cross the intact blood-brain barrier in both normal brain parenchyma and in low-grade tumors. Our results indicate a potential role for DMSO as a contrast agent for brain tumor detection, even in those tumors 'invisible' to standard gadolinium-enhanced MRI, and possibly for monitoring heterogeneities associated with progression or with therapeutic response.

Automated differentiation of glioblastomas from intracranial metastases using 3T MR spectroscopic and perfusion data

PURPOSE

Differentiation of glioblastomas from metastases is clinical important, but may be difficult even for expert observers. To investigate the contribution of machine learning algorithms in the differentiation of glioblastomas multiforme (GB) from metastases, we developed and tested a pattern recognition system based on 3T magnetic resonance (MR) data.

MATERIALS AND METHODS

Single and multi-voxel proton magnetic resonance spectroscopy (1H-MRS) and dynamic susceptibility contrast (DSC) MRI scans were performed on 49 patients with solitary brain tumors (35 glioblastoma multiforme and 14 metastases). Metabolic (NAA/Cr, Cho/Cr, (Lip [Formula: see text] Lac)/Cr) and perfusion (rCBV) parameters were measured in both intratumoral and peritumoral regions. The statistical significance of these parameters was evaluated. For the classification procedure, three datasets were created to find the optimum combination of parameters that provides maximum differentiation. Three machine learning methods were utilized: Naïve-Bayes, Support Vector Machine (SVM) and [Formula: see text]-nearest neighbor (KNN). The discrimination ability of each classifier was evaluated with quantitative performance metrics.

RESULTS

Glioblastoma and metastases were differentiable only in the peritumoral region of these lesions ([Formula: see text]). SVM achieved the highest overall performance (accuracy 98%) for both the intratumoral and peritumoral areas. Naïve-Bayes and KNN presented greater variations in performance. The proper selection of datasets plays a very significant role as they are closely correlated to the underlying pathophysiology.

CONCLUSION

The application of pattern recognition techniques using 3T MR-based perfusion and metabolic features may provide incremental diagnostic value in the differentiation of common intraaxial brain tumors, such as glioblastoma versus metastasis.

Strategies for annotation and curation of translational databases: the eTUMOUR project

The eTUMOUR (eT) multi-centre project gathered in vivo and ex vivo magnetic resonance (MR) data, as well as transcriptomic and clinical information from brain tumour patients, with the purpose of improving the diagnostic and prognostic evaluation of future patients. In order to carry this out, among other work, a database—the eTDB—was developed. In addition to complex permission rules and software and management quality control (QC), it was necessary to develop anonymization, processing and data visualization tools for the data uploaded. It was also necessary to develop sophisticated curation strategies that involved on one hand, dedicated fields for QC-generated meta-data and specialized queries and global permissions for senior curators and on the other, to establish a set of metrics to quantify its contents. The indispensable dataset (ID), completeness and pairedness indices were set. The database contains 1317 cases created as a result of the eT project and 304 from a previous project, INTERPRET. The number of cases fulfilling the ID was 656. Completeness and pairedness were heterogeneous, depending on the data type involved.

Functional imaging in adult and paediatric brain tumours

Imaging is a key component in the management of brain tumours, with MRI being the preferred modality for most clinical scenarios. However, although conventional MRI provides mainly structural information, such as tumour size and location, it leaves many important clinical questions, such as tumour type, aggressiveness and prognosis, unanswered. An increasing number of studies have shown that additional information can be obtained using functional imaging methods (which probe tissue properties), and that these techniques can give key information of clinical importance. These techniques include diffusion imaging, which can assess tissue structure, and perfusion imaging and magnetic resonance spectroscopy, which measures tissue metabolite profiles. Tumour metabolism can also be investigated using PET, with 18F-deoxyglucose being the most readily available tracer. This Review discusses these methods and the studies that have investigated their clinical use. A strong emphasis is placed on the measurement of quantitative parameters, which is a move away from the qualitative nature of conventional radiological reporting and presents major challenges, particularly for multicentre studies.

Accurate classification of childhood brain tumours by in vivo ¹H MRS - a multi-centre study

AIMS

To evaluate the accuracy of single-voxel Magnetic Resonance Spectroscopy ((1)H MRS) as a non-invasive diagnostic aid for paediatric brain tumours in a multi-national study. Our hypotheses are (1) that automated classification based on (1)H MRS provides an accurate non-invasive diagnosis in multi-centre datasets and (2) using a protocol which increases the metabolite information improves the diagnostic accuracy.

METHODS

Seventy-eight patients under 16 years old with histologically proven brain tumours from 10 international centres were investigated. Discrimination of 29 medulloblastomas, 11 ependymomas and 38 pilocytic astrocytomas (PILOAs) was evaluated. Single-voxel MRS was undertaken prior to diagnosis (1.5 T Point-Resolved Spectroscopy (PRESS), Proton Brain Exam (PROBE) or Stimulated Echo Acquisition Mode (STEAM), echo time (TE) 20-32 ms and 135-136 ms). MRS data were processed using two strategies, determination of metabolite concentrations using TARQUIN software and automatic feature extraction with Peak Integration (PI). Linear Discriminant Analysis (LDA) was applied to this data to produce diagnostic classifiers. An evaluation of the diagnostic accuracy was performed based on resampling to measure the Balanced Accuracy Rate (BAR).

RESULTS

The accuracy of the diagnostic classifiers for discriminating the three tumour types was found to be high (BAR 0.98) when a combination of TE was used. The combination of both TEs significantly improved the classification performance (p<0.01, Tukey's test) compared with the use of one TE alone. Other tumour types were classified accurately as glial or primitive neuroectodermal (BAR 1.00).

CONCLUSION

(1)H MRS has excellent accuracy for the non-invasive diagnosis of common childhood brain tumours particularly if the metabolite information is maximised and should become part of routine clinical assessment for these children.

Classification of single-voxel 1H spectra of childhood cerebellar tumors using LCModel and whole tissue representations

In this study, mean tumor spectra are used as the basis functions in LCModel to create a direct classification tool for short echo time (1)H magnetic resonance spectroscopy of pediatric brain tumors. LCModel is a widely used analysis tool designed to fit a linear combination of individual metabolite spectra to in vivo spectra. Here, we have used LCModel to fit mean spectra and corresponding variability components of childhood cerebellar tumors, as calculated using principal component analysis, and assessed for classification accuracy. Classification was performed according to the highest estimated tumor proportion. This method was tested in a leave-one-out analysis discriminating between pediatric brain tumor spectra of medulloblastoma vs. pilocytic astrocytoma and medulloblastoma vs. pilocytic astrocytoma vs. ependymoma. Additionally, the effect of accepting different Cramér-Rao Lower Bound cut-off criteria on classification accuracy and estimated tissue proportions was investigated. The best classification results differentiating medulloblastoma vs. pilocytic astrocytoma and medulloblastoma vs. pilocytic astrocytoma vs. ependymoma were 100 and 87.7%, respectively. These results are comparable to a specialized pattern recognition analysis of this data set and give easy to interpret results in the form of estimated tissue proportions. The method requires minimal user input and is easily transferable across sites and to other magnetic resonance spectroscopy classification problems.

Quantitative short echo time 1H MRSI of the peripheral edematous region of human brain tumors in the differentiation between glioblastoma, metastasis, and meningioma

PURPOSE

To assess metabolite levels in peritumoral edematous (PO) and surrounding apparently normal (SAN) brain regions of glioblastoma, metastasis, and meningioma in humans with (1)H-MRSI to find biomarkers that can discriminate between tumors and characterize infiltrative tumor growth.

MATERIALS AND METHODS

Magnetic resonance (MR) spectra (semi-LASER MRSI, 30 msec echo time, 3T) were selected from regions of interest (ROIs) under MRI guidance, and after quality control of MR spectra. Statistical testing between patient groups was performed for mean metabolite ratios of an entire ROI and for the highest value within that ROI.

RESULTS

The highest ratios of the level of choline compounds and the sum of myo-inositol and glycine over N-acetylaspartate and creatine compounds were significantly increased in PO regions of glioblastoma versus that of metastasis and meningioma. In the SAN region of glioblastoma some of these ratios were increased. Differences were less prominent for metabolite levels averaged over entire ROIs.

CONCLUSION

Specific metabolite ratios in PO and SAN regions can be used to discriminate glioblastoma from metastasis and meningioma. An analysis of these ratios averaged over entire ROIs and those with most abnormal values indicates that infiltrative tumor growth in glioblastoma is inhomogeneous and extends into the SAN region.

Robust discrimination of glioblastomas from metastatic brain tumors on the basis of single-voxel (1)H MRS

This article investigates methods for the accurate and robust differentiation of metastases from glioblastomas on the basis of single-voxel (1)H MRS information. Single-voxel (1)H MR spectra from a total of 109 patients (78 glioblastomas and 31 metastases) from the multicenter, international INTERPRET database, plus a test set of 40 patients (30 glioblastomas and 10 metastases) from three different centers in the Barcelona (Spain) metropolitan area, were analyzed using a robust method for feature (spectral frequency) selection coupled with a linear-in-the-parameters single-layer perceptron classifier. For the test set, a parsimonious selection of five frequencies yielded an area under the receiver operating characteristic curve of 0.86, and an area under the convex hull of the receiver operating characteristic curve of 0.91. Moreover, these accurate results for the discrimination between glioblastomas and metastases were obtained using a small number of frequencies that are amenable to metabolic interpretation, which should ease their use as diagnostic markers. Importantly, the prediction can be expressed as a simple formula based on a linear combination of these frequencies. As a result, new cases could be straightforwardly predicted by integrating this formula into a computer-based medical decision support system. This work also shows that the combination of spectra acquired at different TEs (short TE, 20-32 ms; long TE, 135-144 ms) is key to the successful discrimination between glioblastomas and metastases from single-voxel (1)H MRS.

Using magnetic resonance imaging and spectroscopy in cancer diagnostics and monitoring: preclinical and clinical approaches

Nuclear Magnetic Resonance (MR) based imaging has become an integrated domain in today's oncology research and clinical management of cancer patients. MR is a unique imaging modality among numerous other imaging modalities by providing access to anatomical, physiological, biochemical and molecular details of tumour with excellent spatial and temporal resolutions. In this review we will cover established and investigational MR imaging (MRI) and MR spectroscopy (MRS) techniques used for cancer imaging and demonstrate wealth of information on tumour biology and clinical applications MR techniques offer for oncology research both in preclinical and clinical settings. Emphasis is given not only to the variety of information which may be obtained but also the complementary nature of the techniques. This ability to determine tumour type, grade, invasiveness, degree of hypoxia, microvacular characteristics, and metabolite phenotype, has already profoundly transformed oncology research and patient management. It is evident from the data reviewed that MR techniques will play a key role in uncovering molecular fingerprints of cancer, developing targeted treatment strategies and assessing responsiveness to treatment for personalized patient management, thereby allowing rapid translation of imaging research conclusions into the benefit of clinical oncology.

Prospective diagnostic performance evaluation of single-voxel 1H MRS for typing and grading of brain tumours

The purpose of this study was to evaluate whether single-voxel (1)H MRS could add useful information to conventional MRI in the preoperative characterisation of the type and grade of brain tumours. MRI and MRS examinations from a prospective cohort of 40 consecutive patients were analysed double blind by radiologists and spectroscopists before the histological diagnosis was known. The spectroscopists had only the MR spectra, whereas the radiologists had both the MR images and basic clinical details (age, sex and presenting symptoms). Then, the radiologists and spectroscopists exchanged their predictions and re-evaluated their initial opinions, taking into account the new evidence. Spectroscopists used four different systems of analysis for (1)H MRS data, and the efficacy of each of these methods was also evaluated. Information extracted from (1)H MRS significantly improved the radiologists' MRI-based characterisation of grade IV tumours (glioblastomas, metastases, medulloblastomas and lymphomas) in the cohort [area under the curve (AUC) in the MRI re-evaluation 0.93 versus AUC in the MRI evaluation 0.85], and also of the less malignant glial tumours (AUC in the MRI re-evaluation 0.93 versus AUC in the MRI evaluation 0.81). One of the MRS analysis systems used, the INTERPRET (International Network for Pattern Recognition of Tumours Using Magnetic Resonance) decision support system, outperformed the others, as well as being better than the MRI evaluation for the characterisation of grade III astrocytomas. Thus, preoperative MRS data improve the radiologists' performance in diagnosing grade IV tumours and, for those of grade II-III, MRS data help them to recognise the glial lineage. Even in cases in which their diagnoses were not improved, the provision of MRS data to the radiologists had no negative influence on their predictions.

Magnetic resonance spectroscopy of cancer metabolism and response to therapy

Magnetic resonance spectroscopy allows noninvasive in vivo measurements of biochemical information from living systems, ranging from cultured cells through experimental animals to humans. Studies of biopsies or extracts offer deeper insights by detecting more metabolites and resolving metabolites that cannot be distinguished in vivo. The pharmacokinetics of certain drugs, especially fluorinated drugs, can be directly measured in vivo. This review briefly describes these methods and their applications to cancer metabolism, including glycolysis, hypoxia, bioenergetics, tumor pH, and tumor responses to radiotherapy and chemotherapy.

Non-negative matrix factorisation methods for the spectral decomposition of MRS data from human brain tumours

BACKGROUND

In-vivo single voxel proton magnetic resonance spectroscopy (SV 1H-MRS), coupled with supervised pattern recognition (PR) methods, has been widely used in clinical studies of discrimination of brain tumour types and follow-up of patients bearing abnormal brain masses. SV 1H-MRS provides useful biochemical information about the metabolic state of tumours and can be performed at short (< 45 ms) or long (> 45 ms) echo time (TE), each with particular advantages. Short-TE spectra are more adequate for detecting lipids, while the long-TE provides a much flatter signal baseline in between peaks but also negative signals for metabolites such as lactate. Both, lipids and lactate, are respectively indicative of specific metabolic processes taking place. Ideally, the information provided by both TE should be of use for clinical purposes. In this study, we characterise the performance of a range of Non-negative Matrix Factorisation (NMF) methods in two respects: first, to derive sources correlated with the mean spectra of known tissue types (tumours and normal tissue); second, taking the best performing NMF method for source separation, we compare its accuracy for class assignment when using the mixing matrix directly as a basis for classification, as against using the method for dimensionality reduction (DR). For this, we used SV 1H-MRS data with positive and negative peaks, from a widely tested SV 1H-MRS human brain tumour database.

RESULTS

The results reported in this paper reveal the advantage of using a recently described variant of NMF, namely Convex-NMF, as an unsupervised method of source extraction from SV1H-MRS. Most of the sources extracted in our experiments closely correspond to the mean spectra of some of the analysed tumour types. This similarity allows accurate diagnostic predictions to be made both in fully unsupervised mode and using Convex-NMF as a DR step previous to standard supervised classification. The obtained results are comparable to, or more accurate than those obtained with supervised techniques.

CONCLUSIONS

The unsupervised properties of Convex-NMF place this approach one step ahead of classical label-requiring supervised methods for the discrimination of brain tumour types, as it accounts for their increasingly recognised molecular subtype heterogeneity. The application of Convex-NMF in computer assisted decision support systems is expected to facilitate further improvements in the uptake of MRS-derived information by clinicians.

Identifying malignant transformations in recurrent low grade gliomas using high resolution magic angle spinning spectroscopy

OBJECTIVE

The objective of this study was to determine whether metabolic parameters derived from ex vivo analysis of tissue samples are predictive of biologic characteristics of recurrent low grade gliomas (LGGs). This was achieved by exploring the use of multivariate pattern recognition methods to generate statistical models of the metabolic characteristics of recurrent LGGs that correlate with aggressive biology and poor clinical outcome.

METHODS

Statistical models were constructed to distinguish between patients with recurrent gliomas that had undergone malignant transformation to a higher grade and those that remained grade 2. The pattern recognition methods explored in this paper include three filter-based feature selection methods (chi-square, gain ratio, and two-way conditional probability), a genetic search wrapper-based feature subset selection algorithm, and five classification algorithms (linear discriminant analysis, logistic regression, functional trees, support vector machines, and decision stump logit boost). The accuracy of each pattern recognition framework was evaluated using leave-one-out cross-validation and bootstrapping.

MATERIALS

The population studied included fifty-three patients with recurrent grade 2 gliomas. Among these patients, seven had tumors that transformed to grade 4, twenty-four had tumors that transformed to grade 3, and twenty-two had tumors that remained grade 2. Image-guided tissue samples were obtained from these patients using surgical navigation software. Part of each tissue sample was examined by a pathologist for histological features and for consistency with the tumor grade diagnosis. The other part of the tissue sample was analyzed with ex vivo nuclear magnetic resonance (NMR) spectroscopy.

RESULTS

Distinguishing between recurrent low grade gliomas that transformed to a higher grade and those that remained grade 2 was achieved with 96% accuracy, using areas of the ex vivo NMR spectrum corresponding to myoinositol, 2-hydroxyglutarate, hypo-taurine, choline, glycerophosphocholine, phosphocholine, glutathione, and lipid. Logistic regression and decision stump boosting models were able to distinguish between recurrent gliomas that transformed to a higher grade and those that did not with 100% training accuracy (95% confidence interval [93-100%]), 96% leave-one-out cross-validation accuracy (95% confidence interval [87-100%]), and 96% bootstrapping accuracy (95% confidence interval [95-97%]). Linear discriminant analysis, functional trees, and support vector machines were able to achieve leave-one-out cross-validation accuracy above 90% and bootstrapping accuracy above 85%. The three feature ranking methods were comparable in performance.

CONCLUSIONS

This study demonstrates the feasibility of using quantitative pattern recognition methods for the analysis of metabolic data from brain tissue obtained during the surgical resection of gliomas. All pattern recognition techniques provided good diagnostic accuracies, though logistic regression and decision stump boosting slightly outperform the other classifiers. These methods identified biomarkers that can be used to detect malignant transformations in individual low grade gliomas, and can lead to a timely change in treatment for each patient.

Classification of single-voxel 1H spectra of brain tumours using LCModel

This study presents a novel method for the direct classification of (1)H single-voxel MR brain tumour spectra using the widespread analysis tool LCModel. LCModel is designed to estimate individual metabolite proportions by fitting a linear combination of in vitro metabolite spectra to an in vivo MR spectrum. In this study, it is used to fit representations of complete tumour spectra and to perform a classification according to the highest estimated tissue proportion. Each tumour type is represented by two spectra, a mean component and a variability term, as calculated using a principal component analysis of a training dataset. In the same manner, a mean component and a variability term for normal white matter are also added into the analysis to allow a mixed tissue approach. An unbiased evaluation of the method is carried out through the automatic selection of training and test sets using the Kennard and Stone algorithm, and a comparison of LCModel classification results with those of the INTERPRET Decision Support System (IDSS) which incorporates an advanced pattern recognition method. In a test set of 46 spectra comprising glioblastoma multiforme, low-grade gliomas and meningiomas, LCModel gives a classification accuracy of 90% compared with an accuracy of 95% by IDSS.

Magnetic resonance spectroscopy detectable metabolomic fingerprint of response to antineoplastic treatment

Targeted therapeutic approaches are increasingly being implemented in the clinic, but early detection of response frequently presents a challenge as many new therapies lead to inhibition of tumor growth rather than tumor shrinkage. Development of novel non-invasive methods to monitor response to treatment is therefore needed. Magnetic resonance spectroscopy (MRS) and magnetic resonance spectroscopic imaging are non-invasive imaging methods that can be employed to monitor metabolism, and previous studies indicate that these methods can be useful for monitoring the metabolic consequences of treatment that are associated with early drug target modulation. However, single-metabolite biomarkers are often not specific to a particular therapy. Here we used an unbiased 1H MRS-based metabolomics approach to investigate the overall metabolic consequences of treatment with the phosphoinositide 3-kinase inhibitor LY294002 and the heat shock protein 90 inhibitor 17AAG in prostate and breast cancer cell lines. LY294002 treatment resulted in decreased intracellular lactate, alanine fumarate, phosphocholine and glutathione. Following 17AAG treatment, decreased intracellular lactate, alanine, fumarate and glutamine were also observed but phosphocholine accumulated in every case. Furthermore, citrate, which is typically observed in normal prostate tissue but not in tumors, increased following 17AAG treatment in prostate cells. This approach is likely to provide further information about the complex interactions between signaling and metabolic pathways. It also highlights the potential of MRS-based metabolomics to identify metabolic signatures that can specifically inform on molecular drug action.

Incremental Gaussian Discriminant Analysis based on Graybill and Deal weighted combination of estimators for brain tumour diagnosis

In the last decade, machine learning (ML) techniques have been used for developing classifiers for automatic brain tumour diagnosis. However, the development of these ML models rely on a unique training set and learning stops once this set has been processed. Training these classifiers requires a representative amount of data, but the gathering, preprocess, and validation of samples is expensive and time-consuming. Therefore, for a classical, non-incremental approach to ML, it is necessary to wait long enough to collect all the required data. In contrast, an incremental learning approach may allow us to build an initial classifier with a smaller number of samples and update it incrementally when new data are collected. In this study, an incremental learning algorithm for Gaussian Discriminant Analysis (iGDA) based on the Graybill and Deal weighted combination of estimators is introduced. Each time a new set of data becomes available, a new estimation is carried out and a combination with a previous estimation is performed. iGDA does not require access to the previously used data and is able to include new classes that were not in the original analysis, thus allowing the customization of the models to the distribution of data at a particular clinical center. An evaluation using five benchmark databases has been used to evaluate the behaviour of the iGDA algorithm in terms of stability-plasticity, class inclusion and order effect. Finally, the iGDA algorithm has been applied to automatic brain tumour classification with magnetic resonance spectroscopy, and compared with two state-of-the-art incremental algorithms. The empirical results obtained show the ability of the algorithm to learn in an incremental fashion, improving the performance of the models when new information is available, and converging in the course of time. Furthermore, the algorithm shows a negligible instance and concept order effect, avoiding the bias that such effects could introduce.

A generic and extensible automatic classification framework applied to brain tumour diagnosis in HealthAgents

New biomedical technologies enable the diagnosis of brain tumours by using non-invasive methods. HealthAgents is a European Union-funded research project that aims to build an agent-based distributed decision support system (dDSS) for the diagnosis of brain tumours. This is achieved using the latest biomedical knowledge, information and communication technologies and pattern recognition (PR) techniques. As part of the PR development of HealthAgents, an independent and automatic classification framework (CF) has been developed. This framework has been integrated with the HealthAgents dDSS using the HealthAgents agent platform. The system offers (1) the functionality to search for distributed classifiers to solve specific questions; (2) automatic classification of new cases; (3) instant deployment of new validated classifiers; and (4) the ability to rank a set of classifiers according to their performance and suitability for the case in hand. The CF enables both the deployment of new classifiers using the provided Extensible Markup Language1 classifier specification, and the inclusion of new PR techniques that make the system extensible. These features may enable the rapid integration of PR laboratory results into industrial or research applications, such as the HealthAgents dDSS. Two classification nodes have been deployed and they currently offer classification services by means of dedicated servers connected to the HealthAgents agent platform: one node being located at the Katholieke Universiteit Leuven, Belgium and the other at the Universidad Politécnica de Valencia, Spain. These classification nodes share the current set of brain tumour classifiers that have been trained from in vivo magnetic resonance spectroscopy data. The combination of the CF with a distributed agent system constitutes the basis of the brain tumour dDSS developed in HealthAgents.

A knowledge-rich distributed decision support framework: a case study for brain tumour diagnosis

The HealthAgents project aims to provide a decision support system for brain tumour diagnosis using a collaborative network of distributed agents. The goal is that through the aggregation of the small data sets available at individual hospitals, much better decision support classifiers can be created and made available to the hospitals taking part. In this paper, we describe the technicalities of the HealthAgents framework, in particular how the interoperability of the various agents is managed using semantic web technologies. On the broad scale the architecture is based around distributed data-mart agents that provide ontological access to hospitals’ underlying data that has been anonymized and processed from proprietary formats into a canonical format. Classifier producers have agents that gather the global data from participating hospitals such that classifiers can be created and deployed as agents. The design on a microscale has each agent built upon a generic-layered framework that provides the common agent program code, allowing rapid development of agents for the system. We believe that our framework provides a well-engineered, agent-based approach to data sharing in a medical context. It can provide a better basis on which to investigate the effectiveness of new classification techniques for brain tumour diagnosis.

The development of a graphical user interface, functional elements and classifiers for the non-invasive characterization of childhood brain tumours using magnetic resonance spectroscopy

Magnetic resonance spectroscopy (MRS) is a non-invasive method, which can provide diagnostic information on children with brain tumours. The technique has not been widely used in clinical practice, partly because of the difficulty of developing robust classifiers from small patient numbers and the challenge of providing decision support systems (DSSs) acceptable to clinicians. This paper describes a participatory design approach in the development of an interactive clinical user interface, as part of a distributed DSS for the diagnosis and prognosis of brain tumours. In particular, we consider the clinical need and context of developing interactive elements for an interface that facilitates the classification of childhood brain tumours, for diagnostic purposes, as part of the HealthAgents European Union project. Previous MRS-based DSS tools have required little input from the clinician user and a raw spectrum is essentially processed to provide a diagnosis sometimes with an estimate of error. In childhood brain tumour diagnosis where there are small numbers of cases and a large number of potential diagnoses, this approach becomes intractable. The involvement of clinicians directly in the designing of the DSS for brain tumour diagnosis from MRS led to an alternative approach with the creation of a flexible DSS that, allows the clinician to input prior information to create the most relevant differential diagnosis for the DSS. This approach mirrors that which is currently taken by clinicians and removes many sources of potential error. The validity of this strategy was confirmed for a small cohort of children with cerebellar tumours by combining two diagnostic types, pilocytic astrocytomas (11 cases) and ependymomas (four cases) into a class of glial tumours which then had similar numbers to the other diagnostic type, medulloblastomas (18 cases). Principal component analysis followed by linear discriminant analysis on magnetic resonance spectral data gave a classification accuracy of 91% for a three-class classifier and 94% for a two-class classifier using a leave-one-out analysis. This DSS provides a flexible method for the clinician to use MRS for brain tumour diagnosis in children.

A Web-accessible distributed data warehouse for brain tumour diagnosis

Currently, biological databases (DBs) are a common tool to complement the research of a wide range of biomedical disciplines, but there are only a few specialized medical DBs for human brain tumour magnetic resonance spectroscopy (MRS) data; they typically store a limited range of biological data (i.e. clinical information, magnetic resonance imaging and MRS data) and are not offered as open-source Structured Query Language relational DB schemas. We present a novel approach to biological DBs: a distributed Web-accessible DB for storing and managing clinical and biomedical data related to brain tumours from different clinical centres. This tool is designed for multi-platform systems with dissimilar DB management systems. Being the main data repository of the HealthAgents (HA) project, it uses multi-agent technology and allows the centres to share data and obtain diagnosis classifications from other centres distributed around the world in a reliable way.

The HA project aims to create an agent-based distributed decision support system (DSS) to assist doctors to provide a brain tumour diagnosis and prognosis. The HA DB enables the DSS to totally integrate with its Graphical User Interface to perform classifications with the stored data and visualize the results using the HA distributed agents framework. This new feature converts the system presented in the first application in the world to combine a storage and management tool for brain tumour data and a complete Web-based DSS to obtain automatic diagnosis.

Semi-supervised analysis of human brain tumours from partially labeled MRS information, using manifold learning models

Medical diagnosis can often be understood as a classification problem. In oncology, this typically involves differentiating between tumour types and grades, or some type of discrete outcome prediction. From the viewpoint of computer-based medical decision support, this classification requires the availability of accurate diagnoses of past cases as training target examples. The availability of such labeled databases is scarce in most areas of oncology, and especially so in neuro-oncology. In such context, semi-supervised learning oriented towards classification can be a sensible data modeling choice. In this study, semi-supervised variants of Generative Topographic Mapping, a model of the manifold learning family, are applied to two neuro-oncology problems: the diagnostic discrimination between different brain tumour pathologies, and the prediction of outcomes for a specific type of aggressive brain tumours. Their performance compared favorably with those of the alternative Laplacian Eigenmaps and Semi-Supervised SVM for Manifold Learning models in most of the experiments.

Serial magnetic resonance spectroscopy reveals a direct metabolic effect of cediranib in glioblastoma

Proton magnetic resonance spectroscopy is increasingly used in clinical studies of brain tumor to provide information about tissue metabolic profiles. In this study, we evaluated changes in the levels of metabolites predominant in recurrent glioblastoma multiforme (rGBM) to characterize the response of rGBM to antiangiogenic therapy. We examined 31 rGBM patients treated with daily doses of cediranib, acquiring serial chemical shift imaging data at specific time points during the treatment regimen. We defined spectra from three regions of interest (ROI)--enhancing tumor (ET), peritumoral tissue, and normal tissue on the contralateral side (cNT)--in post-contrast T1-weighted images, and normalized the concentrations of N-acetylaspartate (NAA) and choline (Cho) in each ROI to the concentration of creatine in cNT (norCre). We analyzed the ratios of these normalized metabolites (i.e., NAA/Cho, NAA/norCre, and Cho/norCre) by averaging all patients and categorizing two different survival groups. Relative to pretreatment values, NAA/Cho in ET was unchanged through day 28. However, after day 28, NAA/Cho significantly increased in relation to a significant increase in NAA/norCre and a decrease in Cho/norCre; interestingly, the observed trend was reversed after day 56, consistent with the clinical course of GBM recurrence. Notably, receiver operating characteristic analysis indicated that NAA/Cho in tumor shows a high prediction to 6-month overall survival. These metabolic changes in these rGBM patients strongly suggest a direct metabolic effect of cediranib and might also reflect an antitumor response to antiangiogenic treatment during the first 2 months of treatment. Further study is needed to confirm these findings.

The INTERPRET Decision-Support System version 3.0 for evaluation of Magnetic Resonance Spectroscopy data from human brain tumours and other abnormal brain masses

BACKGROUND

Proton Magnetic Resonance (MR) Spectroscopy (MRS) is a widely available technique for those clinical centres equipped with MR scanners. Unlike the rest of MR-based techniques, MRS yields not images but spectra of metabolites in the tissues. In pathological situations, the MRS profile changes and this has been particularly described for brain tumours. However, radiologists are frequently not familiar to the interpretation of MRS data and for this reason, the usefulness of decision-support systems (DSS) in MRS data analysis has been explored.

RESULTS

This work presents the INTERPRET DSS version 3.0, analysing the improvements made from its first release in 2002. Version 3.0 is aimed to be a program that 1st, can be easily used with any new case from any MR scanner manufacturer and 2nd, improves the initial analysis capabilities of the first version. The main improvements are an embedded database, user accounts, more diagnostic discrimination capabilities and the possibility to analyse data acquired under additional data acquisition conditions. Other improvements include a customisable graphical user interface (GUI). Most diagnostic problems included have been addressed through a pattern-recognition based approach, in which classifiers based on linear discriminant analysis (LDA) were trained and tested.

CONCLUSIONS

The INTERPRET DSS 3.0 allows radiologists, medical physicists, biochemists or, generally speaking, any person with a minimum knowledge of what an MR spectrum is, to enter their own SV raw data, acquired at 1.5 T, and to analyse them. The system is expected to help in the categorisation of MR Spectra from abnormal brain masses.

Combined use of neuroradiology and 1H-MR spectroscopy may provide an intervention limiting diagnosis of glioblastoma multiforme

PURPOSE

To evaluate the accuracy of (1)H-MR spectroscopy ((1)H-MRS) as an intervention limiting diagnostic tool for glioblastoma multiforme. GBM is the most common and aggressive primary brain tumor, with mean survival under a year. Oncological practice currently requires histopathological diagnosis before radiotherapy.

MATERIALS AND METHODS

Eighty-nine patients had clinical computed tomography (CT) and MR imaging and 1.5T SV SE (1)H-MRS with PRESS localization for neuroradiological diagnosis and tumor classification with spectroscopic and automated pattern recognition analysis (TE 30 ms, TR 2000 ms, spectral width 2500 Hz and 2048 data points, 128-256 signal averages were acquired, depending on voxel size (8 cm(3) to 4 cm(3)). Eighteen patients from a cohort of 89 underwent stereotactic biopsy.

RESULTS

The 18 stereotactic biopsies revealed 14 GBM, 2 grade II astrocytomas, 1 lymphoma, and 1 anaplastic astrocytoma. All 14 biopsied GBMs were diagnosed as GBM by a protocol combining an individual radiologist and an automated spectral pattern recognition program.

CONCLUSION

In patients undergoing stereotactic biopsy combined neuroradiological and spectroscopic evaluation diagnoses GBM with accuracy that could replace the need for biopsy. We do not advocate the replacement of biopsy in all patients; instead our data suggest a specific intervention limiting role for the use of (1)H-MRS in brain tumor diagnosis.

Proton MR spectroscopy provides relevant prognostic information in high-grade astrocytomas

BACKGROUND AND PURPOSE

There is a large range of survival times in patients with HGA that can only be partially explained by histologic grade and clinical aspects. This study aims to retrospectively assess the predictive value of single-voxel (1)H-MRS regarding survival in HGA.

MATERIALS AND METHODS

Pretreatment (1)H-MRS in 187 patients with HGA produced 180 spectra at STE (30 ms) and 182 at LTE (136 ms). Patients were dichotomized into 2 groups according to survival better or worse than the median. The spectra of the 2 groups were compared using the Mann-Whitney U test. The points on the spectrum with the most significant differences were selected for discriminating patients with good and poor prognosis. Thresholds were defined with ROC curves, and survival was analyzed by using the Kaplan-Meier method and the Cox proportional hazards model.

RESULTS

Four points on the spectrum showed the most significant differences: 0.98 and 3.67 ppm at STE; and 0.98 and 1.25 ppm at LTE (P between <.001 and .011). These points were useful for stratifying 2 prognostic groups (P between <.001 and .003, Kaplan-Meier). The Cox forward stepwise model selected 3 spectroscopic variables: the intensity values of the points 3.67 ppm at STE (hazard ratio, 2.132; 95% CI, 1.504-3.023), 0.98 ppm at LTE (hazard ratio, 0.499; 95% CI, 0.339-0.736), and 1.25 ppm at LTE (hazard ratio, 0.574; 95% CI, 0.368-0.897).

CONCLUSIONS

(1)H-MRS is of value in predicting the length of survival in patients with HGA and could be used to stratify prognostic groups.