Carles Arús

Brain tumor classification based on long echo proton MRS signals

There has been a growing research interest in brain tumor classification based on proton magnetic resonance spectroscopy (1H MRS) signals. Four research centers within the EU funded INTERPRET project have acquired a significant number of long echo 1H MRS signals for brain tumor classification. In this paper, we present an objective comparison of several classification techniques applied to the discrimination of four types of brain tumors: meningiomas, glioblastomas, astrocytomas grade II and metastases. Linear and non-linear classifiers are compared: linear discriminant analysis (LDA), support vector machines (SVM) and least squares SVM (LS-SVM) with a linear kernel as linear techniques and LS-SVM with a radial basis function (RBF) kernel as a non-linear technique. Kernel-based methods can perform well in processing high dimensional data. This motivates the inclusion of SVM and LS-SVM in this study. The analysis includes optimal input variable selection, (hyper-) parameter estimation, followed by performance evaluation. The classification performance is evaluated over 200 stratified random samplings of the dataset into training and test sets. Receiver operating characteristic (ROC) curve analysis measures the performance of binary classification, while for multiclass classification, we consider the accuracy as performance measure. Based on the complete magnitude spectra, automated binary classifiers are able to reach an area under the ROC curve (AUC) of more than 0.9 except for the hard case glioblastomas versus metastases. Although, based on the available long echo 1H MRS data, we did not find any statistically significant difference between the performances of LDA and the kernel-based methods, the latter have the strength that no dimensionality reduction is required to obtain such a high performance.

Towards a method for automated classification of 1H MRS spectra from brain tumours

Recent studies have shown that MRS can substantially improve the non‐invasive categorization of human brain tumours. However, in order for MRS to be used routinely by clinicians, it will be necessary to develop reliable automated classification methods that can be fully validated. This paper is in two parts: the first part reviews the progress that has been made towards this goal, together with the problems that are involved in the design of automated methods to process and classify the spectra. The second part describes the development of a simple prototype system for classifying 1H single voxel spectra, obtained at an echo time (TE) of 135 ms, of the four most common types of brain tumour (meningioma (MM), astrocytic (AST), oligodendroglioma (OD) and metastasis (ME)) and cysts. This system was developed in two stages: firstly, an initial database of spectra was used to develop a prototype classifier, based on a linear discriminant analysis (LDA) of selected data points. Secondly, this classifier was tested on an independent test set of 15 newly acquired spectra, and the system was refined on the basis of these results. The system correctly classified all the non‐astrocytic tumours. However, the results for the the astrocytic group were poorer (between 55 and 100%, depending on the binary comparison). Approximately 50% of high grade astrocytoma (glioblastoma) spectra in our data base showed very little lipid signal, which may account for thepoorer results for this class. Consequently, for the refined system, the astrocytomas were subdivided into two subgroups for comparison against other tumour classes: those with high lipid content and those without.

In Vivo, Ex Vivo, and In Vitro One‐ and Two‐Dimensional Nuclear Magnetic Resonance Spectroscopy of an Intracerebral Glioma in Rat Brain: Assignment of Resonances

Abstract: An in vivo study of intracerebral rat glioma using proton‐localized NMR spectroscopy showed important modifications of the spectra in the tumor as compared with the contralateral brain. To carry out the assignment of the resonances of the glioma spectra, tumoral and normal rat brain tissues were studied in vivo, ex vivo, and in vitro by one‐dimensional and two‐dimensional proton spectroscopy. N‐Acetylaspartate was found at an extremely low level in the glioma. The change of peak ratio total creatine/3.2 ppm peak was found to be due to a simultaneous decrease of the total creatine content and an increase of the 3.2 ppm peak. The 3.2 ppm resonance in the glioma spectra has been shown to originate from choline, phosphocholine, glycerophosphocholine, taurine, inositol, and phosphoethanolamine. The increase of the 3.2 ppm peak in the glioma was found to result from the increase of taurine and phosphoethanolamine contents. The peak in the 1.3 ppm region of the glioma spectra was due to both lactate and mobile fatty acids. Moreover, two‐dimensional spectroscopy of excised tissues and extracts showed the presence of hypotaurine only in the tumor.

HealthAgents: distributed multi-agent brain tumor diagnosis and prognosis

Abstract We present an agent-based distributed decision support system for the diagnosis and prognosis of brain tumors developed by the HealthAgents project. HealthAgents is a European Union funded research project, which aims to enhance the classification of brain tumors using such a decision support system based on intelligent agents to securely connect a network of clinical centers. The HealthAgents system is implementing novel pattern recognition discrimination methods, in order to analyze in vivo Magnetic Resonance Spectroscopy (MRS) and ex vivo/in vitro High Resolution Magic Angle Spinning Nuclear Magnetic Resonance (HR-MAS) and DNA micro-array data. HealthAgents intends not only to apply forefront agent technology to the biomedical field, but also develop the HealthAgents network, a globally distributed information and knowledge repository for brain tumor diagnosis and prognosis.

Proton MR Spectroscopy Improves Discrimination between Tumor and Pseudotumoral Lesion in Solid Brain Masses

BACKGROUND AND PURPOSE: Differentiating between tumors and pseudotumoral lesions by conventional MR imaging may be a challenging question. This study aims to evaluate the potential usefulness and the added value that single-voxel proton MR spectroscopy could provide on this discrimination. MATERIALS AND METHODS: A total of 84 solid brain lesions were retrospectively included in the study (68 glial tumors and 16 pseudotumoral lesions). Single-voxel spectra at TE 30 ms (short TE) and 136 ms (long TE) were available in all cases. Two groups were defined: “training-set” (56 cases) and “test-set” (28 cases). Tumors and pseudotumors were compared in the training-set with the Mann-Whitney U test. Ratios between resonances were defined as classifiers for new cases, and thresholds were selected with receiver operating characteristic (ROC) curves. The added value of spectroscopy was evaluated by 5 neuroradiologists and assessed with the Wilcoxon signed-rank test. RESULTS: Differences between tumors and pseudotumors were found in myo-inositol (mIns); P < .01) at short TE, and N-acetylaspartate (NAA; P < .001), glutamine (Glx; P < .01), and choline (CHO; P < .05) at long TE. Classifiers suggested tumor when mIns/NAA ratio was more than 0.9 at short TE and also when CHO/NAA ratio was more than 1.9 at long TE. Classifier accuracy was tested in the test-set with the following results: short TE, 82% (23/28); long TE, 79% (22/28). The neuroradiologists’ confidence rating of the test-cases on a 5-point scale (0–4) improved between 5% (from 2.86–3) and 27% (from 2.25–2.86) with spectroscopy (mean, 17%; P < .01). CONCLUSIONS: The proposed ratios of mIns/NAA at short TE and CHO/NAA at long TE provide valuable information to discriminate between brain tumor and pseudotumor by improving neuroradiologists’ accuracy and confidence.

On the Design of a Web-Based Decision Support System for Brain Tumour Diagnosis Using Distributed Agents

This paper introduces HealthAgents, an EC-funded research project to improve the classification of brain tumours through multi-agent decision support over a distributed network of local databases or data marts. HealthAgents will not only develop new pattern recognition methods for a distributed classification and analysis of in vivo MRS and ex vivo/in vitro HRMAS and DNA data, but also define a method to assess the quality and usability of a new candidate local database containing a set of new cases, based on a compatibility score

Genetic programming for classification and feature selection: analysis of 1H nuclear magnetic resonance spectra from human brain tumour biopsies

Genetic programming (GP) is used to classify tumours based on 1H nuclear magnetic resonance (NMR) spectra of biopsy extracts. Analysis of such data would ideally give not only a classification result but also indicate which parts of the spectra are driving the classification (i.e. feature selection). Experiments on a database of variables derived from 1H NMR spectra from human brain tumour extracts (n = 75) are reported, showing GPs classification abilities and comparing them with that of a neural network. GP successfully classified the data into meningioma and non‐meningioma classes. The advantage over the neural network method was that it made use of simple combinations of a small group of metabolites, in particular glutamine, glutamate and alanine. This may help in the choice of the most informative NMR spectroscopy methods for future non‐invasive studies in patients.

Quantitative and qualitative characterization of 1H NMR spectra of colon tumors, normal mucosa and their perchloric acid extracts: decreased levels of myo-inositol in tumours can be detected in intact biopsies.

Sixteen colonic tumours and 10 normal mucosa biopsies have been examined by 1H NMR spectroscopy at 9.4 T. A complete characterization and quantification of the aliphatic region of PCA extract spectra and the analysis of the two‐dimensional COSY spectra of five pairs of intact biopsies (tumor and control mucosa) has been carried out. The analysis of the PCA extracts demonstrated a significant increase in the concentration of the endogenous compounds: lactate, glutamate, aspartate, taurine, spermine, glutathione and glycerophosphoethanolamine, and a significant decrease of myo‐ and scyllo‐inositol, in tumours with respect to mucosae. Among these metabolites, the high myo‐inositol and taurine levels and the reciprocal changes found between them in tumours and mucosae make their resonances interesting as possible malignancy markers if they are detectable in vivo. In contrast to the easy observation of taurine in one‐dimensional spectra of intact biopsies, the difficulty of observing myo‐inositol prompted us to use two‐dimensional COSY spectra for the detection and quantification of both these metabolites. In the two‐dimensional spectra, the use of a ratio between the cross‐peak volumes of both metabolites permits an excellent differentiation between tumours and normal mucosa and suggests its potential to detect malignant changes in the healthy tissue, provided a two‐dimensional approach is used.

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

1H MRS is becoming an accurate, non‐invasive technique for initial examination of brain masses. We investigated if the combination of single‐voxel 1H 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. Copyright

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

BackgroundProton 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.ResultsThis 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.ConclusionsThe 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.