Dataset Information

Deep-learned time-signal intensity pattern analysis using an autoencoder captures magnetic resonance perfusion heterogeneity for brain tumor differentiation.

ABSTRACT: Current image processing methods for dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI) do not capture complex dynamic information of time-signal intensity curves. We investigated whether an autoencoder-based pattern analysis of DSC MRI captured representative temporal features that improves tissue characterization and tumor diagnosis in a multicenter setting. The autoencoder was applied to the time-signal intensity curves to obtain representative temporal patterns, which were subsequently learned by a convolutional neural network. This network was trained with 216 preoperative DSC MRI acquisitions and validated using external data (n = 43) collected with different DSC acquisition protocols. The autoencoder applied to time-signal intensity curves and clustering obtained nine representative clusters of temporal patterns, which accurately identified tumor and non-tumoral tissues. The dominant clusters of temporal patterns distinguished primary central nervous system lymphoma (PCNSL) from glioblastoma (AUC 0.89) and metastasis from glioblastoma (AUC 0.95). The autoencoder captured DSC time-signal intensity patterns that improved identification of tumoral tissues and differentiation of tumor type and was generalizable across centers.

SUBMITTER: Park JE

PROVIDER: S-EPMC7723041 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep-learned time-signal intensity pattern analysis using an autoencoder captures magnetic resonance perfusion heterogeneity for brain tumor differentiation.

Park Ji Eun JE Kim Ho Sung HS Lee Junkyu J Cheong E-Nae EN Shin Ilah I Ahn Sung Soo SS Shim Woo Hyun WH

Scientific reports 20201208 1

Current image processing methods for dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI) do not capture complex dynamic information of time-signal intensity curves. We investigated whether an autoencoder-based pattern analysis of DSC MRI captured representative temporal features that improves tissue characterization and tumor diagnosis in a multicenter setting. The autoencoder was applied to the time-signal intensity curves to obtain representative temporal patterns, which wer ...[more]

PMID: 33293590

Similar Datasets

Project description:BackgroundWhile magnetic resonance imaging (MRI) provides high resolution anatomical images with sharp soft tissue contrast, magnetic resonance spectroscopy (MRS) enables non-invasive detection and measurement of biochemicals and metabolites. However, MRS has low signal-to-noise ratio (SNR) when concentrations of metabolites are in the range of millimolar. Standard approach of using a high number of signal averaging (NSA) to achieve sufficient SNR comes at the cost of a long acquisition time.PurposeWe propose to use deep-learning approaches to denoise MRS data without increasing NSA. This method has potential to reduce the acquisition time as well as improve SNR and quality of spectra, which could enhance the diagnostic value and broaden the clinical applications of MRS.MethodsThe study was conducted using data collected from the brain spectroscopy phantom and human subjects. We utilized a stack auto-encoder (SAE) network to train deep learning models for denoising low NSA data (NSA = 1, 2, 4, 8, and 16) randomly truncated from high SNR data collected with high NSA (NSA = 192), which were also used to obtain the ground truth. We applied both self-supervised and fully-supervised training approaches and compared their performance of denoising low NSA data based on improvement in SNR. To prevent overfitting, the SAE network was trained in a patch-based manner. We then tested the denoising methods on noise-containing data collected from the phantom and human subjects, including data from brain tumor patients. We evaluated their performance by comparing the SNR levels and mean squared errors (MSEs) calculated for the whole spectra against high SNR "ground truth", as well as the value of chemical shift of N-acetyl-aspartate (NAA) before and after denoising.ResultsWith the SAE model, the SNR of low NSA data (NSA = 1) obtained from the phantom increased by 28.5% and the MSE decreased by 42.9%. For low NSA data of the human parietal and temporal lobes, the SNR increased by 32.9% and the MSE decreased by 63.1%. In all cases, the chemical shift of NAA in the denoised spectra closely matched with the high SNR spectra without significant distortion to the spectra after denoising. Furthermore, the denoising performance of the SAE model was more effective in denoising spectra with higher noise levels.ConclusionsThe reported SAE denoising method is a model-free approach to enhance the SNR of MRS data collected with low NSA. With the denoising capability, it is possible to acquire MRS data with a few NSA, shortening the scan time while maintaining adequate spectroscopic information for detecting and quantifying the metabolites of interest. This approach has the potential to improve the efficiency and effectiveness of clinical MRS data acquisition by reducing the scan time and increasing the quality of spectroscopic data.

Project description:Patients with intramedullary signal intensity (SI) changes have a poor prognosis after surgical decompression in cervical compressive myelopathy (CCM); however, some patients show no clear relationship between the SI and postsurgical prognosis. This discrepancy may be because no comprehensive and proper quantitative evaluation exists to assess SI on magnetic resonance imaging (MRI). The purpose of this study was prospectively to evaluate the correlation between the clinical features, neurological outcome of patients with CCM, and the quantitative assessment of SI changes preoperatively and postoperatively, and the correlation with SI severity.A total of 112 patients with CCM at 1 or 2 levels underwent anterior cervical discectomy and fusion. We quantitatively analyzed MR signal changes on T1-weighted MR images (T1WI), gadolinium-diethylenetriaminepentaacetic acid (Gd-DTPA) contrast-enhanced T1WI, and T2-weighted MR images (T2WI) using the signal intensity ratio (SIR). We evaluated the correlations between various variables and neurological outcome using the Japanese Orthopedic Association (JOA) scale, and the severity of SI change by grade (i.e., grade 0 ["none"], grade 1 ["light"], and grade 2 ["bright on T2WI"]).Significant differences between the 3 grades existed in symptom duration, preoperative JOA score, SIR on T2WI, and JOA recovery ratio. The JOA recovery ratio was negatively correlated with symptom duration and the SIR on T2WI, and positively correlated with the preoperative JOA score and cord compression ratio, but not with the SIR on T1WI and contrast-enhanced T1WI. On the postoperative 12-month follow-up MRI, the JOA recovery ratio and SIR on T2WI of the SI reversal patients were better than those of the nonreversal patients. On multiple regression analysis, the SIR on T2WI was the main significant prognostic factor of surgical outcome.The grading system on T2WI provided reliable predictive information for neurological outcome. Quantitative alterations in the SI on preoperative and postoperative T2WI, but not T1WI or contrast-enhanced T1WI, reflected the clinical features, surgical outcomes, and the correlation with SI severity. The patients with a longer duration of symptoms, lower cord compression ratio, severe myelopathy, intense signal change (i.e., grade 2) on the spinal cord, and an SIR greater than 1.55 had a poor recovery after a surgical operation.

Dataset Information

Deep-learned time-signal intensity pattern analysis using an autoencoder captures magnetic resonance perfusion heterogeneity for brain tumor differentiation.

Publications

Deep-learned time-signal intensity pattern analysis using an autoencoder captures magnetic resonance perfusion heterogeneity for brain tumor differentiation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets