Fast nosologic imaging of the brain

Using Single-Voxel Magnetic Resonance Spectroscopy Data Acquired at 1.5T to Classify Multivoxel Data at 3T: A Proof-of-Concept Study

In vivo magnetic resonance spectroscopy (MRS) has two modalities, single-voxel (SV) and multivoxel (MV), in which one or more contiguous grids of SVs are acquired.

PURPOSE

To test whether MV grids can be classified with models trained with SV.

METHODS

Retrospective study. Training dataset: Multicenter multiformat SV INTERPRET, 1.5T. Testing dataset: MV eTumour, 3T. Two classification tasks were completed: 3-class (meningioma vs. aggressive vs. normal) and 4-class (meningioma vs. low-grade glioma vs. aggressive vs. normal). Five different methods were tested for feature selection. The classification was implemented using linear discriminant analysis (LDA), random forest, and support vector machines. The evaluation was completed with balanced error rate (BER) and area under the curve (AUC) on both sets. The accuracy in class prediction was calculated by developing a solid tumor index (STI) and segmentation accuracy with the Dice score.

RESULTS

The best method was sequential forward feature selection combined with LDA, with AUCs = 0.95 (meningioma), 0.89 (aggressive), 0.82 (low-grade glioma), and 0.82 (normal). STI was 66% (4-class task) and 71% (3-class task) because two cases failed completely and two more had suboptimal STI as defined by us.

DISCUSSION

The reasons for failure in the classification of the MV test set were related to the presence of artifacts.

Convolutional neural networks to predict brain tumor grades and Alzheimer’s disease with MR spectroscopic imaging data

Purpose

To evaluate the value of convolutional neural network (CNN) in the diagnosis of human brain tumor or Alzheimer’s disease by MR spectroscopic imaging (MRSI) and to compare its Matthews correlation coefficient (MCC) score against that of other machine learning methods and previous evaluation of the same data. We address two challenges: 1) limited number of cases in MRSI datasets and 2) interpretability of results in the form of relevant spectral regions.

Methods

A shallow CNN with only one hidden layer and an ad-hoc loss function was constructed involving two branches for processing spectral and image features of a brain voxel respectively. Each branch consists of a single convolutional hidden layer. The output of the two convolutional layers is merged and fed to a classification layer that outputs class predictions for the given brain voxel.

Results

Our CNN method separated glioma grades 3 and 4 and identified Alzheimer’s disease patients using MRSI and complementary MRI data with high MCC score (Area Under the Curve were 0.87 and 0.91 respectively). The results demonstrated superior effectiveness over other popular methods as Partial Least Squares or Support Vector Machines. Also, our method automatically identified the spectral regions most important in the diagnosis process and we show that these are in good agreement with existing biomarkers from the literature.

Conclusion

Shallow CNNs models integrating image and spectral features improved quantitative and exploration and diagnosis of brain diseases for research and clinical purposes. Software is available at https://bitbucket.org/TeslaH2O/cnn_mrsi.

Embedding MRI information into MRSI data source extraction improves brain tumour delineation in animal models

Glioblastoma is the most frequent malignant intra-cranial tumour. Magnetic resonance imaging is the modality of choice in diagnosis, aggressiveness assessment, and follow-up. However, there are examples where it lacks diagnostic accuracy. Magnetic resonance spectroscopy enables the identification of molecules present in the tissue, providing a precise metabolomic signature. Previous research shows that combining imaging and spectroscopy information results in more accurate outcomes and superior diagnostic value. This study proposes a method to combine them, which builds upon a previous methodology whose main objective is to guide the extraction of sources. To this aim, prior knowledge about class-specific information is integrated into the methodology by setting the metric of a latent variable space where Non-negative Matrix Factorisation is performed. The former methodology, which only used spectroscopy and involved combining spectra from different subjects, was adapted to use selected areas of interest that arise from segmenting the T2-weighted image. Results showed that embedding imaging information into the source extraction (the proposed semi-supervised analysis) improved the quality of the tumour delineation, as compared to those obtained without this information (unsupervised analysis). Both approaches were applied to pre-clinical data, involving thirteen brain tumour-bearing mice, and tested against histopathological data. On results of twenty-eight images, the proposed Semi-Supervised Source Extraction (SSSE) method greatly outperformed the unsupervised one, as well as an alternative semi-supervised approach from the literature, with differences being statistically significant. SSSE has proven successful in the delineation of the tumour, while bringing benefits such as 1) not constricting the metabolomic-based prediction to the image-segmented area, 2) ability to deal with signal-to-noise issues, 3) opportunity to answer specific questions by allowing researchers/radiologists define areas of interest that guide the source extraction, 4) creation of an intra-subject model and avoiding contamination from inter-subject overlaps, and 5) extraction of meaningful, good-quality sources that adds interpretability, conferring validation and better understanding of each case.

Classification of brain tumours from MR spectra: the INTERPRET collaboration and its outcomes

The INTERPRET project was a multicentre European collaboration, carried out from 2000 to 2002, which developed a decision-support system (DSS) for helping neuroradiologists with no experience of MRS to utilize spectroscopic data for the diagnosis and grading of human brain tumours. INTERPRET gathered a large collection of MR spectra of brain tumours and pseudo-tumoural lesions from seven centres. Consensus acquisition protocols, a standard processing pipeline and strict methods for quality control of the aquired data were put in place. Particular emphasis was placed on ensuring the diagnostic certainty of each case, for which all cases were evaluated by a clinical data validation committee. One outcome of the project is a database of 304 fully validated spectra from brain tumours, pseudotumoural lesions and normal brains, along with their associated images and clinical data, which remains available to the scientific and medical community. The second is the INTERPRET DSS, which has continued to be developed and clinically evaluated since the project ended. We also review here the results of the post-INTERPRET period. We evaluate the results of the studies with the INTERPRET database by other consortia or research groups. A summary of the clinical evaluations that have been performed on the post-INTERPRET DSS versions is also presented. Several have shown that diagnostic certainty can be improved for certain tumour types when the INTERPRET DSS is used in conjunction with conventional radiological image interpretation. About 30 papers concerned with the INTERPRET single-voxel dataset have so far been published. We discuss stengths and weaknesses of the DSS and the lessons learned. Finally we speculate on how the INTERPRET concept might be carried into the future.

Molecular imaging coupled to pattern recognition distinguishes response to temozolomide in preclinical glioblastoma

Non-invasive monitoring of response to treatment of glioblastoma (GB) is nowadays carried out using MRI. MRS and MR spectroscopic imaging (MRSI) constitute promising tools for this undertaking. A temozolomide (TMZ) protocol was optimized for GL261 GB. Sixty-three mice were studied by MRI/MRS/MRSI. The spectroscopic information was used for the classification of control brain and untreated and responding GB, and validated against post-mortem immunostainings in selected animals. A classification system was developed, based on the MRSI-sampled metabolome of normal brain parenchyma, untreated and responding GB, with a 93% accuracy. Classification of an independent test set yielded a balanced error rate of 6% or less. Classifications correlated well both with tumor volume changes detected by MRI after two TMZ cycles and with the histopathological data: a significant decrease (p < 0.05) in the proliferation and mitotic rates and a 4.6-fold increase in the apoptotic rate. A surrogate response biomarker based on the linear combination of 12 spectral features has been found in the MRS/MRSI pattern of treated tumors, allowing the non-invasive classification of growing and responding GL261 GB. The methodology described can be applied to preclinical treatment efficacy studies to test new antitumoral drugs, and begets translational potential for early response detection in clinical studies.

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.

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.

Data analysis and tissue type assignment for glioblastoma multiforme

Glioblastoma multiforme (GBM) is characterized by high infiltration. The interpretation of MRSI data, especially for GBMs, is still challenging. Unsupervised methods based on NMF by Li et al. (2013, NMR in Biomedicine) and Li et al. (2013, IEEE Transactions on Biomedical Engineering) have been proposed for glioma recognition, but the tissue types is still not well interpreted. As an extension of the previous work, a tissue type assignment method is proposed for GBMs based on the analysis of MRSI data and tissue distribution information. The tissue type assignment method uses the values from the distribution maps of all three tissue types to interpret all the information in one new map and color encodes each voxel to indicate the tissue type. Experiments carried out on in vivo MRSI data show the feasibility of the proposed method. This method provides an efficient way for GBM tissue type assignment and helps to display information of MRSI data in a way that is easy to interpret.

Blind source separation and automatic tissue typing of microdiffraction data by hierarchical nonnegative matrix factorization

In this article a nonnegative blind source separation technique, known as nonnegative matrix factorization, is applied to microdiffraction data in order to extract characteristic patterns and to determine their spatial distribution in tissue typing problems occurring in bone-tissue engineering. In contrast to other blind source separation methods, nonnegative matrix factorization only requires nonnegative constraints on the extracted sources and corresponding weights, which makes it suitable for the analysis of data occurring in a variety of applications. In particular, here nonnegative matrix factorization is hierarchically applied to two-dimensional meshes of X-ray diffraction data measured in bone samples with implanted tissue. Such data are characterized by nonnegative profiles and their analysis provides significant information about the structure of possibly new deposited bone tissue. A simulation and real data studies show that the proposed method is able to retrieve the patterns of interest and to provide a reliable and accurate segmentation of the given X-ray diffraction data.

Unsupervised nosologic imaging for glioma diagnosis

In this letter a novel approach to create nosologic images of the brain using magnetic resonance spectroscopic imaging (MRSI) data in an unsupervised way is presented. Different tissue patterns are identified from the MRSI data using nonnegative matrix factorization and are then coded as different primary colors (i.e. red, green, and blue) in an RGB image, so that mixed tissue regions are automatically visualized as mixtures of primary colors. The approach is useful in assisting glioma diagnosis, where several tissue patterns such as normal, tumor, and necrotic tissue can be present in the same voxel/spectrum. Error-maps based on linear least squares estimation are computed for each nosologic image to provide additional reliability information, which may help clinicians in decision making. Tests on in vivo MRSI data show the potential of this new approach.

Convex non-negative matrix factorization for brain tumor delimitation from MRSI data

Background

Pattern Recognition techniques can provide invaluable insights in the field of neuro-oncology. This is because the clinical analysis of brain tumors requires the use of non-invasive methods that generate complex data in electronic format. Magnetic Resonance (MR), in the modalities of spectroscopy (MRS) and spectroscopic imaging (MRSI), has been widely applied to this purpose. The heterogeneity of the tissue in the brain volumes analyzed by MR remains a challenge in terms of pathological area delimitation.

Methodology/Principal Findings

A pre-clinical study was carried out using seven brain tumor-bearing mice. Imaging and spectroscopy information was acquired from the brain tissue. A methodology is proposed to extract tissue type-specific sources from these signals by applying Convex Non-negative Matrix Factorization (Convex-NMF). Its suitability for the delimitation of pathological brain area from MRSI is experimentally confirmed by comparing the images obtained with its application to selected target regions, and to the gold standard of registered histopathology data. The former showed good accuracy for the solid tumor region (proliferation index (PI)>30%). The latter yielded (i) high sensitivity and specificity in most cases, (ii) acquisition conditions for safe thresholds in tumor and non-tumor regions (PI>30% for solid tumoral region; ≤5% for non-tumor), and (iii) fairly good results when borderline pixels were considered.

Conclusions/Significance

The unsupervised nature of Convex-NMF, which does not use prior information regarding the tumor area for its delimitation, places this approach one step ahead of classical label-requiring supervised methods for discrimination between tissue types, minimizing the negative effect of using mislabeled voxels. Convex-NMF also relaxes the non-negativity constraints on the observed data, which allows for a natural representation of the MRSI signal. This should help radiologists to accurately tackle one of the main sources of uncertainty in the clinical management of brain tumors, which is the difficulty of appropriately delimiting the pathological area.

Improving the classification of brain tumors in mice with perturbation enhanced (PE)-MRSI

Classifiers based on statistical pattern recognition analysis of MRSI data are becoming important tools for the non-invasive diagnosis of human brain tumors. Here we investigate the potential interest of perturbation-enhanced MRSI (PE-MRSI), in this case acute hyperglycemia, for improving the discrimination between mouse brain MRS patterns of glioblastoma multiforme (GBM), oligodendroglioma (ODG), and non-tumor brain parenchyma (NT). Six GBM-bearing mice and three ODG-bearing mice were scanned at 7 Tesla by PRESS-MRSI with 12 and 136 ms echo-time, during euglycemia (Eug) and also during induced acute hyperglycemia (Hyp), generating altogether four datasets per animal (echo time + glycemic condition): 12Eug, 136Eug, 12Hyp, and 136Hyp. For classifier development all spectral vectors (spv) selected from the MRSI matrix were unit length normalized (UL2) and used either as a training set (76 GBM spv, four mice; 70 ODG spv, two mice; 54 NT spv) or as an independent testing set (61 GBM spv, two mice; 31 ODG, one mouse; 23 NT spv). All Fisher's LDA classifiers obtained were evaluated as far as their descriptive performance-correctly classified cases of the training set (bootstrapping)-and predictive accuracy-balanced error rate of independent testing set classification. MRSI-based classifiers at 12Hyp were consistently more efficient in separating GBM, ODG, and NT regions, with overall accuracies always >80% and up to 95-96%; remaining classifiers were within the 48-85% range. This was also confirmed by user-independent selection of training and testing sets, using leave-one-out (LOO). This highlights the potential interest of perturbation-enhanced MRSI protocols for improving the non-invasive characterization of preclinical brain tumors.

Exploiting spatial information to estimate metabolite levels in two-dimensional MRSI of heterogeneous brain lesions

MRSI provides MR spectra from multiple adjacent voxels within a body volume represented as a two- or three-dimensional matrix, allowing the measurement of the distribution of metabolites over this volume. The spectra of these voxels are usually analyzed one by one, without exploiting their spatial context. In this article, we present an advanced metabolite quantification method for MRSI data, in which the available spatial information is considered. A nonlinear least-squares algorithm is proposed in which prior knowledge is included in the form of proximity constraints on the spectral parameters within a grid and optimized starting values. A penalty term that promotes a spatially smooth spectral parameter map is added to the fitting algorithm. This method is adaptive, in the sense that several sweeps through the grid are performed and each solution may tune some hyperparameters at run-time. Simulation studies of MRSI data showed significantly improved metabolite estimates after the inclusion of spatial information. Improved metabolite maps were also demonstrated by applying the method to in vivo MRSI data. Overlapping peaks or peaks of compounds present at low concentration can be better quantified with the proposed method than with single-voxel approaches. The new approach compares favorably against the multivoxel approach embedded in the well-known quantification software LCModel.

Pattern recognition of MRSI data shows regions of glioma growth that agree with DTI markers of brain tumor infiltration

Gliomas are the most common primary brain tumors and the majority are highly malignant, with one of the worst prognoses for patients. Gliomas are characterized by invasive growth into normal brain tissue that makes complete surgical resection and accurate radiotherapy planning extremely difficult. We have performed independent component analysis of magnetic resonance spectroscopy imaging data from human gliomas to segment brain tissue into tumor core, tumor infiltration, and normal brain, with confirmation by diffusion tensor imaging analysis. Our data are consistent with previous studies that compared anomalies in isotropic and anisotropic diffusion images to determine regions of potential glioma infiltration. We show that coefficients of independent components can be used to create colored images for easy visual identification of regions of infiltrative tumor growth.