Automatic quality control in clinical ¹H MRSI of brain cancer

Abstract

MRSI grids frequently show spectra with poor quality, mainly because of the high sensitivity of MRS to field inhomogeneities. These poor quality spectra are prone to quantification and/or interpretation errors that can have a significant impact on the clinical use of spectroscopic data. Therefore, quality control of the spectra should always precede their clinical use. When performed manually, quality assessment of MRSI spectra is not only a tedious and time-consuming task, but is also affected by human subjectivity. Consequently, automatic, fast and reliable methods for spectral quality assessment are of utmost interest. In this article, we present a new random forest-based method for automatic quality assessment of (1)H MRSI brain spectra, which uses a new set of MRS signal features. The random forest classifier was trained on spectra from 40 MRSI grids that were classified as acceptable or non-acceptable by two expert spectroscopists. To account for the effects of intra-rater reliability, each spectrum was rated for quality three times by each rater. The automatic method classified these spectra with an area under the curve (AUC) of 0.976. Furthermore, in the subset of spectra containing only the cases that were classified every time in the same way by the spectroscopists, an AUC of 0.998 was obtained. Feature importance for the classification was also evaluated. Frequency domain skewness and kurtosis, as well as time domain signal-to-noise ratios (SNRs) in the ranges 50-75 ms and 75-100 ms, were the most important features. Given that the method is able to assess a whole MRSI grid faster than a spectroscopist (approximately 3 s versus approximately 3 min), and without loss of accuracy (agreement between classifier trained with just one session and any of the other labelling sessions, 89.88%; agreement between any two labelling sessions, 89.03%), the authors suggest its implementation in the clinical routine. The method presented in this article was implemented in jMRUI's SpectrIm plugin.

A convolutional neural network to filter artifacts in spectroscopic MRI

“…Furthermore, the network was wrapped into a framework that enabled rapid deployment of the filter into the clinical research workflow, and could be applied in under 2 minutes to high‐resolution EPSI data with whole‐brain coverage. The accuracy achieved is similar to that reported in previous studies using machine learning for MR spectral quality analysis, such as those using random forests with engineered spectral features . It is difficult to compare results across studies, as a result of variation in study design (e.g., how data were collected, the biases of the raters generating ground truth, and which parameters were chosen as features).…”

“…Another challenge in developing algorithms for spectral quality filtering is the low percentage of poor‐quality voxels present in a whole‐brain volume compared with good quality voxels, which yields an imbalance in class proportions and consequently can hinder algorithm performance . In the data set collected in this work, 72% of spectra were of good quality and 28% were of poor quality, which is similar to proportions (65‐84% acceptable spectra) observed in other works . To assess whether balanced class proportions would affect CNN performance, a random minority oversampling scheme was implemented, in which data from the minority class (poor spectra) were randomly sampled multiple times to artificially increase the number of samples.…”

“…Several approaches have been developed to filter poor quality spectra from MRSI data sets, including exclusion criteria based on peak linewidths, reliability testing, Cramer‐Rao bounds, and machine‐learning techniques such as random forest classifiers . In each of these algorithms, the classification of spectral quality is based on a collection of “engineered” features as input; these features are categorical or ordinal vectors that seek to summarize the information encoded in each spectrum.…”

“…Common to all of these approaches is that the definition of features is required before performing any analysis. Thus, these features capture criteria that MRS experts explicitly believe are important to spectral quality, and machine‐learning algorithms built using these engineered features have shown promise as spectral quality filters . In contrast to systems that model expert beliefs explicitly, deep learning broadly defines a category of machine learning algorithms that can learn underlying features from raw data without the need for any a priori definition of such features, and have been able to shatter benchmarks in natural language processing, medical image segmentation, survival analysis, and identification of pathology in medical images .…”

Improving labeling efficiency in automatic quality control of MRSI data

“…Residual water peak removal was performed using HLSVD prior to feature extraction. A total of 47 features were extracted from time‐domain (TD) and frequency‐domain (FD) magnitude spectra, as in . The following groups of features were used: Local maximum peak SNR (FD) Local mean SNR (FD and TD) Local relative change (TD) Global TD features (maximum signal strength, time point of maximum signal strength, mean, standard deviation, skewness, kurtosis) Global FD features (maximum signal strength, parts per million value (ppm) of maximum signal strength, mean, standard deviation, skewness, kurtosis) …”