Evaluation of deep learning models for quality control of MR spectra

Frontiers in Neuroscience

Neuroscience

Emily Xie¹  Yan Li¹  Huawei Liu¹  Sana Vaziri¹  Janine M. Lupo²  Duan Xu²  Hélène Ratiney³  Michaël Sdika³ 

[1] Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States;Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States;UC San Francisco/UC Berkeley Graduate Program in Bioengineering, San Francisco, CA, United States;Univ Lyon, INSA-Lyon, Université Claude Bernard Lyon 1, UJM-Saint Etienne, CNRS, Inserm, CREATIS UMR 5220, Lyon, France;

关键词: MR spectroscopy;    convolutional neural network;    random forest;    quality control;    machine learning;   

DOI  :  10.3389/fnins.2023.1219343

 received in 2023-05-09, accepted in 2023-08-10,  发布年份 2023

来源: Frontiers

【 摘 要 】

PurposeWhile 3D MR spectroscopic imaging (MRSI) provides valuable spatial metabolic information, one of the hurdles for clinical translation is its interpretation, with voxel-wise quality control (QC) as an essential and the most time-consuming step. This work evaluates the accuracy of machine learning (ML) models for automated QC filtering of individual spectra from 3D healthy control and patient datasets.MethodsA total of 53 3D MRSI datasets from prior studies (30 neurological diseases, 13 brain tumors, and 10 healthy controls) were included in the study. Three ML models were evaluated: a random forest classifier (RF), a convolutional neural network (CNN), and an inception CNN (ICNN) along with two hybrid models: CNN + RF, ICNN + RF. QC labels used for training were determined manually through consensus of two MRSI experts. Normalized and cropped real-valued spectra was used as input. A cross-validation approach was used to separate datasets into training/validation/testing sets of aggregated voxels.ResultsAll models achieved a minimum AUC of 0.964 and accuracy of 0.910. In datasets from neurological disease and controls, the CNN model produced the highest AUC (0.982), while the RF model achieved the highest AUC in patients with brain tumors (0.976). Within tumor lesions, which typically exhibit abnormal metabolism, the CNN AUC was 0.973 while that of the RF was 0.969. Data quality inference times were on the order of seconds for an entire 3D dataset, offering drastic time reduction compared to manual labeling.ConclusionML methods accurately and rapidly performed automated QC. Results in tumors highlights the applicability to a variety of metabolic conditions.

【 授权许可】

Unknown   
Copyright © 2023 Vaziri, Liu, Xie, Ratiney, Sdika, Lupo, Xu and Li.