Exploring the Potential of Machine Learning Algorithms to Improve Diffusion Nuclear Magnetic Resonance Imaging Models Analysis.

Purpose: This paper explores different machine learning (ML) algorithms for analyzing diffusion nuclear magnetic resonance imaging (dMRI) models when analytical fitting shows restrictions. It reviews various ML techniques for dMRI analysis and evaluates their performance on different b-values range datasets, comparing them with analytical methods. Materials and Methods: After standard fitting for reference, four sets of diffusion-weighted nuclear magnetic resonance images were used to train/test various ML algorithms for prediction of diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), and kurtosis (K). ML classification algorithms, including extra-tree classifier (ETC), logistic regression, C-support vector, extra-gradient boost, and multilayer perceptron (MLP), were used to determine the existence of diffusion parameters (D, D*, f, and K) within single voxels. Regression algorithms, including linear regression, polynomial regression, ridge, lasso, random forest (RF), elastic-net, and support-vector machines, were used to estimate the value of the diffusion parameters. Performance was evaluated using accuracy (ACC), area under the curve (AUC) tests, and cross-validation root mean square error (RMSECV). Computational timing was also assessed. Results: ETC and MLP were the best classifiers, with 94.1% and 91.7%, respectively, for the ACC test and 98.7% and 96.3% for the AUC test. For parameter estimation, RF algorithm yielded the most accurate results The RMSECV percentages were: 8.39% for D, 3.57% for D*, 4.52% for f, and 3.53% for K. After the training phase, the ML methods demonstrated a substantial decrease in computational time, being approximately 232 times faster than the conventional methods. Conclusions: The findings suggest that ML algorithms can enhance the efficiency of dMRI model analysis and offer new perspectives on the microstructural and functional organization of biological tissues. This paper also discusses the limitations and future directions of ML-based dMRI analysis.

Contrast-agent-free state-of-the-art MRI on cerebral small vessel disease-part 1. ASL, IVIM, and CVR.

Cerebral small vessel disease (cSVD), a common cause of stroke and dementia, is traditionally considered the small vessel equivalent of large artery occlusion or rupture that leads to cortical and subcortical brain damage. Microvessel endothelial dysfunction can also contribute to it. Brain imaging, including MRI, is useful to show the presence of lesions of several types, although the association between conventional MRI measures and clinical features of cSVD is not always concordant. We assessed the additional contribution of contrast-agent-free, state-of-the-art MRI techniques such as arterial spin labeling (ASL), diffusion tensor imaging, functional MRI, and intravoxel incoherent motion (IVIM) applied to cSVD in the existing literature. We performed a review following the PICO Worksheet and Search Strategy, including original papers in English, published between 2000 and 2022. For each MRI method, we extracted information about their contributions, in addition to those established with traditional MRI methods and related information about the origins, pathology, markers, and clinical outcomes in cSVD. This paper presents the first part of the review, which includes 37 studies focusing on ASL, IVIM, and cerebrovascular reactivity (CVR) measures. In general, they have shown that, in addition to white matter hyperintensities, alterations in other neuroimaging parameters such as blood flow and CVR also indicate the presence of cSVD. Such quantitative parameters were also related to cSVD risk factors. Therefore, they are promising, noninvasive tools to explore questions that have not yet been clarified about this clinical condition. However, protocol standardization is essential to increase their clinical use.

Feasibility of intravoxel incoherent motion in the assessment of tumor microvasculature and blood-brain barrier integrity: a case-based evaluation of gliomas.

OBJECTIVE: To evaluate the feasibility of intravoxel incoherent motion (IVIM) in assessing blood-brain barrier (BBB) integrity and microvasculature in tumoral tissue of glioma patients. METHODS: Images from 8 high-grade and 4 low-grade glioma patients were acquired on a 3 T MRI scanner. Acquisition protocol included pre- and post-contrast T1- and T2-weighted imaging, FLAIR, dynamic susceptibility contrast (DSC), and susceptibility-weighted imaging (SWI). In addition, IVIM was acquired with 15 b-values and fitted under the non-negative least square (NNLS) model to output the diffusion (D) and pseudo-diffusion (D*) coefficients, perfusion fraction (f), and f times D* (fD*) maps. RESULTS: IVIM perfusion-related maps were sensitive to (1) blood flow and perfusion alterations within the microvasculature of brain tumors, in agreement with intra-tumoral susceptibility signal (ITSS); (2) enhancing areas of BBB breakdown in agreement with DSC maps as well as areas of BBB abnormality that was not detected on DSC maps; (3) enhancing perfusion changes within edemas; (4) detecting early foci of increased perfusion within low-grade gliomas. CONCLUSION: The results suggest IVIM may be a promising approach to delineate tumor extension and progression in size, and to predict histological grade, which are clinically relevant information that characterize tumors and guide therapeutic decisions in patients with glioma.

Comparative study between monoexponential and biexponential diffusion weighted imaging sequences in multiparametric prostate magnetic resonance imaging / Estudo comparativo entre as sequências de difusão monoexponencial e biexponencial na ressonância magnética multiparamétrica da próstata

ABSTRACT Objective: To compare qualitatively and quantitatively, in terms of image quality, a new biexponential diffusion sequence protocol with the standard monoexponential diffusion protocol on multiparametric prostate magnetic resonance imaging. Methods: This study had a prospective data collection and cross-sectional analysis. Between August and November 2017, a total of 70 patients who underwent multiparametric prostate magnetic resonance imaging due to clinical suspicion of prostatic neoplasia were recruited. The images obtained were evaluated by two independent readers regarding subjective/qualitative criteria (six criteria) and objective/quantitative criteria (three criteria), always comparing the monoexponential to biexponential acquisition protocols. The results were compared by statistical analysis (interobserver agreement − Gwet coefficient; analysis of the qualitative variables − Stuart-Maxwell test; and analysis of the quantitative variables − Wilcoxon test). Results: After exclusion of four patients, the final sample consisted of 66 patients. A good/excellent inter observer agreement was stablished for subjective criteria (except in one criteria). For the qualitative analysis the amount of good or excellent evaluations was higher for the monoexponential protocol (except in one category), with evidence of significant differences for three criteria (diffusion weighted imaging global quality; diffusion weighted imaging signal-to-noise ratio; and apparent diffusion coefficient signal-to-noise ratio). For the quantitative data analysis, the monoexponential protocol showed less variability of the anteroposterior diameters, meaning less distortion of the images, and better estimated signal-to-noise ratio. Conclusion: In our data, the quality of the images of the monoexponential standard diffusion sequence was qualitatively and quantitatively superior to those of the biexponential diffusion weighted imaging sequence.

RESUMO Objetivo: Comparar qualitativa e quantitativamente, em termos de qualidade de imagem, um novo protocolo de sequência de difusão biexponencial com o protocolo de difusão monoexponencial padrão, em ressonância magnética multiparamétrica da próstata. Métodos: Estudo com coleta prospectiva e análise transversal. Entre agosto e novembro de 2017, foram recrutados 70 pacientes que realizaram ressonância magnética multiparamétrica da próstata, por suspeita de neoplasia prostática. As imagens obtidas por ambas as sequências foram avaliadas por dois leitores independentes, quanto a critérios de avaliação subjetiva/qualitativa (seis critérios) e objetiva/quantitativa (três critérios), sempre comparando os protocolos de aquisição monoexponencial e biexponencial. Os resultados foram comparados por análise estatística (concordância interobservador − coeficiente de Gwet; análise das variáveis qualitativas − teste de Stuart-Maxwell; e análise das variáveis quantitativas − testes de Wilcoxon). Resultados: Após exclusão de quatro pacientes, a amostra final foi composta por 66 pacientes. Uma boa/excelente concordância interobservador foi estabelecida para critérios subjetivos (exceto em um critério). Para a análise qualitativa, a quantidade de avaliações boas ou excelentes foi maior para o protocolo monoexponencial (exceto em uma categoria), com evidências de diferenças significativas para três critérios (qualidade global da imagem ponderada em difusão, relação sinal-ruído na imagem ponderada em difusão e relação sinal-ruído ADC). Para a análise quantitativa dos dados, o protocolo monoexponencial apresentou menor variabilidade dos diâmetros anteroposteriores, o que significou menos distorção das imagens, e melhor relação sinal-ruído estimada. Conclusão: Em nossos dados, a qualidade das imagens da sequência de difusão padrão monoexponencial foi qualitativa e quantitativamente superior àquelas da sequência teste biexponencial.