利用自动病变检测规划立体定向脑电图：可行性回顾性研究_Journal of Epilepsy

Authors：

 WagstylKonrad , AdlerSophie , PimpelBirgit , 郭玉洁　译 , 慕洁　审

Corresponding author：

WagstylKonrad, Email: k.wagstyl@ucl.ac.uk

Keywords：

深度学习; 癫痫; 神经影像; 儿科学; 立体定向脑电图

DOI：

10.7507/2096-0247.20210059

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

本回顾性横断面研究评估了将深度学习的难治性癫痫患儿的结构性磁共振成像（MRI）纳入到规划立体定向脑电图（SEEG）植入的可行性和潜在益处。本研究旨在评估自动病变检测与 SEEG 检测出癫痫发作起始区（SOZ）之间的共定位程度。将神经网络分类器应用于基于皮层 MRI 数据的三个队列：① 对 34 例局灶性皮质发育不良（FCD）患者的神经网络进行学习、训练和交叉验证；② 对 20 名健康儿童对照者进行特异性评估；③ 对 34 例患儿纳入 SEEG 植入计划的可行性进行了评价。SEEG 电极触点的坐标与分类器预测的病变进行核验。临床神经生理学家鉴定癫痫发作起源和易激惹区的 SEEG 电极触点位置。若 SOZ 坐标点和分类器预测的病变之间的距离<10 mm 则被认为是共定位的。影像学诊断病灶的分类敏感度为 74%（25/34）。对照组中未检测到异常（特异性=100%）。在 34 例 SEEG 植入患者中，21 例有局灶性皮层 SOZ，其中 8 例经病理证实为 FCD。分类器正确地检测了这 8 例 FCD 患者中的 7 例（86%）。组织病理学存在异质性的局灶性皮层病变患者中，62% 的患者分类器输出结果与 SOZ 之间存在共定位。3 例患者中，电临床提示为局灶性癫痫，SEEG 上无 SOZ 定位点，但在这些患者中，分类器识别了尚未植入的额外异常点。自动病变检测与 SEEG 之间的共定位存在高度的一致性。我们已经建立了一个框架，将基于深度学习的 MRI 自动病变检测纳入到 SEEG 植入计划。我们的发现支持了对自动 MRI 分析的前瞻性评估，以规划最佳电极植入轨迹方案。

Citation： WagstylKonrad, AdlerSophie, PimpelBirgit, 郭玉洁　译, 慕洁　审. 利用自动病变检测规划立体定向脑电图：可行性回顾性研究. Journal of Epilepsy, 2021, 7(4): 358-366. doi: 10.7507/2096-0247.20210059 Copy

1.	Kwan P, Schachter SC, Brodie MJ. Drug-resistant epilepsy. N Engl J Med, 2011, 365: 919-926.
2.	Wyllie E. Surgical treatment of epilepsy in children. Pediatr Neurol, 1998, 19(1): 179-188.
3.	Gonzalez-Martinez J, Mullin J, Bulacio J, et al. Stereoelectroencephalography in children and adolescents withdifficult-to-localize refractory focal epilepsy. Neurosurgery, 2014, 75(3): 258-268.
4.	McGovern RA, Knight EP, Gupta A, et al. Robot-assisted stereoelectroencephalographyin children. J Neurosurg Pediatr, 2018, 23: 288-296.
5.	Huppertz H-J, Grimm C, Fauser S, et al. Enhanced visualization ofblurred gray–white matter junctions in focal cortical dysplasia byvoxel-based 3D MRI analysis. Epilepsy Res, 2005, 67(1): 35-50.
6.	Wagner J, Weber B, Urbach H, et al. Morphometric MRI analysis improves detection of focal corticaldysplasia type II. Brain, 2011, 134: 2844-2854.
7.	Wang ZI, Jones SE, Jaisani Z, et al. Voxel-based morphometricmagnetic resonance imaging (MRI) postprocessing in MRInegativeepilepsies. Ann Neurol, 2015, 77: 1060-1075.
8.	Hong SJ, Kim H, Schrader D, Bernasconi N, Bernhardt BC, Bernasconi A. Automated detection of cortical dysplasia type II inMRI-negative epilepsy. Neurology, 2014, 83(1): 48-55.
9.	Adler S, Wagstyl K, Gunny R, et al. Novel surface features for automateddetection of focal cortical dysplasias in paediatric epilepsy. Neuroimage Clin, 2017, 14(1): 18-27.
10.	Hoyos-Osorio K, álvarez AM, Orozco áA, et al. Clustering-based undersampling to supportautomatic detection of focal cortical dysplasias. In: MendozaM, Velastín S, editors Progress in Pattern Recognition, ImageAnalysis, Computer Vision, and Applications. Cham, Switzerland: Springer, 2017: 298-305.
11.	Jin B, Krishnan B, Adler S, et al. Automated detection of focalcortical dysplasia type II with surface-based magnetic resonanceimaging postprocessing and machine learning. Epilepsia, 2018, 59(4): 982-992.
12.	Mo JJ, Zhang JG, Li WL, et al. Clinical value of machine learningin the automated detection of focal cortical dysplasia usingquantitative multimodal surface-based features. Front Neurosci, 2018, 12: 1008.
13.	Fischl B. FreeSurfer. Neuroimage, 2012, 62(3): 774-781.
14.	Greve DN, Van der Haegen L, Cai Q, et al. A surface-based analysisof language lateralization and cortical asymmetry. J Cogn Neurosci, 2013, 25(6): 1477-1492.
15.	Sharma JD, Seunarine KK, Tahir MZ, et al. Accuracy ofrobot-assisted versus optical frameless navigated stereoelectroencephalographyelectrode placement in children. J Neurosurg Pediatr, 2019, 23(2): 297-302.
16.	Jenkinson M, Bannister P, Brady M, et al. Improved optimizationfor the robust and accurate linear registration and motioncorrection of brain images. Neuroimage, 2002, 17(5): 825-841.
17.	Smith SM, Jenkinson M, Woolrich MW, et al. Advances in functionaland structural MR image analysis and implementation asFSL. Neuroimage, 2004, 23(Suppl): 208-219.
18.	Kikinis R, Pieper SD, Vosburgh KG. 3D slicer: a platform for subject-specific image analysis, visualization, and clinical support. In: Jolesz FA, ed. Intraoperative Imaging and Image-Guided Therapy. New York, NY: Springer, 2014: 277–289.
19.	Narizzano M, Arnulfo G, Ricci S, et al. SEEG assistant: a 3DSlicerextension to support epilepsy surgery. BMC Bioinformatics, 2017, 18(1): 124.
20.	Perucca P, Dubeau F, Gotman J. Intracranial electroencephalographicseizure-onset patterns: effect of underlying pathology. Brain, 2014, 137(1): 183-196.
21.	Lagarde S, Buzori S, Trebuchon A, et al. The repertoire of seizureonset patterns in human focal epilepsies: determinants and prognosticvalues. Epilepsia, 2019, 60(1): 85-95.
22.	David O, Blauwblomme T, Job AS, et al. Imaging the seizureonset zone with stereo-electroencephalography. Brain, 2011, 134(11): 2898-2911.
23.	Abraham A, Pedregosa F, Eickenberg M, et al. Machine learningfor neuroimaging with scikit-learn. Front Neuroinform, 2014, 8(1): 14.
24.	Allen M, Poggiali D, Whitaker K, et al. Raincloud plots: a multi-platform tool for robust data visualization. Wellcome Open Res, 2019, 4(1): 63.
25.	Adler S, Blackwood M, Northam GB, et al. Multimodal computationalneocortical anatomy in pediatric hippocampal sclerosis. AnnClin Transl Neurol, 2018, 5(5): 1200-1210.
26.	McGovern RA, Ruggieri P, Bulacio J, et al. Risk analysis of hemorrhage in stereo-electroencephalography procedures. Epilepsia, 2019, 60(3): 571-580.
27.	Adler S, Hong SJ, Liu M, et al. Topographic principles of corticalfluid-attenuated inversion recovery signal in temporal lobe epilepsy. Epilepsia, 2018, 5(5): 627-635.
28.	Galovic M, van Dooren VQH, Postma T, et al. Progressive corticalthinning in patients with focal epilepsy. JAMA Neurol, 2019, 76(10): 1230.
29.	Adler S, Whitaker K, Semmelroch M, et al. Multi-centre EpilepsyLesion Detection (MELD) project: conducting clinical research inan open-science framework. f1000research. 2018.
30.	Rosen BE. Ensemble learning using decorrelated neural networks. Conn Sci, 1996, 8: 373-384.
31.	Cucurull G, Wagstyl K, Casanova A, et al. Convolutional neuralnetworks for mesh-based parcellation of the cerebral cortex. 2018. [cited 2018 Jun 14]. Available from https://openr eview.net/pdf?id=rkKvB Aiiz.

1. Kwan P, Schachter SC, Brodie MJ. Drug-resistant epilepsy. N Engl J Med, 2011, 365: 919-926.
2. Wyllie E. Surgical treatment of epilepsy in children. Pediatr Neurol, 1998, 19(1): 179-188.
3. Gonzalez-Martinez J, Mullin J, Bulacio J, et al. Stereoelectroencephalography in children and adolescents withdifficult-to-localize refractory focal epilepsy. Neurosurgery, 2014, 75(3): 258-268.
4. McGovern RA, Knight EP, Gupta A, et al. Robot-assisted stereoelectroencephalographyin children. J Neurosurg Pediatr, 2018, 23: 288-296.
5. Huppertz H-J, Grimm C, Fauser S, et al. Enhanced visualization ofblurred gray–white matter junctions in focal cortical dysplasia byvoxel-based 3D MRI analysis. Epilepsy Res, 2005, 67(1): 35-50.
6. Wagner J, Weber B, Urbach H, et al. Morphometric MRI analysis improves detection of focal corticaldysplasia type II. Brain, 2011, 134: 2844-2854.
7. Wang ZI, Jones SE, Jaisani Z, et al. Voxel-based morphometricmagnetic resonance imaging (MRI) postprocessing in MRInegativeepilepsies. Ann Neurol, 2015, 77: 1060-1075.
8. Hong SJ, Kim H, Schrader D, Bernasconi N, Bernhardt BC, Bernasconi A. Automated detection of cortical dysplasia type II inMRI-negative epilepsy. Neurology, 2014, 83(1): 48-55.
9. Adler S, Wagstyl K, Gunny R, et al. Novel surface features for automateddetection of focal cortical dysplasias in paediatric epilepsy. Neuroimage Clin, 2017, 14(1): 18-27.
10. Hoyos-Osorio K, álvarez AM, Orozco áA, et al. Clustering-based undersampling to supportautomatic detection of focal cortical dysplasias. In: MendozaM, Velastín S, editors Progress in Pattern Recognition, ImageAnalysis, Computer Vision, and Applications. Cham, Switzerland: Springer, 2017: 298-305.
11. Jin B, Krishnan B, Adler S, et al. Automated detection of focalcortical dysplasia type II with surface-based magnetic resonanceimaging postprocessing and machine learning. Epilepsia, 2018, 59(4): 982-992.
12. Mo JJ, Zhang JG, Li WL, et al. Clinical value of machine learningin the automated detection of focal cortical dysplasia usingquantitative multimodal surface-based features. Front Neurosci, 2018, 12: 1008.
13. Fischl B. FreeSurfer. Neuroimage, 2012, 62(3): 774-781.
14. Greve DN, Van der Haegen L, Cai Q, et al. A surface-based analysisof language lateralization and cortical asymmetry. J Cogn Neurosci, 2013, 25(6): 1477-1492.
15. Sharma JD, Seunarine KK, Tahir MZ, et al. Accuracy ofrobot-assisted versus optical frameless navigated stereoelectroencephalographyelectrode placement in children. J Neurosurg Pediatr, 2019, 23(2): 297-302.
16. Jenkinson M, Bannister P, Brady M, et al. Improved optimizationfor the robust and accurate linear registration and motioncorrection of brain images. Neuroimage, 2002, 17(5): 825-841.
17. Smith SM, Jenkinson M, Woolrich MW, et al. Advances in functionaland structural MR image analysis and implementation asFSL. Neuroimage, 2004, 23(Suppl): 208-219.
18. Kikinis R, Pieper SD, Vosburgh KG. 3D slicer: a platform for subject-specific image analysis, visualization, and clinical support. In: Jolesz FA, ed. Intraoperative Imaging and Image-Guided Therapy. New York, NY: Springer, 2014: 277–289.
19. Narizzano M, Arnulfo G, Ricci S, et al. SEEG assistant: a 3DSlicerextension to support epilepsy surgery. BMC Bioinformatics, 2017, 18(1): 124.
20. Perucca P, Dubeau F, Gotman J. Intracranial electroencephalographicseizure-onset patterns: effect of underlying pathology. Brain, 2014, 137(1): 183-196.
21. Lagarde S, Buzori S, Trebuchon A, et al. The repertoire of seizureonset patterns in human focal epilepsies: determinants and prognosticvalues. Epilepsia, 2019, 60(1): 85-95.
22. David O, Blauwblomme T, Job AS, et al. Imaging the seizureonset zone with stereo-electroencephalography. Brain, 2011, 134(11): 2898-2911.
23. Abraham A, Pedregosa F, Eickenberg M, et al. Machine learningfor neuroimaging with scikit-learn. Front Neuroinform, 2014, 8(1): 14.
24. Allen M, Poggiali D, Whitaker K, et al. Raincloud plots: a multi-platform tool for robust data visualization. Wellcome Open Res, 2019, 4(1): 63.
25. Adler S, Blackwood M, Northam GB, et al. Multimodal computationalneocortical anatomy in pediatric hippocampal sclerosis. AnnClin Transl Neurol, 2018, 5(5): 1200-1210.
26. McGovern RA, Ruggieri P, Bulacio J, et al. Risk analysis of hemorrhage in stereo-electroencephalography procedures. Epilepsia, 2019, 60(3): 571-580.
27. Adler S, Hong SJ, Liu M, et al. Topographic principles of corticalfluid-attenuated inversion recovery signal in temporal lobe epilepsy. Epilepsia, 2018, 5(5): 627-635.
28. Galovic M, van Dooren VQH, Postma T, et al. Progressive corticalthinning in patients with focal epilepsy. JAMA Neurol, 2019, 76(10): 1230.
29. Adler S, Whitaker K, Semmelroch M, et al. Multi-centre EpilepsyLesion Detection (MELD) project: conducting clinical research inan open-science framework. f1000research. 2018.
30. Rosen BE. Ensemble learning using decorrelated neural networks. Conn Sci, 1996, 8: 373-384.
31. Cucurull G, Wagstyl K, Casanova A, et al. Convolutional neuralnetworks for mesh-based parcellation of the cerebral cortex. 2018. [cited 2018 Jun 14]. Available from https://openr eview.net/pdf?id=rkKvB Aiiz.

Previous Article
新医科+5G 背景下神经病学课堂教学模式新探索
Next Article
癫痫持续状态的脑脊液和血液生物标志物

Journal of Epilepsy

利用自动病变检测规划立体定向脑电图：可行性回顾性研究

Abstract Full text Figures/Tables Video References Cited by

Previous Article

Next Article

Format

Content