Noninvasive diagnostic assessment of brain tumors using combined in vivo MR imaging and spectroscopy

Advances in Neuro-Oncological Imaging: An Update on Diagnostic Approach to Brain Tumors

This study delineates the pivotal role of imaging within the field of neurology, emphasizing its significance in the diagnosis, prognostication, and evaluation of treatment responses for central nervous system (CNS) tumors. A comprehensive understanding of both the capabilities and limitations inherent in emerging imaging technologies is imperative for delivering a heightened level of personalized care to individuals with neuro-oncological conditions. Ongoing research in neuro-oncological imaging endeavors to rectify some limitations of radiological modalities, aiming to augment accuracy and efficacy in the management of brain tumors. This review is dedicated to the comparison and critical examination of the latest advancements in diverse imaging modalities employed in neuro-oncology. The objective is to investigate their respective impacts on diagnosis, cancer staging, prognosis, and post-treatment monitoring. By providing a comprehensive analysis of these modalities, this review aims to contribute to the collective knowledge in the field, fostering an informed approach to neuro-oncological care. In conclusion, the outlook for neuro-oncological imaging appears promising, and sustained exploration in this domain is anticipated to yield further breakthroughs, ultimately enhancing outcomes for individuals grappling with CNS tumors.

Predicting Survival in Patients with Brain Tumors: Current State-of-the-Art of AI Methods Applied to MRI

Given growing clinical needs, in recent years Artificial Intelligence (AI) techniques have increasingly been used to define the best approaches for survival assessment and prediction in patients with brain tumors. Advances in computational resources, and the collection of (mainly) public databases, have promoted this rapid development. This narrative review of the current state-of-the-art aimed to survey current applications of AI in predicting survival in patients with brain tumors, with a focus on Magnetic Resonance Imaging (MRI). An extensive search was performed on PubMed and Google Scholar using a Boolean research query based on MeSH terms and restricting the search to the period between 2012 and 2022. Fifty studies were selected, mainly based on Machine Learning (ML), Deep Learning (DL), radiomics-based methods, and methods that exploit traditional imaging techniques for survival assessment. In addition, we focused on two distinct tasks related to survival assessment: the first on the classification of subjects into survival classes (short and long-term or eventually short, mid and long-term) to stratify patients in distinct groups. The second focused on quantification, in days or months, of the individual survival interval. Our survey showed excellent state-of-the-art methods for the first, with accuracy up to ∼98%. The latter task appears to be the most challenging, but state-of-the-art techniques showed promising results, albeit with limitations, with C-Index up to ∼0.91. In conclusion, according to the specific task, the available computational methods perform differently, and the choice of the best one to use is non-univocal and dependent on many aspects. Unequivocally, the use of features derived from quantitative imaging has been shown to be advantageous for AI applications, including survival prediction. This evidence from the literature motivates further research in the field of AI-powered methods for survival prediction in patients with brain tumors, in particular, using the wealth of information provided by quantitative MRI techniques.

Multi-level Kronecker Convolutional Neural Network (ML-KCNN) for Glioma Segmentation from Multi-modal MRI Volumetric Data

The development of an automated glioma segmentation system from MRI volumes is a difficult task because of data imbalance problem. The ability of deep learning models to incorporate different layers for data representation assists medical experts like radiologists to recognize the condition of the patient and further make medical practices easier and automatic. State-of-the-art deep learning algorithms enable advancement in the medical image segmentation area, such a segmenting the volumes into sub-tumor classes. For this task, fully convolutional network (FCN)-based architectures are used to build end-to-end segmentation solutions. In this paper, we proposed a multi-level Kronecker convolutional neural network (MLKCNN) that captures information at different levels to have both local and global level contextual information. Our ML-KCNN uses Kronecker convolution, which overcomes the missing pixels problem by dilated convolution. Moreover, we used a post-processing technique to minimize false positive from segmented outputs, and the generalized dice loss (GDL) function handles the data-imbalance problem. Furthermore, the combination of connected component analysis (CCA) with conditional random fields (CRF) used as a post-processing technique achieves reduced Hausdorff distance (HD) score of 3.76 on enhancing tumor (ET), 4.88 on whole tumor (WT), and 5.85 on tumor core (TC). Dice similarity coefficient (DSC) of 0.74 on ET, 0.90 on WT, and 0.83 on TC. Qualitative and visual evaluation of our proposed method shown effectiveness of the proposed segmentation method can achieve performance that can compete with other brain tumor segmentation techniques.

Assessment of Overall Survival in Glioma Patients as Predicted by Metabolomic Criteria

Objective: We assess the efficacy of the metabolomic profile from glioma biopsies in providing estimates of postsurgical Overall Survival in glioma patients.

Methods: Tumor biopsies from 46 patients bearing gliomas, obtained neurosurgically in the period 1992–1998, were analyzed by high resolution ¹H magnetic resonance spectroscopy (HR- ¹H MRS), following retrospectively individual postsurgical Overall Survival up to 720 weeks.

Results: The Overall Survival profile could be resolved in three groups; Short (shorter than 52 weeks, n = 19), Intermediate (between 53 and 364 weeks, n = 19) or Long (longer than 365 weeks, n = 8), respectively. Classical histopathological analysis assigned WHO grades II–IV to every biopsy but notably, some patients with low grade glioma depicted unexpectedly Short Overall Survival, while some patients with high grade glioma, presented unpredictably Long Overall Survival. To explore the reasons underlying these different responses, we analyzed HR-¹H MRS spectra from acid extracts of the same biopsies, to characterize the metabolite patterns associated to OS predictions. Poor prognosis was found in biopsies with higher contents of alanine, acetate, glutamate, total choline, phosphorylcholine, and glycine, while more favorable prognosis was achieved in biopsies with larger contents of total creatine, glycerol-phosphorylcholine, and myo-inositol. We then implemented a multivariate analysis to identify hierarchically the influence of metabolomic biomarkers on OS predictions, using a Classification Regression Tree (CRT) approach. The CRT based in metabolomic biomarkers grew up to three branches and split into eight nodes, predicting correctly the outcome of 94.7% of the patients in the Short Overall Survival group, 78.9% of the patients in the Intermediate Overall Survival group, and 75% of the patients in the Long Overall Survival group, respectively.

Conclusion: Present results indicate that metabolic profiling by HR-¹H MRS improves the Overall Survival predictions derived exclusively from classical histopathological gradings, thus favoring more precise therapeutic decisions.

Diagnostic accuracy and added value of qualitative radiological review of 1H-magnetic resonance spectroscopy in evaluation of childhood brain tumors

Background

¹H-magnetic resonance spectroscopy (MRS) facilitates noninvasive diagnosis of pediatric brain tumors by providing metabolite profiles. Prospective studies of diagnostic accuracy and comparisons with conventional MRI are lacking. We aimed to evaluate diagnostic accuracy of MRS for childhood brain tumors and determine added clinical value compared with conventional MRI.

Methods

Children presenting to a tertiary pediatric center with brain lesions from December 2015 through 2017 were included. MRI and single-voxel MRS were acquired on 52 tumors and sequentially interpreted by 3 radiologists, blinded to histopathology. Proportions of correct diagnoses and interrater agreement at each stage were compared. Cases were reviewed to determine added value of qualitative radiological review of MRS through increased certainty of correct diagnosis, reduced number of differentials, or diagnosis following spectroscopist evaluation. Final diagnosis was agreed by the tumor board at study end.

Results

Radiologists' principal MRI diagnosis was correct in 69%, increasing to 77% with MRS. MRI + MRS resulted in significantly more additional correct diagnoses than MRI alone (P = .035). There was a significant increase in interrater agreement when correct with MRS (P = .046). Added value following radiologist interpretation of MRS occurred in 73% of cases, increasing to 83% with additional spectroscopist review. First histopathological diagnosis was available a median of 9.5 days following imaging, with 25% of all patients managed without conclusive histopathology.

Conclusions

MRS can improve the accuracy of noninvasive diagnosis of pediatric brain tumors and add value in the diagnostic pathway. Incorporation into practice has the potential to facilitate early diagnosis, guide treatment planning, and improve patient care.

The value of magnetic resonance spectroscopy as a supplement to MRI of the brain in a clinical setting

BACKGROUND

There are different opinions of the clinical value of MRS of the brain. In selected materials MRS has demonstrated good results for characterisation of both neoplastic and non-neoplastic lesions. The aim of this study was to evaluate the supplemental value of MR spectroscopy (MRS) in a clinical setting.

MATERIAL AND METHODS

MRI and MRS were re-evaluated in 208 cases with a clinically indicated MRS (cases with uncertain or insufficient information on MRI) and a confirmed diagnosis. Both single voxel spectroscopy (SVS) and chemical shift imaging (CSI) were performed in 105 cases, only SVS or CSI in 54 and 49 cases, respectively. Diagnoses were grouped into categories: non-neoplastic disease, low-grade tumour, and high-grade tumour. The clinical value of MRS was considered very beneficial if it provided the correct category or location when MRI did not, beneficial if it ruled out suspected diseases or was more specific than MRI, inconsequential if it provided the same level of information, or misleading if it provided less or incorrect information.

RESULTS

There were 70 non-neoplastic lesions, 43 low-grade tumours, and 95 high-grade tumours. For MRI, the category was correct in 130 cases (62%), indeterminate in 39 cases (19%), and incorrect in 39 cases (19%). Supplemented with MRS, 134 cases (64%) were correct, 23 cases (11%) indeterminate, and 51 (25%) incorrect. Additional information from MRS was beneficial or very beneficial in 31 cases (15%) and misleading in 36 cases (17%).

CONCLUSION

In most cases MRS did not add to the diagnostic value of MRI. In selected cases, MRS may be a valuable supplement to MRI.

Evaluation of the added value of 1H-magnetic resonance spectroscopy for the diagnosis of pediatric brain lesions in clinical practice

Background

Magnetic resonance spectroscopy (MRS) aids noninvasive diagnosis of pediatric brain tumors, but use in clinical practice is not well documented. We aimed to review clinical use of MRS, establish added value in noninvasive diagnosis, and investigate potential impact on patient care.

Methods

Sixty-nine children with lesions imaged using MRS and reviewed by the tumor board from 2014 to 2016 met inclusion criteria. Contemporaneous MRI diagnosis, spectroscopy analysis, histopathology, and clinical information were reviewed. Final diagnosis was agreed on by the tumor board at study end.

Results

Five cases were excluded for lack of documented MRI diagnosis. The principal MRI diagnosis by pediatric radiologists was correct in 59%, increasing to 73% with addition of MRS. Of the 73%, 19.1% (95% CI, 9.1%-33.3%) were incorrectly diagnosed with MRI alone. MRS led to a significant improvement in correct diagnosis over all tumor types (P = .012). Of diagnoses correctly made with MRI, confidence increased by 37% when adding MRS, with no patients incorrectly re-diagnosed. Indolent lesions were diagnosed noninvasively in 85% of cases, with MRS a major contributor to 91% of these diagnoses. Of all patients, 39% were managed without histopathological diagnosis. MRS contributed to diagnosis in 68% of this group, modifying it in 12%. MRS influenced management in 33% of cases, mainly through avoiding and guiding biopsy and aiding tumor characterization.

Conclusion

MRS can improve accuracy and confidence in noninvasive diagnosis of pediatric brain lesions in clinical practice. There is potential to improve outcomes through avoiding biopsy of indolent lesions, aiding tumor characterization, and facilitating earlier family discussions and treatment planning.

Clinical Imaging of Tumor Metabolism with ¹H Magnetic Resonance Spectroscopy

Magnetic resonance spectroscopy (MRS) is a noninvasive functional technique to evaluate the biochemical behavior of human tissues. This property has been widely used in assessment and therapy monitoring of brain tumors. MRS studies can be implemented outside the brain, with successful and promising results in the evaluation of prostate and breast cancer, although still with limited reproducibility. As a result of technical improvements, malignancies of the musculoskeletal system and abdominopelvic organs can benefit from the molecular information that MRS provides. The technical challenges and main applications in oncology of (1)H MRS in a clinical setting are the focus of this review.

Combined MRI and MRS improves pre-therapeutic diagnoses of pediatric brain tumors over MRI alone

INTRODUCTION

The specific goal of this study was to determine whether the inclusion of MRS had a measureable and positive impact on the accuracy of pre-surgical MR examinations of untreated pediatric brain tumors over that of MRI alone in clinical practice.

METHODS

Final imaging reports of 120 pediatric patients with newly detected brain tumors who underwent combined MRI/MRS examinations were retrospectively reviewed. Final pathology was available in all cases. Group A comprised 60 subjects studied between June 2001 and January 2005, when MRS was considered exploratory and radiologists utilized only conventional MRI to arrive at a diagnosis. For group B, comprising 60 subjects studied between January 2005 and March 2008, the radiologists utilized information from both MRI and MRS. Furthermore, radiologists revisited group A (blind review, time lapse >4 years) to determine whether the additional information from MRS would have altered their interpretation.

RESULTS

Sixty-three percent of patients in group A were diagnosed correctly, whereas in 10% the report was partially correct with the final tumor type mentioned (but not mentioned as most likely tumor), while in 27% of cases the reports were wrong. For group B, the diagnoses were correct in 87%, partially correct in 5%, and incorrect in 8% of the cases, which is a significant improvement (p < 0.005). Re-review of combined MRI and MRS of group A resulted 87% correct, 7% partially correct, and 7% incorrect diagnoses, which is a significant improvement over the original diagnoses (p < 0.05).

CONCLUSION

Adding MRS to conventional MRI significantly improved diagnostic accuracy in preoperative pediatric patients with untreated brain tumors.

Added value of advanced over conventional magnetic resonance imaging in grading gliomas and other primary brain tumors

BACKGROUND

Although conventional MR imaging (MRI) is the most widely used non-invasive technique for brain tumor grading, its accuracy has been reported to be relatively low. Advanced MR techniques, such as perfusion-weighted imaging (PWI), diffusion-weighted imaging (DWI), and magnetic resonance spectroscopy (MRS), could predict neoplastic histology, but their added value over conventional MRI is still open to debate.

METHODS

We prospectively analyzed 129 patients diagnosed with primary brain tumors (118 gliomas) classified as low-grade in 30 cases and high-grade in 99 cases.

RESULTS

Significant differences were obtained in high-grade tumors for conventional MRI variables (necrosis, enhancement, edema, hemorrhage, and neovascularization); high relative cerebral blood volume values (rCBV), low relative apparent diffusion coefficients (rADC), high ratio of N-acetyl-aspartate/creatine at short echo time (TE) and high choline/creatine at long TE. Among conventional MRI variables, the presence of enhancement and necrosis were demonstrated to be the best predictors of high grade in primary brain tumors (sensitivity 95.9%; specificity 70%). The best results in primary brain tumors were obtained for enhancement, necrosis, and rADC (sensitivity 98.9%; specificity 75.9%). Necrosis and enhancement were the only predictors of high grade in gliomas (sensitivity 97.6%; specificity 76%) when all the magnetic resonance variables were combined.

CONCLUSIONS

MRI is highly accurate in the assessment of tumor grade. The combination of conventional MRI features with advanced MR variables showed only improved tumor grading by adding rADC to conventional MRI variables in primary brain tumors.

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.

Awake surgery between art and science. Part I: clinical and operative settings

Awake surgery requires coordinated teamwork and communication between the surgeon and the anesthesiologist, as he monitors the patient, the neuroradiologist as he interprets the images for intraoperative confirmation, and the neuropsychologist and neurophysiologist as they evaluate in real-time the patient's responses to commands and questions. To improve comparison across published studies on clinical assessment and operative settings in awake surgery, we reviewed the literature, focusing on methodological differences and aims. In complex, interdisciplinary medical care, such differences can affect the outcome and the cost-benefit ratio of the treatment. Standardization of intraoperative mapping and related controversies will be discussed in Part II.

Metabolomic approach to human brain spectroscopy identifies associations between clinical features and the frontal lobe metabolome in multiple sclerosis

Proton magnetic resonance spectroscopy ((1)H-MRS) is capable of noninvasively detecting metabolic changes that occur in the brain tissue in vivo. Its clinical utility has been limited so far, however, by analytic methods that focus on independently evaluated metabolites and require prior knowledge about which metabolites to examine. Here, we applied advanced computational methodologies from the field of metabolomics, specifically partial least squares discriminant analysis and orthogonal partial least squares, to in vivo (1)H-MRS from frontal lobe white matter of 27 patients with relapsing-remitting multiple sclerosis (RRMS) and 14 healthy controls. We chose RRMS, a chronic demyelinating disorder of the central nervous system, because its complex pathology and variable disease course make the need for reliable biomarkers of disease progression more pressing. We show that in vivo MRS data, when analyzed by multivariate statistical methods, can provide reliable, distinct profiles of MRS-detectable metabolites in different patient populations. Specifically, we find that brain tissue in RRMS patients deviates significantly in its metabolic profile from that of healthy controls, even though it appears normal by standard MRI techniques. We also identify, using statistical means, the metabolic signatures of certain clinical features common in RRMS, such as disability score, cognitive impairments, and response to stress. This approach to human in vivo MRS data should promote understanding of the specific metabolic changes accompanying disease pathogenesis, and could provide biomarkers of disease progression that would be useful in clinical trials.

Multimodal MR imaging (diffusion, perfusion, and spectroscopy): is it possible to distinguish oligodendroglial tumor grade and 1p/19q codeletion in the pretherapeutic diagnosis?

BACKGROUND AND PURPOSE

Pretherapeutic determination of tumor grade and genotype in grade II and III oligodendroglial tumors is clinically important but is still challenging. Tumor grade and 1p/19q status are currently the 2 most important factors in therapeutic decision making for patients with these tumors. Histopathology and cMRI studies are still limited in some cases. In the present study, we were interested in determining whether the combination of PWI, DWI, and MR spectroscopy could help distinguish oligodendroglial tumors according to their histopathologic grade and genotype.

MATERIALS AND METHODS

We retrospectively reviewed 50 adult patients with grade II and III oligodendrogliomas and oligoastrocytomas who had DWI, PWI, and MR spectroscopy at short and long TE data and known 1p/19q status. Univariate analyses and multivariate random forest models were performed to determine which criteria could differentiate between grades and genotypes.

RESULTS

ADC, rCBV, rCBF, and rK2 were significantly different between grade II and III oligodendroglial tumors. DWI, PWI, and MR spectroscopy showed no significant difference between tumors with and without 1p/19q loss. Separation between tumor grades and genotypes with cMRI alone showed 31% and 48% misclassification rates, respectively. Multimodal MR imaging helps to determine tumor grade and 1p/19q genotype more accurately (misclassification rates of 17% and 40%, respectively).

CONCLUSIONS

Although multimodal investigation of oligodendroglial tumors has a lower contribution to 1p/19q genotyping compared with cMRI alone, it greatly improves the accuracy of grading of these neoplasms. Use of multimodal MR imaging could thus provide valuable information that may assist clinicians in patient preoperative management and treatment decision making.

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.

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.

(1)H MR Spectroscopy in Gliomatosis: Is there a Sensitivity Issue?

Objective. ¹H MR spectroscopy (MRS) is widely performed for assessment of brain tumours and is considered a highly sensitive technique capable of differentiating benign from malignant conditions and tumour grading. Method. We present a case of a 69 year old woman who was suspected to have gliomatosis on MRI. Results. MRS performed using single voxel and chemical shift/multivoxel techniques was within normal limits. A repeat scan 6 months later showed progressive disease, and biopsy was performed that proved the diagnosis of glioblastoma. Conclusion. Normal MRS in a patient with suspicion of gliomatosis on MRI should not reassure clinicians into assuming a benign aetiology or a good prognosis in short term.

[Brain metastasis of breast cancer; imaging evaluation]

CNS involvement in breast cancer modifies the prognosis and the treatment of the disease. Imaging plays a leading role for the diagnosis, the pretherapeutic assessment and the follow-up. MRI is the most sensitive modality for the detection of infraclinic lesions, reported in about 15% of metastatic breast cancers. In addition to conventional MR study, diffusion MR, perfusion MR and spectroscopy have a diagnostic value with specificity of more than 95%; 3D study is required if neurosurgical resection or stereotactic radiosurgery is contemplated. The use of new drugs in clinical trials needs a precise and accurate follow up to assess their usefulness; appreciation of the response is based on the precise measure of the number of targets and of their size; The WHO and recently the RECIST have established the guidelines for measurement of the tumoral targets and to assess the response to treatments. Brain modifications related to surgery or stereotactic radiosurgery are well studied by MRI.

Strengths and weaknesses of 1.5T and 3T MRS data in brain glioma classification

Although magnetic resonance spectroscopy (MRS) methods of 1.5Tesla (T) and 3T have been widely applied during the last decade for noninvasive diagnostic purposes, only a few studies have been reported on the value of the information extracted in brain cancer discrimination. The purpose of this study is threefold. First, to show that the diagnostic value of the information extracted from two different MRS scanners of 1.5T and 3T is significantly influenced in terms of brain gliomas discrimination. Second, to statistically evaluate the discriminative potential of publicly known metabolic ratio markers, obtained from these two types of scanners in classifying low-, intermediate-, and high-grade gliomas. Finally, to examine the diagnostic value of new metabolic ratios in the discrimination of complex glioma cases where the diagnosis is both challenging and critical. Our analysis has shown that although the information extracted from 3T MRS scanner is expected to provide better brain gliomas discrimination; some factors like the features selected, the pulse-sequence parameters, and the spectroscopic data acquisition methods can influence the discrimination efficiency. Finally, it is shown that apart from the bibliographical known, new metabolic ratio features such as N-acetyl aspartate/ S, Choline/ S, Creatine/ S , and myo-Inositol/ S play significant role in gliomas grade discrimination.

pH optimization for a reliable quantification of brain tumor cell and tissue extracts with (1)H NMR: focus on choline-containing compounds and taurine

The aim of this study was to define the optimal pH for (1)H nuclear magnetic resonance (NMR) spectroscopy analysis of perchloric acid or methanol-chloroform-water extracts from brain tumor cells and tissues. The systematic study of the proton chemical shift variations as a function of pH of 13 brain metabolites in model solutions demonstrated that recording (1)H NMR spectra at pH 10 allowed resolving resonances that are overlapped at pH 7, especially in the 3.2-3.3 ppm choline-containing-compounds region. (1)H NMR analysis of extracts at pH 7 or 10 showed that quantitative measurements of lactate, alanine, glutamate, glutamine (Gln), creatine + phosphocreatine and myo-inositol (m-Ino) can be readily performed at both pHs. The concentrations of glycerophosphocholine, phosphocholine and choline that are crucial metabolites for tumor brain malignancy grading were accurately measured at pH 10 only. Indeed, the resonances of their trimethylammonium moieties are cleared of any overlapping signal, especially those of taurine (Tau) and phosphoethanolamine. The four non-ionizable Tau protons resonating as a singlet in a non-congested spectral region permits an easier and more accurate quantitation of this apoptosis marker at pH 10 than at pH 7 where the triplet at 3.43 ppm can be overlapped with the signals of glucose or have an intensity too low to be measured. Glycine concentration was determined indirectly at both pHs after subtracting the contribution of the overlapped signals of m-Ino at pH 7 or Gln at pH 10.

Combined diffusion imaging and MR spectroscopy in the diagnosis of human prion diseases

BACKGROUND AND PURPOSE

The physiopathologic bases underlying the signal intensity changes and reduced diffusibility observed in prion diseases (TSEs) are still poorly understood. We evaluated the interest of MRS combined with DWI both as a diagnostic tool and a way to understand the mechanism underlying signal intensity and ADC changes in this setting.

MATERIALS AND METHODS

We designed a prospective study of multimodal MR imaging in patients with suspected TSEs. Forty-five patients with a suspicion of TSE and 11 age-matched healthy volunteers were included. The MR imaging protocol included T1, FLAIR, and DWI sequences. MRS was performed on the cerebellum, pulvinar, right lenticular nucleus, and frontal cortex. MR images were assessed visually, and ADC values were calculated.

RESULTS

Among the 45 suspected cases, 31 fulfilled the criteria for probable or definite TSEs (19 sCJDs, 3 iCJDs, 2 vCJDs, and 7 genetic TSEs); and 14 were classified as AltDs. High signals in the cortex and/or basal ganglia were observed in 26/31 patients with TSEs on FLAIR and 29/31 patients on DWI. In the basal ganglia, high DWI signals corresponded to a decreased ADC. Metabolic alterations, increased mIns, and decreased NAA were observed in all patients with TSEs. ADC values and metabolic changes were not correlated; this finding suggests that neuronal stress (vacuolization), neuronal loss, and astrogliosis do not alone explain the decrease of ADC.

CONCLUSIONS

MRS combined with other MR imaging is of interest in the diagnosis of TSE and provides useful information for understanding physiopathologic processes underlying prion diseases.

Experience of diffusion tensor imaging and 1H spectroscopy for outcome prediction in severe traumatic brain injury: Preliminary results

OBJECTIVE

The objective of the study is to test whether multimodal magnetic resonance imaging can provide a reliable outcome prediction of the clinical status, focusing on consciousness at 1 year after severe traumatic brain injury (TBI).

DESIGN

Single center prospective cohort with consecutive inclusions.

SETTING

Critical Care Neurosurgical Unit of a university hospital.

PATIENTS

Forty-three TBI patients not responding to simple orders after sedation cessation and 15 healthy controls.

INTERVENTIONS

A multimodal magnetic resonance imaging combining morphologic sequences, diffusion tensor imaging (DTI), and H proton magnetic resonance spectroscopy (MRS) was performed 24 +/- 11 days after severe TBI. The ability of DTI and MRS to predict 1-year outcome was assessed by linear discriminant analysis (LDA). Robustness of the classification was tested using a bootstrap procedure.

MEASUREMENTS AND MAIN RESULTS

Fractional anisotropy (FA) was computed as the mean of values at discrete brain sites in the infratentorial and supratentorial regions. The N-acetyl aspartate/creatine (NAA/Cr) ratio was measured in the thalamus, lenticular nucleus, insular cortex, occipital periventricular white matter, and pons. After 1 year, 19 (44%) patients had unfavorable outcomes (death, persistent vegetative state, or minimally conscious state) and 24 (56%) favorable outcomes (normal consciousness with or without functional impairments). Analysis of variance was performed to compare FA and NAA/Cr in the two outcome groups and controls. FA and MRS findings showed highly significant differences between the outcome groups, with significant variables by LDA being supratentorial FA, NAA/Cr (pons), NAA/Cr (thalamus), NAA/Cr (insula), and infratentorial FA. LDA of combined FA and MRS data clearly separated the unfavorable outcome, favorable outcome, and control groups, with no overlap. Unfavorable outcome was predicted with up to 86% sensitivity and 97% specificity; these values were better than those obtained with DTI or MRS alone.

CONCLUSION

FA and NAA/Cr hold potential as quantitative outcome-prediction tools at the subacute phase of TBI.

Nosologic imaging of the brain: segmentation and classification using MRI and MRSI

A new technique is presented to create nosologic images of the brain based on magnetic resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI). A nosologic image summarizes the presence of different tissues and lesions in a single image by color coding each voxel or pixel according to the histopathological class it is assigned to. The proposed technique applies advanced methods from image processing as well as pattern recognition to segment and classify brain tumors. First, a registered brain atlas and a subject-specific abnormal tissue prior, obtained from MRSI data, are used for the segmentation. Next, the detected abnormal tissue is classified based on supervised pattern recognition methods. Class probabilities are also calculated for the segmented abnormal region. Compared to previous approaches, the new framework is more flexible and able to better exploit spatial information leading to improved nosologic images. The combined scheme offers a new way to produce high-resolution nosologic images, representing tumor heterogeneity and class probabilities, which may help clinicians in decision making.

Multiproject-multicenter evaluation of automatic brain tumor classification by magnetic resonance spectroscopy

JUSTIFICATION

Automatic brain tumor classification by MRS has been under development for more than a decade. Nonetheless, to our knowledge, there are no published evaluations of predictive models with unseen cases that are subsequently acquired in different centers. The multicenter eTUMOUR project (2004-2009), which builds upon previous expertise from the INTERPRET project (2000-2002) has allowed such an evaluation to take place.

MATERIALS AND METHODS

A total of 253 pairwise classifiers for glioblastoma, meningioma, metastasis, and low-grade glial diagnosis were inferred based on 211 SV short TE INTERPRET MR spectra obtained at 1.5 T (PRESS or STEAM, 20-32 ms) and automatically pre-processed. Afterwards, the classifiers were tested with 97 spectra, which were subsequently compiled during eTUMOUR.

RESULTS

In our results based on subsequently acquired spectra, accuracies of around 90% were achieved for most of the pairwise discrimination problems. The exception was for the glioblastoma versus metastasis discrimination, which was below 78%. A more clear definition of metastases may be obtained by other approaches, such as MRSI + MRI.

CONCLUSIONS

The prediction of the tumor type of in-vivo MRS is possible using classifiers developed from previously acquired data, in different hospitals with different instrumentation under the same acquisition protocols. This methodology may find application for assisting in the diagnosis of new brain tumor cases and for the quality control of multicenter MRS databases.

The effect of combining two echo times in automatic brain tumor classification by MRS

(1)H MRS is becoming an accurate, non-invasive technique for initial examination of brain masses. We investigated if the combination of single-voxel (1)H MRS at 1.5 T at two different (TEs), short TE (PRESS or STEAM, 20-32 ms) and long TE (PRESS, 135-136 ms), improves the classification of brain tumors over using only one echo TE. A clinically validated dataset of 50 low-grade meningiomas, 105 aggressive tumors (glioblastoma and metastasis), and 30 low-grade glial tumors (astrocytomas grade II, oligodendrogliomas and oligoastrocytomas) was used to fit predictive models based on the combination of features from short-TEs and long-TE spectra. A new approach that combines the two consecutively was used to produce a single data vector from which relevant features of the two TE spectra could be extracted by means of three algorithms: stepwise, reliefF, and principal components analysis. Least squares support vector machines and linear discriminant analysis were applied to fit the pairwise and multiclass classifiers, respectively. Significant differences in performance were found when short-TE, long-TE or both spectra combined were used as input. In our dataset, to discriminate meningiomas, the combination of the two TE acquisitions produced optimal performance. To discriminate aggressive tumors from low-grade glial tumours, the use of short-TE acquisition alone was preferable. The classifier development strategy used here lends itself to automated learning and test performance processes, which may be of use for future web-based multicentric classifier development studies.

Identification and characterisation of childhood cerebellar tumours by in vivo proton MRS

(1)H MRS has great potential for the clinical investigation of childhood brain tumours, but the low incidence in, and difficulties of performing trials on, children have hampered progress in this area. Most studies have used a long-TE, thus limiting the metabolite information obtained, and multivariate analysis has been largely unexplored. Thirty-five children with untreated cerebellar tumours (18 medulloblastomas, 12 pilocytic astrocytomas and five ependymomas) were investigated using a single-voxel short-TE PRESS sequence on a 1.5 T scanner. Spectra were analysed using LCModel to yield metabolite profiles, and key metabolite assignments were verified by comparison with high-resolution magic-angle-spinning NMR of representative tumour biopsy samples. In addition to univariate metabolite comparisons, the use of multivariate classifiers was investigated. Principal component analysis was used for dimension reduction, and linear discriminant analysis was used for variable selection and classification. A bootstrap cross-validation method suitable for estimating the true performance of classifiers in small datasets was used. The discriminant function coefficients were stable and showed that medulloblastomas were characterised by high taurine, phosphocholine and glutamate and low glutamine, astrocytomas were distinguished by low creatine and high N-acetylaspartate, and ependymomas were differentiated by high myo-inositol and glycerophosphocholine. The same metabolite features were seen in NMR spectra of ex vivo samples. Successful classification was achieved for glial-cell (astrocytoma + ependymoma) versus non-glial-cell (medulloblastoma) tumours, with a bootstrap 0.632 + error, e(B.632+), of 5.3%. For astrocytoma vs medulloblastoma and astrocytoma vs medulloblastoma vs ependymoma classification, the e(B.632+) was 6.9% and 7.1%, respectively. The study showed that (1)H MRS detects key differences in the metabolite profiles for the main types of childhood cerebellar tumours and that discriminant analysis of metabolite profiles is a promising tool for classification. The findings warrant confirmation by larger multi-centre studies.

Differentiation of High-Grade and Low-Grade Gliomas Using Pattern Analysis of Long-Echo Single-Voxel Proton Magnetic Resonance Spectroscopy ((1)H-MRS)

The usefulness of proton magnetic resonance spectroscopy ((1)H-MRS) for glioma grading is not clear, particularly due to the absence of standard criteria for data analysis. Previously we had developed an original classification of the pathological (1)H-MRS spectra based on the identification of the predominant metabolite peak, N-acetylaspartate (NAA) for Type I, choline-containing compounds (Cho) for Type II, and mobile lipids (Lip) for Type III, and presence or absence of other metabolite peaks: lactate (Lac), Lip, or Cho. The present study evaluated the effectiveness of this classification in grading of previously non-treated gliomas. A total of 38 low-grade and 33 high-grade neoplasms were investigated. Four tumors had (1)H-MRS spectra Type I, and all of those were low-grade. Three tumors had (1)H-MRS spectra Type III, and all those were glioblastomas. Fifteen tumors with (1)H-MRS spectra Type II had a Lip/NAA ratio more than 1 (Type II C with moderate elevation of lipids), and 12 of those neoplasms were high-grade. The differences in distribution of high-grade and low-grade gliomas among another 49 gliomas with (1)H-MRS spectra Type II did not depend on the presence of Lac and/or Lip peaks, and in this subgroup NAA/Cho ratio was also evaluated. Inclusion of both characteristics (type of the (1)H-MRS spectrum and NAA/Cho ratio with defined cut-off level of 0.6) into the diagnostic algorithm yielded 72% diagnostic accuracy (95% confidence interval: 62%-82%) in discriminating high-grade and low-grade neoplasms. In conclusion, pattern analysis of the pathological (1)H-MRS spectra using the proposed classification along with evaluation of NAA/Cho ratio might be helpful for non-invasive glioma grading.

Multivariate analysis of diffusion tensor imaging data improves the detection of microstructural damage in young professional boxers

In this study, we present two different methods of multivariate analysis of voxel-based diffusion tensor imaging (DTI) data, using as an example data derived from 59 professional boxers and 12 age-matched controls. Conventional univariate analysis ignores much of the diffusion information contained in the tensor. Our first multivariate method uses the Hotelling's T(2) statistic and the second uses linear discriminant analysis to generate the linear discriminant function at each voxel to form a separability metric. Both multivariate methods confirm the findings from the individual metrics of large-scale changes in the bilateral inferior temporal gyri of the boxers, but they also reveal greater sensitivity as well as identifying major subcortical changes that had not been evident in the univariate analyses. Linear discriminant analysis has the added strength of providing a quantitative measure of the relative contribution of each metric to any differences between the two subject groups. This novel adaptation of statistical and mathematical techniques to neuroimaging analysis is important for two reasons. Clinically, it develops the findings of a previous mild head injury study, and, methodologically, it could equally well be applied to multivariate studies of other pathologies.

(1)H MR spectroscopy of human brain tumours: a practical approach

Magnetic resonance spectroscopy (MRS) is proposed in addition to magnetic resonance imaging (MRI) to help in the characterization of brain tumours by detecting metabolic alterations that may be indicative of the tumour class. MRS can be routinely performed on clinical magnets, within a reasonable acquisition time and if performed under adequate conditions, MRS is reproducible and thus can be used for longitudinal follow-up of treatment. MRS can also be performed in clinical practice to guide the neurosurgeon into the most aggressive part of the lesions or to avoid unnecessary surgery, which may furthermore decrease the risk of surgical morbidity.

Proton MR spectroscopy of cerebral gliomas at 3 T: spatial heterogeneity, and tumour grade and extent

This study aimed to evaluate the usefulness of proton MR spectroscopic imaging ((1)H-MRSI) at 3 T in differentiating high- from low-grade gliomas, and tumour from necrosis, oedema or normal tissue. Forty-four patients with brain gliomas and four with meningiomas were retrospectively reviewed. The normalised metabolites choline (nCho), N-acetylaspartate (nNAA), creatine (nCr) and lactate/lipids (nLL), and the metabolite ratios Cho/NAA, NAA/Cr and Cho/Cr were calculated. Necrotic-appearing areas showed two spectroscopic patterns: "necrosis" with variable nCho and high nLL, and "cystic necrosis" with variable nLL or nonevident peaks. Peri-enhancing oedematous-appearing areas showed three spectroscopic patterns ("tumour" with abnormal Cho/NAA, "oedema" with normal Cho/NAA and "tumour/oedema" with normal nCho and abnormal Cho/NAA) in gliomas, and one ("oedema") in meningiomas. Peri-enhancing or peri-tumour normal-appearing areas showed two patterns ("infiltrated" with abnormal nCho and/or Cho/NAA and "normal" with normal spectra) in gliomas and one ("normal") in meningiomas. Discriminant analysis showed that classification accuracy between high- and low-grade glioma masses was better with normalised metabolites or all parameters together than metabolite ratios and that among peri-enhancing areas was much better with normalised metabolites. The analysis of spatial distribution of normalised metabolites by 3-T (1)H-MRSI helps to discriminate among different tissues, offering information not available with conventional MRI.

Multiparametric differentiation of posterior fossa tumors in children using diffusion-weighted imaging and short echo-time 1H-MR spectroscopy

PURPOSE

To assess the combined value of diffusion-weighted imaging (DWI) and proton magnetic resonance spectroscopy (1H-MRS) in differentiating medulloblastoma, ependymoma, pilocytic astrocytoma, and infiltrating glioma in a pediatric population.

MATERIALS AND METHODS

A total of 17 children with untreated posterior fossa tumors (seven medulloblastoma, four infiltrating glioma, two ependymoma, and four pilocytic astrocytoma), were investigated with conventional MRI, DWI, and MRS using a single-voxel technique. Within the nonnecrotic tumor core, apparent diffusion coefficient (ADC) values using a standardized region of interest (ROI) were retrieved. Quantification of water signal and analysis of metabolite signals from MRS measurements in the same tumorous area were reviewed using multivariant linear discriminant analysis.

RESULTS

Combination of ADC values and metabolites, which were normalized using water as an internal standard, allowed discrimination between the four tumor groups with a likelihood below 1 x 10(-9). Positive predictive value was 1 in all cases. Tumors could not be discriminated when using metabolite ratios or ADC values alone, nor could they be differentiated using creatine (Cr) as an internal reference even in combination with ADC values.

CONCLUSION

Linear discriminant analysis using DWI and MRS using water as internal reference, fully discriminates the four most frequent posterior fossa tumors in children.

Perturbation of mouse glioma MRS pattern by induced acute hyperglycemia

(1)H MRS is evolving into an invaluable tool for brain tumor classification in humans based on pattern recognition analysis, but there is still room for improvement. Here we propose a new approach: to challenge tumor metabolism in vivo by a defined perturbation, and study the induced changes in MRS pattern. For this we recorded single voxel (1)H MR spectra from mice bearing a stereotactically induced GL261 grade IV brain glioma during a period of induced acute hyperglycemia. A total of 29 C57BL/6 mice were used. Single voxel spectra were acquired at 7 T with point resolved spectroscopy and TE of 12, 30 and 136 ms. Tumors were induced by stereotactic injection of 10(5) GL261cells in 17 mice. Hyperglycemia (up to 338 +/- 36 mg/dL glucose in the blood) was induced by intraperitoneal bolus injection. Maximal increases in glucose resonances of up to 2.4-fold were recorded from tumors in vivo. Our observations are in agreement with extracellular accumulation of glucose, which may suggest that glucose transport and/or metabolism are working close to their maximum capacity in GL261 tumors. The significant and specific MRS pattern changes observed when comparing euglycemia and hyperglycemia may be of use for future pattern-recognition studies of animal and human brain tumors by enhancing MRS-based discrimination between tumor types and grades.

Correlations between MR and endothelial hyperplasia in low-grade gliomas

PURPOSE

To investigate the capacity of multimodal MR to detect endothelial hyperplasia (EH), which has been linked to the aggressiveness of gliomas and which is so far only detected by biopsy, an invasive technique that prevents repeated measurements and early detection.

MATERIALS AND METHODS

A total of 26 patients with low-grade gliomas participated in the study. All underwent a histopathological analysis and a multimodal MR examination (spectroscopic, anatomic, diffusion, perfusion, and postcontrast imaging).

RESULTS

EH was present (EH+) in 15 patients and absent (EH-) in 11. No differences were observed between EH- and EH+ groups when comparing spectroscopic and diffusion parameters. Perfusion measurements, however, allowed us to distinguish EH+ from EH-: the relative regional cerebral blood flow (rCBF) was found equal to 3.23 +/- 2.05 for EH+ and 1.33 +/- 0.46 for EH- (P = 0.006).

CONCLUSION

We have observed a strong correlation between the presence of EH and the increase of rCBF. Compared to conventional imaging, MR perfusion provides additional and complementary information that may be used for biopsy guidance, early detection of tumor aggressiveness, and noninvasive follow-ups.

Generating prior probabilities for classifiers of brain tumours using belief networks

Background

Numerous methods for classifying brain tumours based on magnetic resonance spectra and imaging have been presented in the last 15 years. Generally, these methods use supervised machine learning to develop a classifier from a database of cases for which the diagnosis is already known. However, little has been published on developing classifiers based on mixed modalities, e.g. combining imaging information with spectroscopy. In this work a method of generating probabilities of tumour class from anatomical location is presented.

Methods

The method of "belief networks" is introduced as a means of generating probabilities that a tumour is any given type. The belief networks are constructed using a database of paediatric tumour cases consisting of data collected over five decades; the problems associated with using this data are discussed. To verify the usefulness of the networks, an application of the method is presented in which prior probabilities were generated and combined with a classification of tumours based solely on MRS data.

Results

Belief networks were constructed from a database of over 1300 cases. These can be used to generate a probability that a tumour is any given type. Networks are presented for astrocytoma grades I and II, astrocytoma grades III and IV, ependymoma, pineoblastoma, primitive neuroectodermal tumour (PNET), germinoma, medulloblastoma, craniopharyngioma and a group representing rare tumours, "other". Using the network to generate prior probabilities for classification improves the accuracy when compared with generating prior probabilities based on class prevalence.

Conclusion

Bayesian belief networks are a simple way of using discrete clinical information to generate probabilities usable in classification. The belief network method can be robust to incomplete datasets. Inclusion of a priori knowledge is an effective way of improving classification of brain tumours by non-invasive methods.

[Brain magnetic resonance spectroscopy]

MR spectroscopy (MRS) sequences allow noninvasive exploration of brain metabolism during a MRI examination. Their day-to-day use in a clinical setting has recently been improved by simple programming of sequences and automated quantification of metabolites. However, a few simple rules should be observed in the choice of sequences and the location of the voxels so as to obtain an informative, high-quality examination. The research applications of MR spectroscopy, where use of this examination seeks to better understand the pathophysiology of the disease, must be distinguished from its clinical indications, where MRS provides information that can be used directly in patient management. The most significant of the clinical uses are imaging intracranial tumors (positive and differential diagnosis, extension, treatment follow-up), diffuse brain injury, encephalopathies (especially hepatic and HIV-related), and the diagnosis of metabolic disorders.

[Indications for cerebral MR proton spectroscopy in 2007]

Magnetic resonance spectroscopy (MRS) is being increasingly performed alongside the more conventional MRI sequences in the exploration of neurological disorders. It is however important to clearly differentiate its clinical applications aiming at improving the differential diagnosis or the prognostic evaluation of the patient, from the research protocols, when MRS can contribute to a better understanding of the pathophysiology of the disease or to the evaluation of new treatments. The most important applications in clinical practice are intracranial space occupying lesions (especially the positive diagnosis of intracranial abscesses and gliomatosis cerebri and the differential diagnosis between edema and tumor infiltration), alcoholic, hepatic, and HIV-related encephalopathies and the exploration of metabolic diseases. Among the research applications, MRS is widely used in multiple sclerosis, ischemia and brain injury, epilepsy and neuro degenerative diseases.

Spectroscopie par résonance magnétique des tumeurs cérébrales

MR spectroscopy (MRS) can complement MRI in the evaluation of intracranial tumors. Before treatment, MRS can contribute to the differential diagnosis between tumor and non tumoral lesion (especially intracranial abscesses), to assess the aggressiveness of a glial tumor or to determine its extension to better delineate the surgical removal or the target volume of radiotherapy. During treatment follow-up, MRS helps differentiate recurrent tumor from radionecrosis or physiological post-surgical contrast enhancement. The current studies are trying to determine if the indications of MRS, alone or in association with other MR sequences can further be extended in the study of brain tumors, in particular the follow-up of lesions undergoing chemo or radiotherapy.