To investigate the performance of deep learning (DL) based on fully convolutional neural network (FCNN) in segmenting brain tissues in a large cohort of multiple sclerosis (MS) patients.
Methods:
We developed a FCNN model to segment brain tissues, including T2-hyperintense MS lesions. The training, validation, and testing of FCNN were based on ~1000 magnetic resonance imaging (MRI) datasets acquired on relapsing–remitting MS patients, as a part of a phase 3 randomized clinical trial. Multimodal MRI data (dual-echo, FLAIR, and T1-weighted images) served as input to the network. Expert validated segmentation was used as the target for training the FCNN. We cross-validated our results using the leave-one-center-out approach.
Results:
We observed a high average (95% confidence limits) Dice similarity coefficient for all the segmented tissues: 0.95 (0.92–0.98) for white matter, 0.96 (0.93–0.98) for gray matter, 0.99 (0.98–0.99) for cerebrospinal fluid, and 0.82 (0.63–1.0) for T2 lesions. High correlations between the DL segmented tissue volumes and ground truth were observed (R2 > 0.92 for all tissues). The cross validation showed consistent results across the centers for all tissues.
Conclusion:
The results from this large-scale study suggest that deep FCNN can automatically segment MS brain tissues, including lesions, with high accuracy.
ReichDSLucchinettiCFCalabresiPA.Multiple sclerosis. N Engl J Med2018; 378(2): 169–180.
2.
LublinFDCofieldSSCutterGR, et al. Randomized study combining interferon and glatiramer acetate in multiple sclerosis. Ann Neurol2013; 73(3): 327–340.
3.
DanelakisATheoharisTVerganelakisDA.Survey of automated multiple sclerosis lesion segmentation techniques on magnetic resonance imaging. Comput Med Imaging Graph2018; 70: 83–100.
4.
de SitterASteenwijkMDRuetA, et al. Performance of five research-domain automated WM lesion segmentation methods in a multi-center MS study. NeuroImage2017; 163: 106–114.
5.
EggerCOpferRWangC, et al. MRI FLAIR lesion segmentation in multiple sclerosis: Does automated segmentation hold up with manual annotation?NeuroImage Clin2017; 13: 264–270.
6.
CarassARoySJogA, et al. Longitudinal multiple sclerosis lesion segmentation: Resource and challenge. NeuroImage2017; 148: 77–102.
7.
GhribiOSellamiLBenSlimaM, et al. An advanced MRI multi-modalities segmentation methodology dedicated to multiple sclerosis lesions exploration and differentiation. IEEE Trans Nanobioscience2017; 16(8): 656–665.
8.
ValcarcelAMLinnKAVandekarSN, et al. MIMoSA: An automated method for intermodal segmentation analysis of multiple sclerosis brain lesions. J Neuroimaging2018; 28: 389–398.
9.
RoccaMAComiGFilippiM.The role of T1-weighted derived measures of neurodegeneration for assessing disability progression in multiple sclerosis. Front Neurol2017; 8: 433.
10.
GoodfellowIBengioYCourvilleA, et al. Deep learning. Cambridge: MIT Press, 2016.
RonnebergerOFischerPBroxT. U-Net: Convolutional networks for biomedical image segmentation. In: International conference on medical image computing and computer-assisted intervention, Munich, 5–9October2015, pp. 234–241. Cham: Springer.
13.
MoeskopsPde BresserJKuijfHJ, et alEvaluation of a deep learning approach for the segmentation of brain tissues and white matter hyperintensities of presumed vascular origin in MRI. NeuroImage: Clinical2018; 17: 251–262.
14.
RachmadiMFValdés-HernándezMdelC, et al. Segmentation of white matter hyperintensities using convolutional neural networks with global spatial information in routine clinical brain MRI with none or mild vascular pathology. Comput Med Imaging Graph2018; 66: 28–43.
15.
LaukampKRThieleFShakirinG, et al. Fully automated detection and segmentation of meningiomas using deep learning on routine multiparametric MRI. Eur Radiol2019; 29: 124–132.
16.
LindseyJWScottTFLynchSG, et al. The CombiRx trial of combined therapy with interferon and glatiramer cetate in relapsing remitting MS: Design and baseline characteristics. Mult Scler Relat Disord2012; 1(2): 81–86.
17.
NarayanaPAGovindarajanKAGoelP, et al. Regional cortical thickness in relapsing remitting multiple sclerosis: A multi-center study. NeuroImage Clin2013; 2(1): 120–131.
18.
PeronaPMalikJ.Scale-space and edge detection using anisotropic diffusion. IEEE Trans Pattern Anal Mach Intell1990; 12(7): 629–639.
19.
GerigGKublerOKikinisR, et al. Nonlinear anisotropic filtering of MRI data. IEEE Trans Med Imaging1992; 11(2): 221–232.
20.
NyúlLGUdupaJKZhangX.New variants of a method of MRI scale standardization. IEEE Trans Med Imaging2000; 19(2): 143–150.
21.
SajjaBRDattaSHeR, et al. Unified approach for multiple sclerosis lesion segmentation on brain MRI. Ann Biomed Eng2006; 341: 142–151.
22.
GlorotXBengioY. Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the thirteenth international conference on artificial intelligence and statistics, Sardinia, 13–15 May 2010, pp. 249–256.
23.
IsenseeFKickingerederPWickW, et al. Brain tumor segmentation and radiomics survival prediction: Contribution to the BRATS 2017 challenge. In: International MICCAI brainlesion workshop, Quebec City, 14 September 2017, pp. 287–297. Heidelberg, Germany: Springer Verlag.
24.
KingmaDPBaJL. Adam: A method for stochastic optimization. In: International conference on learning representations (ICLR), 2015, https://arxiv.org/pdf/1412.6980.pdf
AbadiMBarhamPChenJ, et al. TensorFlow: A system for large-scale machine learning. In: 12th USENIX symposium on operating systems design and implementation (OSDI‘16), Savannah, GA, 2–4 November 2016.
27.
BroschTTangLYWYooY, et al. Deep 3D convolutional encoder networks with shortcuts for multiscale feature integration applied to multiple sclerosis lesion segmentation. IEEE Trans Med Imaging2016; 35(5): 1229–1239.
28.
ValverdeSCabezasMRouraE, et al. Improving automated multiple sclerosis lesion segmentation with a cascaded 3D convolutional neural network approach. NeuroImage2017; 155: 159–168.
29.
TorreyLShavlikJ.Transfer learning. In: OlivasESGuerreroJDMSoberMM, et al. (eds) Handbook of research on machine learning applications and trends: Algorithms, methods, and techniques. Hershey, PA: IGI Global, 2010, pp. 242–264.
30.
NelsonFPoonawallaAHHouP, et al. Improved identification of intracortical lesions in multiple sclerosis with phase-sensitive inversion recovery in combination with fast double inversion recovery MR imaging. AJNR Am J Neuroradiol2007; 28(9): 1645–1649.
