Codificador-decodificador con autoatención para la segmentación de materia gris en resonancia magnética cerebral
DOI:
https://doi.org/10.24054/rcta.v1i47.4282Palabras clave:
imágenes de resonancia magnética, segmentación de imágenes, visión artificial, aprendizaje profundoResumen
La resonancia magnetica cerebral es clave para el diagnostico asistido por computador, pero la segmentacion manual es costosa, lenta y dependiente del operador. Se presenta un pipeline para segmentar materia gris (SG) que combina un preprocesamiento minimo (armonizacion de forma, teselado por parches) con un codificador-decodificador 3D que integra autoatencion global y conexiones de salto multiescala. En el conjunto de prueba MRBrainS18, el modelo logra un Dice de 0.650 +/- 0.043, un IoU de 0.483 +/- 0.046, una precision de 0.700 +/- 0.046 y un recall de 0.610 +/- 0.059; las distribuciones por sujeto son compactas, lo que refleja consistencia entre casos. Las metricas de solapamiento muestran ausencia de relacion monotona con la distancia de Hausdorff (r ~ 0.08), lo que resalta la sensibilidad de los bordes aun con buen solapamiento global. El analisis de Bland-Altman evidencia un sesgo volumetrico negativo (RVE = -12.54 % +/- 9.53 %), consistente con una precision mayor que el recall y con predicciones conservadoras en las interfaces tisulares. Las lineas base clasicas (Otsu multiumbral y crecimiento de regiones), evaluadas en un subconjunto representativo, presentan menor rendimiento; el modelo mejora Dice e IoU en un 50 % frente a la mejor linea base.
Descargas
Referencias
L. Wang, T. Chitiboi, H. Meine, M. Günther, and H. K. Hahn, “Principles and methods for automatic and semi-automatic tissue segmentation in MRI data,” MAGMA, vol. 29, pp. 95–110, 2016, doi: 10.1007/s10334-015-0520-5.
Y. Gao, Y. Jiang, Y. Peng, F. Yuan, X. Zhang, and J. Wang, “Medical image segmentation: A comprehensive review of deep learning-based methods,” Tomography, vol. 11, Art. no. 52, 2025, doi: 10.3390/tomography11050052.
B. Kim, H. Kim, S. Kim, and Y. Hwang, “A brief review of non-invasive brain imaging technologies and the near-infrared optical bioimaging,” Applied Microscopy, vol. 51, Art. no. 9, 2021, doi: 10.1186/s42649-021-00058-7.
C. A. Nelson, “Incidental findings in magnetic resonance imaging (MRI) brain research,” J. Law Med. Ethics, vol. 36, pp. 315–213, 2008, doi: 10.1111/j.1748-720X.2008.00275.x.
Md. E. Rayed, S. M. S. Islam, S. I. Niha, J. R. Jim, M. M. Kabir, and M. F. Mridha, “Deep learning for medical image segmentation: State-of-the-art advancements and challenges,” Informatics in Medicine Unlocked, vol. 47, Art. no. 101504, 2024, doi: 10.1016/j.imu.2024.101504.
W. E. Kwok, “Radiofrequency interference in magnetic resonance imaging: Identification and rectification,” J. Clin. Imaging Sci., vol. 14, Art. no. 33, 2024, doi: 10.25259/JCIS_74_2024.
E. D. Angelini, T. Song, B. D. Mensh, and A. F. Laine, “Brain MRI segmentation with multiphase minimal partitioning: A comparative study,” Int. J. Biomed. Imaging, vol. 2007, Art. no. 10526, 2007, doi: 10.1155/2007/10526.
D. Feng, L. Tierney, and V. Magnotta, “MRI tissue classification using high-resolution Bayesian hidden Markov normal mixture models,” J. Amer. Statist. Assoc., vol. 107, pp. 102–119, 2012, doi: 10.1198/jasa.2011.ap09529.
P. Kaur, G. Singh, and P. Kaur, “A review of denoising medical images using machine learning approaches,” Curr. Med. Imaging Rev., vol. 14, pp. 675–685, 2018, doi: 10.2174/1573405613666170428154156.
A. Pereira Neto and F. J. B. Barros, “Noise reduction in brain magnetic resonance imaging using adaptive wavelet thresholding based on linear prediction factor,” Front. Neurosci., vol. 18, 2025, doi: 10.3389/fnins.2024.1516514.
I. Despotovic, B. Goossens, and W. Philips, “MRI segmentation of the human brain: Challenges, methods, and applications,” Computational and Mathematical Methods in Medicine, vol. 2015, Art. no. 450341, 2015, doi: 10.1155/2015/450341.
A. L. Alexander, S. A. Hurley, A. A. Samsonov, N. Adluru, A. P. Hosseinbor, P. Mossahebi, D. P. M. Tromp, E. Zakszewski, and A. S. Field, “Characterization of cerebral white matter properties using quantitative magnetic resonance imaging stains,” Brain Connect., vol. 1, pp. 423–446, 2011, doi: 10.1089/brain.2011.0071.
Md. E. Rayed, S. M. S. Islam, S. I. Niha, J. R. Jim, M. M. Kabir, and M. F. Mridha, “Deep learning for medical image segmentation: State-of-the-art advancements and challenges,” Informatics in Medicine Unlocked, vol. 47, Art. no. 101504, 2024, doi: 10.1016/j.imu.2024.101504.
V. Fonov, A. C. Evans, K. Botteron, C. R. Almli, R. C. McKinstry, D. L. Collins, and the Brain Development Cooperative Group, “Unbiased average age-appropriate atlases for pediatric studies,” NeuroImage, vol. 54, pp. 313–327, 2011, doi: 10.1016/j.neuroimage.2010.07.033.
V. Fonov, A. Evans, R. McKinstry, C. Almli, and D. Collins, “Unbiased nonlinear average age-appropriate brain templates from birth to adulthood,” NeuroImage, vol. 47, p. S102, 2009, doi: 10.1016/S1053-8119(09)70884-5.
D. L. Collins, A. P. Zijdenbos, W. F. C. Baaré, and A. C. Evans, “ANIMAL+INSECT: Improved cortical structure segmentation,” in Proc. 16th Int. Conf. Information Processing in Medical Imaging (IPMI), Berlin, Heidelberg, Germany: Springer, Jun. 1999, pp. 210–223.
N. R. Pal and S. K. Pal, “A review on image segmentation techniques,” Pattern Recognit., vol. 26, pp. 1277–1294, 1993, doi: 10.1016/0031-3203(93)90135-J.
M. K. Kar, M. K. Nath, and D. R. Neog, “A review on progress in semantic image segmentation and its application to medical images,” SN Comput. Sci., vol. 2, Art. no. 397, 2021, doi: 10.1007/s42979-021-00784-5.
R. Kadri, M. Tmar, and B. Bouaziz, “Brain image processing using deep learning: An overview,” in Digital Health in Focus of Predictive, Preventive and Personalised Medicine, L. Chaari, Ed. Cham, Switzerland: Springer, 2020, pp. 77–86.
G. Zhu, B. Jiang, L. Tong, Y. Xie, G. Zaharchuk, and M. Wintermark, “Applications of deep learning to neuro-imaging techniques,” Front. Neurol., vol. 10, 2019.
S. Lemm, B. Blankertz, T. Dickhaus, and K.-R. Müller, “Introduction to machine learning for brain imaging,” NeuroImage, vol. 56, pp. 387–399, 2011, doi: 10.1016/j.neuroimage.2010.11.004.
A.-I. Barranco-Gutiérrez, “Machine learning for brain images classification of two language speakers,” Comput. Intell. Neurosci., vol. 2020, Art. no. 9045456, 2020, doi: 10.1155/2020/9045456.
S. Wang and R. M. Summers, “Machine learning and radiology,” Med. Image Anal., vol. 16, pp. 933–951, 2012, doi: 10.1016/j.media.2012.02.005.
L. Wu, S. Wang, J. Liu, L. Hou, N. Li, F. Su, X. Yang, W. Lu, J. Qiu, M. Zhang, et al., “A survey of MRI-based brain tissue segmentation using deep learning,” Complex Intell. Syst., vol. 11, Art. no. 64, 2024, doi: 10.1007/s40747-024-01639-1.
Z. Guo, X. Li, H. Huang, N. Guo, and Q. Li, “Deep learning-based image segmentation on multimodal medical imaging,” IEEE Trans. Radiat. Plasma Med. Sci., vol. 3, pp. 162–169, 2019, doi: 10.1109/TRPMS.2018.2890359.
D. L. Pham, C. Xu, and J. L. Prince, “Current methods in medical image segmentation,” Annu. Rev. Biomed. Eng., vol. 2, pp. 315–337, 2000, doi: 10.1146/annurev.bioeng.2.1.315.
Md. E. Rayed, S. M. S. Islam, S. I. Niha, J. R. Jim, M. M. Kabir, and M. F. Mridha, “Deep learning for medical image segmentation: State-of-the-art advancements and challenges,” Informatics in Medicine Unlocked, vol. 47, Art. no. 101504, 2024, doi: 10.1016/j.imu.2024.101504.
N. Otsu, “A threshold selection method from gray-level histograms,” IEEE Trans. Syst., Man, Cybern., vol. 9, pp. 62–66, 1979, doi: 10.1109/TSMC.1979.4310076.
W. A. Hussein, S. Sahran, and S. N. H. S. Abdullah, “A fast scheme for multilevel thresholding based on a modified bees algorithm,” Knowl.-Based Syst., vol. 101, pp. 114–134, 2016, doi: 10.1016/j.knosys.2016.03.010.
D.-Y. Huang and C.-H. Wang, “Optimal multi-level thresholding using a two-stage Otsu optimization approach,” Pattern Recognit. Lett., vol. 30, pp. 275–284, 2009, doi: 10.1016/j.patrec.2008.10.003.
R. Panda, L. Samantaray, A. Das, S. Agrawal, and A. Abraham, “A novel evolutionary row class entropy based optimal multi-level thresholding technique for brain MR images,” Expert Syst. Appl., vol. 168, Art. no. 114426, 2021, doi: 10.1016/j.eswa.2020.114426.
S. Tan, L. Li, W. Choi, M. K. Kang, W. D. D’Souza, and W. Lu, “Adaptive region-growing with maximum curvature strategy for tumor segmentation in F-FDG PET,” Phys. Med. Biol., vol. 62, no. 13, pp. 5383–5402, 2017, doi: 10.1088/1361-6560/aa6e20.
M. Tamal, “A hybrid region growing tumour segmentation method for low contrast and high noise nuclear medicine images by combining a novel nonlinear diffusion filter and global gradient measure,” Heliyon, vol. 5, no. 11, Art. no. e02993, 2019, doi: 10.1016/j.heliyon.2019.e02993.
S. Bama, R. Velumani, N. B. Prakash, G. R. Hemalakshmi, and A. Mohanarathinam, “Automatic segmentation of melanoma using superpixel region growing technique,” Mater. Today Proc., vol. 45, pp. 1726–1732, 2021, doi: 10.1016/j.matpr.2020.08.618.
X. Zhao and X.-M. Zhao, “Deep learning of brain magnetic resonance images: A brief review,” Methods, vol. 192, pp. 131–140, 2021, doi: 10.1016/j.ymeth.2020.09.007.
Y. Fu, Y. Lei, T. Wang, W. J. Curran, T. Liu, and X. Yang, “A review of deep learning based methods for medical image multi-organ segmentation,” Physica Medica, vol. 85, pp. 107–122, 2021, doi: 10.1016/j.ejmp.2021.05.003.
R. Balakrishnan, M. del C. Valdés Hernández, and A. J. Farrall, “Automatic segmentation of white matter hyperintensities from brain magnetic resonance images in the era of deep learning and big data—A systematic review,” Comput. Med. Imaging Graph., vol. 88, Art. no. 101867, 2021, doi: 10.1016/j.compmedimag.2021.101867.
H. Izadkhah, “Medical image processing: An insight to convolutional neural networks,” in Deep Learning in Bioinformatics, H. Izadkhah, Ed. Academic Press, 2022, pp. 175–213.
Y. Xu et al., “Transformers in computational visual media: A survey,” Comput. Visual Media, vol. 8, no. 1, pp. 33–62, 2022, doi: 10.1007/s41095-021-0247-3.
H. J. Kuijf et al., “MR brain segmentation challenge 2018 data,” 2024.
O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional networks for biomedical image segmentation,” in Proc. MICCAI, Cham, Switzerland: Springer, 2015, pp. 234–241.
V. Dumoulin and F. Visin, “A guide to convolution arithmetic for deep learning,” arXiv:1603.07285, 2018.
H. Gao, H. Yuan, Z. Wang, and S. Ji, “Pixel transposed convolutional networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 42, no. 5, pp. 1218–1227, 2020, doi: 10.1109/TPAMI.2019.2893965.
V. Nair and G. E. Hinton, “Rectified linear units improve restricted Boltzmann machines,” in Proc. 27th Int. Conf. Mach. Learn. (ICML), Madison, WI, USA, 2010, pp. 807–814.
C. H. Sudre, W. Li, T. Vercauteren, S. Ourselin, and M. J. Cardoso, “Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations,” in Proc. MICCAI, 2017, pp. 240–248, doi: 10.1007/978-3-319-67558-9_28.
T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, “Focal loss for dense object detection,” arXiv:1708.02002, 2018.
W. R. Crum, O. Camara, and D. L. G. Hill, “Generalized overlap measures for evaluation and validation in medical image analysis,” IEEE Trans. Med. Imaging, vol. 25, no. 11, pp. 1451–1461, 2006, doi: 10.1109/TMI.2006.880587.
D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv:1412.6980, 2017.
K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification,” arXiv:1502.01852, 2015.
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2026 Camilo Andres Laiton Bonadiez, German Sanchez Torres, Carlos Nelson Henríquez Miranda
Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0.
