IMPLEMENTACIÓN DE UN ALGORITMO DE SEGMENTACIÓN DE ACCIDENTES CEREBROVASCULARES EN IMÁGENES DE RESONANCIA MAGNÉTICA UTILIZANDO REDES CONVOLUCIONALES (IMPLEMENTATION OF A STROKE SEGMENTATION ALGORITHM IN MAGNETIC RESONANCE IMAGING USING CONVOLUTIONAL NETWORKS)
Resumen
Un accidente cerebrovascular es una emergencia médica que puede provocar complicaciones graves si no se detecta a tiempo. De modo que, si se diagnóstica a con anticipación aumenta la posibilidad de que el paciente pueda sobrevivir. La detección de las lesiones por accidente cerebrovascular isquémico mediante imágenes de resonancia magnética (RM) es fundamental para un diagnóstico preciso y decisivo. En este artículo, se propone desarrollar un algoritmo mediante técnicas de procesamiento de imágenes y una arquitectura basada en U-Net para realizar la segmentación automática de accidentes cerebrovasculares a partir de IRM que permitan identificar la zona de la lesión. Se utilizo la base de datos ISLES 2015, dos optimizadores (SGD y AdamW) y tres funciones de pérdida (Dice, Focal, GDFL) para evaluación del método propuesto obteniendo mejores resultados con AdamW y GDFL con un valor de intersección sobre unión de 0.791 que compite con métodos del estado del arte.
Palabras clave:Accidente cerebrovascular, Generalized dice loss, Imágenes multimodales, Segmentación de imágenes de resonancia magnética, U-Net.
Abstract
A stroke is a medical emergency that can lead to serious complications if not detected early. Thus, early diagnosis increases the patient's chance of survival. Detection of ischemic stroke lesions by magnetic resonance imaging (MRI) is essential for accurate and decisive diagnosis. In this article, we propose to develop an algorithm using image processing techniques and a U-Net based architecture to perform stroke segmentation from MRI to identify the lesion area. The ISLES 2015 database, two optimizers (SGD and AdamW) and three loss functions (Dice, Focal, GDFL) were used to evaluate the performance of the proposed method obtaining better results with AdamW and GDFL with an intersection over union value of 0.791 that compete with State-of-the-art methods.
Keywords: Generalized dice loss, Magnetic resonance image segmentation, Multimodal imaging, Stroke, U-Net.
Texto completo:
389-408 PDFReferencias
Allen, M., Poggiali, D., Whitaker, K., Marshall, TR., van Langen, J., & Kievit, RA. Raincloud plots: a multi-platform tool for robust data visualization. Wellcome Open Research. 2019; 4:63. DOI: 10.12688/wellcomeopenres.15191.2. PMID: 31069261; PMCID: PMC6480976.
Castellino, R. A. Computer aided detection (CAD): an overview. Cancer Imaging, 2005 Aug 23; 5(1):17-9. doi: 10.1102/1470-7330.2005.0018. PMID: 16154813; PMCID: PMC1665219.
Che-Wei, L., and Jeen-Shing, W. A digital circuit design of hyperbolic tangent sigmoid function for neural networks, 2008 IEEE International Symposium on Circuits and Systems (ISCAS), Seattle, WA, USA, 2008, pp. 856-859, doi: 10.1109/ISCAS.2008.4541553.
Choi, D., Shallue, C. J., Nado, Z., Lee, J., Maddison, C. J., & Dahl, G. E., (2019). On Empirical Comparisons of Optimizers for Deep Learning. 1. http://arxiv.org/abs/1910.05446.
Gobierno de México, (2015). Accidentes Cerebrovasculares. Recuperado de https://n9.cl/ergng.
Gobierno de México, (2022). En 2021, ictus o enfermedad vascular cerebral ocasionó más de 37 mil decesos en México. Recuperado de https://n9.cl/r8z64.
Hema-Rajini, N., & Bhavani, R., (2013). Computer aided detection of ischemic stroke using segmentation and texture features. Measurement: Journal of the International Measurement Confederation, 46(6), 1865–1874. https://doi.org/10.1016/j.measurement.2013.01.010.
IMSS, (2015). Enfermedad Vascular Cerebral. IMSS. Recuperado de https://n9.cl/mlwn3.
K. Revanth, T. D. Varsha shree, N. S. Madhava Raja and V. Rajinikanth, Computational Investigation of Stroke Lesion Segmentation from Flair/DW Modality MRI, 2018 Fourth International Conference on Biosignals, Images and Instrumentation (ICBSII), Chennai, India, 2018, pp. 206-212, doi: 10.1109/ICBSII.2018.8524617.
Karthik, R., Gupta, U., Jha, A., Rajalakshmi, R., & Menaka, R., (2019). A deep supervised approach for ischemic lesion segmentation from multimodal MRI using Fully Convolutional Network. Applied Soft Computing Journal, 84(105685). https://doi.org/10.1016/j.asoc.2019.105685.
Le Bihan, D., & Lima, M., (2015). Diffusion magnetic resonance imaging: What water tells us about biological tissues. PLoS Biology, 13(7), e1002203. https://doi.org/10.1371/journal.pbio.1002203.
Liu, L., Kurgan, L., Wu, F. X., & Wang, J., (2020). Attention convolutional neural network for accurate segmentation and quantification of lesions in ischemic stroke disease. Medical Image Analysis, 65(101791), 101791. https://doi.org/10.1016/j.media.2020.101791.
M. Beauchemin, K.P.B. Thomson & G. Edwards, (1998) On the Hausdorff Distance Used for the Evaluation of Segmentation Results, Canadian Journal of Remote Sensing, 24:1, 3-8, DOI: 10.1080/07038992.1998.10874685.
Maier, O., Menze, B. H., von der Gablentz, J., Häni, L., Heinrich, M. P., Liebrand, M., Winzeck, S., Basit, A., Bentley, P., Chen, L., Christiaens, D., Dutil, F., Egger, K., Feng, C., Glocker, B., Götz, M., Haeck, T., Halme, H.-L., Havaei, M., Reyes, M., (2017). ISLES 2015 - A public evaluation benchmark for ischemic stroke lesion segmentation from multispectral MRI. Medical Image Analysis, 35, 250–269. https://doi.org/10.1016/j.media.2016.07.009.
MemorialCare, (2023). Stroke. MemorialCare. Recuperado de https://www.memorialcare.org/services/stroke-care/stroke.
Müller, D., Soto-Rey, I. & Kramer, F. Towards a guideline for evaluation metrics in medical image segmentation. BMC Res Notes 15, 210, 2022. https://doi.org/10.1186/s13104-022-06096-y.
Oktay, O., Schlemper, J., Folgoc, L. L., Lee, M., Heinrich, M., Misawa, K., Mori, K., McDonagh, S., Hammerla, N. Y., Kainz, B., Glocker, B., & Rueckert, D., (2018). Attention U-Net: Learning where to look for the pancreas. In arXiv [cs.CV]. http://arxiv.org/abs/1804.03999.
OMS, (2020). Las 10 principales causas de defunción. OMS. recuperado de https://n9.cl/07m39.
Perkins, C. J., Kahya, E., Roque, C. T., Roche, P. E., & Newman, G. C., (2001). Fluid-attenuated inversion recovery and Diffusion- and perfusion-weighted MRI abnormalities in 117 consecutive patients with stroke symptoms. Stroke; a Journal of Cerebral Circulation, 32(12), 2774–2781. https://doi.org/10.1161/hs1201.099634.
Ronneberger, O., Fischer, P. & Brox, T., (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. In N. Navab, J. Hornegger, W. M. Wells & A. F. Frangi (eds.), Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III (pp. 234--241). Springer International Publishing. ISBN: 978-3-319-24574-4.
Sriramakrishnan, P., Kalaiselvi, T., Thirumalaiselvi, M., Padmapriya, S. T., & Ramkumar, S., (2019). A Role of Medical Imaging Techniques in Human Brain Tumor Treatment. International Journal of Recent Technology and Engineering, 8(4S2), 565–568. https://n9.cl/glu2f.
Subudhi, A., Jena, S. & Sabut, S. Delineation of the ischemic stroke lesion based on watershed and relative fuzzy connectedness in brain MRI. Med Biol Eng Comput 56, 795–807, 2018. https://n9.cl/0zgdp.
Yeung, M., Sala, E., Schönlieb, C. B., & Rundo, L., (2022). Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation. Computerized Medical Imaging and Graphics: The Official Journal of the Computerized Medical Imaging Society, 95(102026), 102026. https://doi.org/10.1016/j.compmedimag.2021.102026.
URL de la licencia: https://creativecommons.org/licenses/by/3.0/deed.es
Pistas Educativas está bajo la Licencia Creative Commons Atribución 3.0 No portada. 
TECNOLÓGICO NACIONAL DE MÉXICO / INSTITUTO TECNOLÓGICO DE CELAYA
Antonio García Cubas Pte #600 esq. Av. Tecnológico, Celaya, Gto. México
Tel. 461 61 17575 Ext 5450 y 5146
pistaseducativas@itcelaya.edu.mx