Share this content in WeChat
Technical Article
Normalizing radiomics features from multiscale structural MRI of the adolescent brain
WANG Yao  MA Huan  ZHANG Dafu  WANG Hongbo  SUN Dewei  LI Kun  YANG Jianzhong 

WANG Y, MA H, ZHANG D F, et al. Normalizing radiomics features from multiscale structural MRI of the adolescent brain[J]. Chin J Magn Reson Imaging, 2023, 14(8): 100-107. DOI:10.12015/issn.1674-8034.2023.08.016.

[Abstract] Objectives To explore the value of combination of N4 bias field correction, histogram-matching (HM) normalization and ComBat harmonization to reduce the "scanner effect" of radiomics features from brain MRI.Materials and Methods Three-dimensional T1 weighted image (3D-T1WI) and diffusion tensor imaging (DTI) of the brain was performed in 23 healthy volunteers with three MRI scanners (Philip 1.5 T, Philip 3.0 T, GE 3.0 T). Computational Anatomy toolbox (Cat 12) and FMRIB's software library (FSL) were used for preprocessing. Then, N4 bias field correction and HM normalization were performed on the preprocessed T1WI and DTI. Finally, LIFEx software was used to extract radiomics features of gray and white matter and then Combat harmonization were carried out. The Shapiro-Wilk test was used to exam the normality and the analysis of variance (ANOVA) and Tukey honestly significant difference (Tukey-HSD) test were used to compare the radiomics features of the three scanners, and the Bartlett spherical test was used to estimate whether the variance was uniform. The differences between scanners in the number of radiomics features and numerical statistical distribution in each processing were qualitatively and quantitatively evaluated.Results A total of 10 males and 13 females were enrolled. There was no significant difference in age (t=1.090, P=0.316), education years (t=-0.638, P=0.574) and CES-D score (t=-0.670, P=0.510) between the males and females (P>0.05). In the original images acquired by the three MRI scanners, the distribution range and peak value of the intensity histogram were not aligned. When N4 bias field correction was performed using 5-level (50 iterations) full mask, the intensity variation coefficient of brain tissue among the three scanners was the lowest. N4 correction sharpened the intensity peak, HM normalized and aligned each intensity peak, and Combat harmonization further aligned the image intensity distribution range and peak of the three MRI. The process (N4 bias field correction, HM normalization and ComBat harmonization) had the same influence on T1WI and DTI sequences. Through the combination of N4 correction, HM normalization and Combat harmonization, the percentage of radiomics features with differences between scanners was reduced from 88.6% (70/79) before bias field correction to 3.8% (3/79) after ComBat harmonization. At the same time, the percentage of radiomics features with differences between VOI of gray and white matter increased from 43.0% (34/79) before bias field correction to 84.8% (67/79) after ComBat harmonization.Conclusions The combination of N4 bias field correction, HM normalization and ComBat harmonization can effectively eliminate the "scanner effect" of the brain structural MRI and thereby help to incorporate multi-center MRI data across scanners.
[Keywords] adolescent;brain;radiomics feature;harmonization;magnetic resonance imaging

WANG Yao1   MA Huan1   ZHANG Dafu1   WANG Hongbo1   SUN Dewei1   LI Kun1*   YANG Jianzhong2  

1 Department of Radiology, the Third Affiliated Hospital of Kunming Medical University, Kunming 650018, China

2 Department of Psychiatry, the Second Affiliated Hospital of Kunming Medical University, Kunming 650101, China

Corresponding author: Li K, E-mail:

Conflicts of interest   None.

Received  2023-03-13
Accepted  2023-07-27
DOI: 10.12015/issn.1674-8034.2023.08.016
WANG Y, MA H, ZHANG D F, et al. Normalizing radiomics features from multiscale structural MRI of the adolescent brain[J]. Chin J Magn Reson Imaging, 2023, 14(8): 100-107. DOI:10.12015/issn.1674-8034.2023.08.016.

BENTO M, FANTINI I, PARK J, et al. Deep learning in large and multi-site structural brain MR imaging datasets[J/OL]. Front Neuroinform, 2021, 15: 805669 [2023-03-13]. DOI: 10.3389/fninf.2021.805669.
MONTE-RUBIO G C, SEGURA B, STRAFELLA A P, et al. Parameters from site classification to harmonize MRI clinical studies: application to a multi-site Parkinson's disease dataset[J]. Hum Brain Mapp, 2022, 43(10): 3130-3142. DOI: 10.1002/hbm.25838.
KOIKE S, TANAKA S C, OKADA T, et al. Brain/MINDS beyond human brain MRI project: a protocol for multi-level harmonization across brain disorders throughout the lifespan[J/OL]. Neuroimage Clin, 2021, 30: 102600 [2023-03-13]. DOI: 10.1016/j.nicl.2021.102600.
JOHNSON W E, LI C, RABINOVIC A. Adjusting batch effects in microarray expression data using empirical Bayes methods[J]. Biostatistics, 2007, 8(1): 118-127. DOI: 10.1093/biostatistics/kxj037.
LEEK J T, SCHARPF R B, BRAVO H C, et al. Tackling the widespread and critical impact of batch effects in high-throughput data[J]. Nat Rev Genet, 2010, 11(10): 733-739. DOI: 10.1038/nrg2825.
LEITHNER D, SCHÖDER H, HAUG A, et al. Impact of ComBat harmonization on PET radiomics-based tissue classification: a dual-center PET/MRI and PET/CT study[J]. J Nucl Med, 2022, 63(10): 1611-1616. DOI: 10.2967/jnumed.121.263102.
SAINT MARTIN M J, ORLHAC F, AKL P, et al. A radiomics pipeline dedicated to Breast MRI: validation on a multi-scanner phantom study[J]. MAGMA, 2021, 34(3): 355-366. DOI: 10.1007/s10334-020-00892-y.
KEENAN K E, GIMBUTAS Z, DIENSTFREY A, et al. Multi-site, multi-platform comparison of MRI T1 measurement using the system phantom[J/OL]. PLoS One, 2021, 16(6): e0252966 [2023-03-13]. DOI: 10.1371/journal.pone.0252966.
FORTIN J P, PARKER D, TUNÇ B, et al. Harmonization of multi-site diffusion tensor imaging data[J/OL]. Neuroimage, 2017, 161: 149-170 [2023-03-13]. DOI: 10.1016/j.neuroimage.2017.08.047.
ORLHAC F, LECLER A, SAVATOVSKI J, et al. How can we combat multicenter variability in MR radiomics? Validation of a correction procedure[J]. Eur Radiol, 2021, 31(4): 2272-2280. DOI: 10.1007/s00330-020-07284-9.
BLODGETT J M, LACHANCE C C, STUBBS B, et al. A systematic review of the latent structure of the Center for Epidemiologic Studies Depression Scale (CES-D) amongst adolescents[J/OL]. BMC Psychiatry, 2021, 21(1): 197 [2023-03-13]. DOI: 10.1186/s12888-021-03206-1.
KLIEM S, BELLER J, TIBUBOS A N, et al. A reanalysis of the center for epidemiological studies depression scale (CES-D) using non-parametric item response theory[J/OL]. Psychiatry Res, 2020, 290: 113132 [2023-03-13]. DOI: 10.1016/j.psychres.2020.113132.
American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 5th edn[M]. Washington, DC: American Psychiatric Association, 2013.
GHOSH A, KAUR S, SHAH R, et al. Surface-based brain morphometry in schizophrenia vs. cannabis-induced psychosis: a controlled comparison[J/OL]. J Psychiatr Res, 2022, 155: 286-294 [2023-03-13]. DOI: 10.1016/j.jpsychires.2022.09.034.
GUO C J, FERREIRA D, FINK K, et al. Repeatability and reproducibility of FreeSurfer, FSL-SIENAX and SPM brain volumetric measurements and the effect of lesion filling in multiple sclerosis[J]. Eur Radiol, 2019, 29(3): 1355-1364. DOI: 10.1007/s00330-018-5710-x.
PIRZADA S, UDDIN M N, FIGLEY T D, et al. Spatial normalization of multiple sclerosis brain MRI data depends on analysis method and software package[J/OL]. Magn Reson Imaging, 2020, 68: 83-94 [2023-03-13]. DOI: 10.1016/j.mri.2020.01.016.
TUSTISON N J, AVANTS B B, COOK P A, et al. N4ITK: improved N3 bias correction[J]. IEEE Trans Med Imaging, 2010, 29(6): 1310-1320. DOI: 10.1109/TMI.2010.2046908.
YUSHKEVICH P A, YANG-GAO, GERIG G. ITK-SNAP: an interactive tool for semi-automatic segmentation of multi-modality biomedical images[J/OL]. Annu Int Conf IEEE Eng Med Biol Soc, 2016: 3342-3345 [2023-03-13]. DOI: 10.1109/EMBC.2016.7591443.
CLARKE W T, STAGG C J, JBABDI S. FSL-MRS: an end-to-end spectroscopy analysis package[J]. Magn Reson Med, 2021, 85(6): 2950-2964. DOI: 10.1002/mrm.28630.
MISHRA V R, SREENIVASAN K R, ZHUANG X W, et al. Influence of analytic techniques on comparing DTI-derived measurements in early stage Parkinson's disease[J/OL]. Heliyon, 2019, 5(4): e01481 [2023-03-13]. DOI: 10.1016/j.heliyon.2019.e01481.
LI Y P, AMMARI S, BALLEYGUIER C, et al. Impact of preprocessing and harmonization methods on the removal of scanner effects in brain MRI radiomic features[J/OL]. Cancers, 2021, 13(12): 3000 [2023-03-13]. DOI: 10.3390/cancers13123000.
NIOCHE C, ORLHAC F, BOUGHDAD S, et al. LIFEx: a freeware for radiomic feature calculation in multimodality imaging to accelerate advances in the characterization of tumor heterogeneity[J]. Cancer Res, 2018, 78(16): 4786-4789. DOI: 10.1158/0008-5472.CAN-18-0125.
REINHOLD J C, DEWEY B E, CARASS A, et al. Evaluating the impact of intensity normalization on MR image synthesis[J/OL]. Proc SPIE Int Soc Opt Eng, 2019, 10949: 109493H [2023-03-13]. DOI: 10.1117/12.2513089.
SHAH M, XIAO Y M, SUBBANNA N, et al. Evaluating intensity normalization on MRIs of human brain with multiple sclerosis[J]. Med Image Anal, 2011, 15(2): 267-282. DOI: 10.1016/
ZWANENBURG A, VALLIÈRES M, ABDALAH M A, et al. The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping[J]. Radiology, 2020, 295(2): 328-338. DOI: 10.1148/radiol.2020191145.
LAMBIN P, LEIJENAAR R T H, DEIST T M, et al. Radiomics: the bridge between medical imaging and personalized medicine[J]. Nat Rev Clin Oncol, 2017, 14(12): 749-762. DOI: 10.1038/nrclinonc.2017.141.
MADABHUSHI A, UDUPA J K. Interplay between intensity standardization and inhomogeneity correction in MR image processing[J]. IEEE Trans Med Imaging, 2005, 24(5): 561-576. DOI: 10.1109/TMI.2004.843256.
LIU Z Y, LI Z L, QU J R, et al. Radiomics of multiparametric MRI for pretreatment prediction of pathologic complete response to neoadjuvant chemotherapy in breast cancer: a multicenter study[J]. Clin Cancer Res, 2019, 25(12): 3538-3547. DOI: 10.1158/1078-0432.CCR-18-3190.
ISAKSSON L J, RAIMONDI S, BOTTA F, et al. Effects of MRI image normalization techniques in prostate cancer radiomics[J/OL]. Phys Med, 2020, 71: 7-13 [2023-03-13]. DOI: 10.1016/j.ejmp.2020.02.007.
WACHINGER C, RIECKMANN A, PÖLSTERL S, et al. Detect and correct bias in multi-site neuroimaging datasets[J/OL]. Med Image Anal, 2021, 67: 101879 [2023-03-13]. DOI: 10.1016/
UM H, TIXIER F, BERMUDEZ D, et al. Impact of image preprocessing on the scanner dependence of multi-parametric MRI radiomic features and covariate shift in multi-institutional glioblastoma datasets[J/OL]. Phys Med Biol, 2019, 64(16): 165011 [2023-03-13]. DOI: 10.1088/1361-6560/ab2f44.

PREV A preliminary study on diagnostic model of placenta implantation based on magnetic resonance image feature machine learning
NEXT Value of frequency offset in image quality optimization of fast spin echo T1WI SPIR fat suppression sequence for pediatric thoracic spine MRI

Tel & Fax: +8610-67113815    E-mail: