Data-driven, voxel-based analysis of brain PET images: Application of PCA and LASSO methods to visualize and quantify patterns of neurodegeneration

被引:13
作者
Klyuzhin, Ivan S. [1 ]
Fu, Jessie F. G. [2 ]
Hong, Andy [2 ]
Sacheli, Matthew [3 ]
Shenkov, Nikolay [2 ]
Matarazzo, Michele [3 ]
Rahmim, Arman [4 ]
Stoessl, A. Jon [3 ]
Sossi, Vesna [2 ]
机构
[1] Univ British Columbia, Dept Med, Div Neurol, Vancouver, BC, Canada
[2] Univ British Columbia, Dept Phys & Astron, Vancouver, BC, Canada
[3] Univ British Columbia, Pacific Parkinsons Res Ctr, Vancouver, BC, Canada
[4] Johns Hopkins Univ, Dept Radiol, Baltimore, MD USA
来源
PLOS ONE | 2018年 / 13卷 / 11期
基金
加拿大自然科学与工程研究理事会; 加拿大健康研究院;
关键词
PARKINSONS-DISEASE; SELECTION;
D O I
10.1371/journal.pone.0206607
中图分类号
O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];
学科分类号
07 ; 0710 ; 09 ;
摘要
Spatial patterns of radiotracer binding in positron emission tomography (PET) images may convey information related to the disease topology. However, this information is not captured by the standard PET image analysis that quantifies the mean radiotracer uptake within a region of interest (ROI). On the other hand, spatial analyses that use more advanced radiomic features may be difficult to interpret. Here we propose an alternative data-driven, voxel-based approach to spatial pattern analysis in brain PET, which can be easily interpreted. We apply principal component analysis (PCA) to identify voxel covariance patterns, and optimally combine several patterns using the least absolute shrinkage and selection operator (LASSO). The resulting models predict clinical disease metrics from raw voxel values, allowing for inclusion of clinical covariates. The analysis is performed on high-resolution PET images from healthy controls and subjects affected by Parkinson's disease (PD), acquired with a pre-synaptic and a post-synaptic dopaminergic PET tracer. We demonstrate that PCA identifies robust and tracer-specific binding patterns in sub-cortical brain structures; the patterns evolve as a function of disease progression. Principal component LASSO (PC-LASSO) models of clinical disease metrics achieve higher predictive accuracy compared to the mean tracer binding ratio (BR) alone: the cross-validated test mean squared error of adjusted disease duration (motor impairment score) was 16.3 +/- 0.17 years(2) (9.7 +/- 0.15) with mean BR, versus 14.4 +/- 0.18 years(2) (8.9 +/- 0.16) with PC-LASSO. We interpret the best-performing PC-LASSO models in the spatial sense and discuss them with reference to the PD pathology and somatotopic organization of the striatum. PC-LASSO is thus shown to be a useful method to analyze clinically-relevant tracer binding patterns, and to construct interpretable, imaging-based predictive models of clinical metrics.
引用
收藏
页数:20
相关论文
共 46 条
  • [1] [Anonymous], IEEE S C REC NUCL SC
  • [2] [Anonymous], P SPIE INT SOC OPTIC
  • [3] Beyond imaging: The promise of radiomics
    Avanzo, Michele
    Stancanello, Joseph
    El Naqa, Issam
    [J]. PHYSICA MEDICA-EUROPEAN JOURNAL OF MEDICAL PHYSICS, 2017, 38 : 122 - 139
  • [4] Revisiting the Robustness of PET-Based Textural Features in the Context of Multi-Centric Trials
    Bailly, Clement
    Bodet-Milin, Caroline
    Couespel, Solene
    Necib, Hatem
    Kraeber-Bodere, Franoise
    Ansquer, Catherine
    Carlier, Thomas
    [J]. PLOS ONE, 2016, 11 (07):
  • [5] VALID POST-SELECTION INFERENCE
    Berk, Richard
    Brown, Lawrence
    Buja, Andreas
    Zhang, Kai
    Zhao, Linda
    [J]. ANNALS OF STATISTICS, 2013, 41 (02) : 802 - 837
  • [6] Blinder SAL, 2014, 2014 IEEE NUCLEAR SCIENCE SYMPOSIUM AND MEDICAL IMAGING CONFERENCE (NSS/MIC)
  • [7] Classical Statistics and Statistical Learning in Imaging Neuroscience
    Bzdok, Danilo
    [J]. FRONTIERS IN NEUROSCIENCE, 2017, 11
  • [8] Inference in the age of big data: Future perspectives on neuroscience
    Bzdok, Danilo
    Yeo, B. T. Thomas
    [J]. NEUROIMAGE, 2017, 155 : 549 - 564
  • [9] Neuroimaging Research: From Null-Hypothesis Falsification to Out-of-Sample Generalization
    Bzdok, Danilo
    Varoquaux, Gael
    Thirion, Bertrand
    [J]. EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 2017, 77 (05) : 868 - 880
  • [10] False Discovery Rates in PET and CT Studies with Texture Features: A Systematic Review
    Chalkidou, Anastasia
    O'Doherty, Michael J.
    Marsden, Paul K.
    [J]. PLOS ONE, 2015, 10 (05):