A Combined Radiomics and Machine Learning Approach to Overcome the Clinicoradiologic Paradox in Multiple Sclerosis

被引:15
作者
Pontillo, G. [1 ,2 ]
Tommasin, S. [6 ]
Cuocolo, R. [3 ,4 ]
Petracca, M. [5 ]
Petsas, N. [7 ]
Ugga, L. [1 ]
Carotenuto, A. [5 ]
Pozzilli, C. [6 ]
Iodice, R. [5 ]
Lanzillo, R. [5 ]
Quarantelli, M. [2 ,8 ]
Morra, V. Brescia [5 ]
Tedeschi, E. [1 ]
Pantano, P. [6 ,7 ]
Cocozza, S. [1 ]
机构
[1] Univ Naples Federico II, Dept Adv Biomed Sci, Naples, Italy
[2] Univ Naples Federico II, Dept Elect Engn & Informat Technol, Naples, Italy
[3] Univ Naples Federico II, Dept Clin Med & Surg, Naples, Italy
[4] Univ Naples Federico II, Lab Augmented Real Hlth Monitoring, Dept Elect Engn & Informat Technol, Naples, Italy
[5] Univ Naples Federico II, Dept Neurosci & Reprod & Odontostomatol Sci, Naples, Italy
[6] Sapienza Univ Rome, Dept Human Neurosci, Rome, Italy
[7] IRCCS, Ist Neurol Mediterraneo, Pozzilli, Italy
[8] CNR, Inst Biostruct & Bioimaging, Naples, Italy
关键词
MAGNIMS CONSENSUS GUIDELINES; MRI; DISABILITY; ATROPHY;
D O I
10.3174/ajnr.A7274
中图分类号
R74 [神经病学与精神病学];
学科分类号
摘要
BACKGROUND AND PURPOSE: Conventional MR imaging explains only a fraction of the clinical outcome variance in multiple sclerosis. We aimed to evaluate machine learning models for disability prediction on the basis of radiomic, volumetric, and connectivity features derived from routine brain MR images. MATERIALS AND METHODS: In this retrospective cross-sectional study, 3T brain MR imaging studies of patients with multiple sclerosis, including 3D T1-weighted and T2-weighted FLAIR sequences, were selected from 2 institutions. T1-weighted images were processed to obtain volume, connectivity score (inferred from the T2 lesion location), and texture features for an atlas-based set of GM regions. The site 1 cohort was randomly split into training (n?=?400) and test (n?=?100) sets, while the site 2 cohort (n?=?104) constituted the external test set. After feature selection of clinicodemographic and MR imaging?derived variables, different machine learning algorithms predicting disability as measured with the Expanded Disability Status Scale were trained and cross-validated on the training cohort and evaluated on the test sets. The effect of different algorithms on model performance was tested using the 1-way repeated-measures ANOVA. RESULTS: The selection procedure identified the 9 most informative variables, including age and secondary-progressive course and a subset of radiomic features extracted from the prefrontal cortex, subcortical GM, and cerebellum. The machine learning models predicted disability with high accuracy (r approaching 0.80) and excellent intra- and intersite generalizability (r???0.73). The machine learning algorithm had no relevant effect on the performance. CONCLUSIONS: The multidimensional analysis of brain MR images, including radiomic features and clinicodemographic data, is highly informative of the clinical status of patients with multiple sclerosis, representing a promising approach to bridge the gap between conventional imaging and disability.
引用
收藏
页码:1927 / 1933
页数:7
相关论文
共 36 条
[1]   Machine learning for neuroirnaging with scikit-learn [J].
Abraham, Alexandre ;
Pedregosa, Fabian ;
Eickenberg, Michael ;
Gervais, Philippe ;
Mueller, Andreas ;
Kossaifi, Jean ;
Gramfort, Alexandre ;
Thirion, Bertrand ;
Varoquaux, Gael .
FRONTIERS IN NEUROINFORMATICS, 2014, 8
[2]   Thalamic atrophy in multiple sclerosis: A magnetic resonance imaging marker of neurodegeneration throughout disease [J].
Azevedo, Christina J. ;
Cen, Steven Y. ;
Khadka, Sankalpa ;
Liu, Shuang ;
Kornak, John ;
Shi, Yonggang ;
Zheng, Ling ;
Hauser, Stephen L. ;
Pelletier, Daniel .
ANNALS OF NEUROLOGY, 2018, 83 (02) :223-234
[3]   MRI in multiple sclerosis: correlation with expanded disability status scale (EDSS) [J].
Barkhof, F .
MULTIPLE SCLEROSIS JOURNAL, 1999, 5 (04) :283-286
[4]   The clinico-radiological paradox in multiple sclerosis revisited [J].
Barkhof, F .
CURRENT OPINION IN NEUROLOGY, 2002, 15 (03) :239-245
[5]   Technical Note: Virtual phantom analyses for preprocessing evaluation and detection of a robust feature set for MRI-radiomics of the brain [J].
Bologna, Marco ;
Corino, Valentina ;
Mainardi, Luca .
MEDICAL PHYSICS, 2019, 46 (11) :5116-5123
[6]   Cortical lesions in multiple sclerosis [J].
Calabrese, Massimiliano ;
Filippi, Massimo ;
Gallo, Paolo .
NATURE REVIEWS NEUROLOGY, 2010, 6 (08) :438-444
[7]  
Chard Declan, 2017, F1000Res, V6, P1828, DOI 10.12688/f1000research.11932.1
[8]   Magnetization transfer ratio evolution with demyelination and remyelination in multiple sclerosis lesions [J].
Chen, Jacqueline T. ;
Collins, D. Louis ;
Atkins, Harold L. ;
Freedman, Mark S. ;
Arnold, Douglas L. .
ANNALS OF NEUROLOGY, 2008, 63 (02) :254-262
[9]   Cerebellar lobule atrophy and disability in progressive MS [J].
Cocozza, Sirio ;
Petracca, Maria ;
Mormina, Enricomaria ;
Buyukturkoglu, Korhan ;
Podranski, Kornelius ;
Heinig, Monika M. ;
Pontillo, Giuseppe ;
Russo, Camilla ;
Tedeschi, Enrico ;
Russo, Cinzia Valeria ;
Costabile, Teresa ;
Lanzillo, Roberta ;
Harel, Asaff ;
Klineova, Sylvia ;
Miller, Aaron ;
Brunetti, Arturo ;
Morra, Vincenzo Brescia ;
Lublin, Fred ;
Inglese, Matilde .
JOURNAL OF NEUROLOGY NEUROSURGERY AND PSYCHIATRY, 2017, 88 (12) :1065-1072
[10]   Deep gray matter volume loss drives disability worsening in multiple sclerosis [J].
Eshaghi, Arman ;
Prados, Ferran ;
Brownlee, Wallace J. ;
Altmann, Daniel R. ;
Tur, Carmen ;
Cardoso, M. Jorge ;
De Angelis, Floriana ;
van de Pavert, Steven H. ;
Cawley, Niamh ;
De Stefano, Nicola ;
Stromillo, M. Laura ;
Battaglini, Marco ;
Ruggieri, Serena ;
Gasperini, Claudio ;
Filippi, Massimo ;
Rocca, Maria A. ;
Rovira, Alex ;
Sastre-Garriga, Jaume ;
Vrenken, Hugo ;
Leurs, Cyra E. ;
Killestein, Joep ;
Pirpamer, Lukas ;
Enzinger, Christian ;
Ourselin, Sebastien ;
Wheeler-Kingshott, Claudia A. M. Gandini ;
Chard, Declan ;
Thompson, Alan J. ;
Alexander, Daniel C. ;
Barkhof, Frederik ;
Ciccarelli, Olga .
ANNALS OF NEUROLOGY, 2018, 83 (02) :210-222