Self-supervised learning for histopathological image analysis using limited annotated data

dc.contributor.advisorGonzález Osorio, Fabio Augusto
dc.contributor.advisorCruz Roa, Angel Alfonso
dc.contributor.authorBetancourth Castro, Landneyker
dc.contributor.cvlacBetancourth Castro, Landneyker [0001826131]
dc.contributor.researchgroupMindlab
dc.coverage.countryColombia
dc.date.accessioned2025-09-10T21:57:06Z
dc.date.available2025-09-10T21:57:06Z
dc.date.issued2025
dc.descriptionilustraciones (principalmente a color), diagramas, fotografíasspa
dc.description.abstractThe analysis of digital histopathology slides plays a crucial role in cancer research and diagnosis, including prostate cancer. However, acquiring sufficient annotated data to train deep learning models in this domain is challenging due to the scarcity of pathologists, the expertise required for precise annotations, and the time-consuming nature of the process. This study proposes a self-supervised learning approach based on SimCLR (Simple Contrastive Learning of Representations) for histopathological image analysis, enabling the learning of visual representations from unannotated data. These representations are then used to train a supervised classifier with a small amount of labeled data, facilitating accurate and generalizable prostate cancer grading. The dataset used in this study is the Prostate cANcer graDe Assessment (PANDA) dataset Bulten2022, which contains histopathological images of prostate tissue samples along with expert annotations for cancer grade assessment. We evaluated various scenarios and configurations by varying the amount of annotated data and using self-supervised representation learning on unannotated datasets from either a general domain (natural images) or a specific domain (histopathology). In our proposed SimCLR-based approach, we demonstrate the effectiveness of self-supervised techniques in learning meaningful representations from unannotated data. The method leverages inherent structures and patterns in histopathological images to learn rich representations, which can later be fine-tuned on a small annotated dataset for specific downstream tasks. The proposed SimCLR-based framework was evaluated on the task of prostate cancer grade assessment using a limited number of annotated samples. Experimental results show that the self-supervised model is capable of generalizing to unseen data and achieving competitive performance compared to supervised approaches trained on larger annotated datasets. Notably, better performance was observed when the model was pre-trained directly on domain-specific histopathological images, reaching scores of 0.96 for stroma, 0.76 for healthy tissue, and 0.73, 0.71, 0.42 for Gleason 3, 4, 5, respectively with only 15% of the annotated data. In contrast, when using a model pre-trained on natural images (STL-10), slightly lower scores were obtained: 0.9, 0.58, 0.56, 0.52, and 0.3, respectively. Despite this difference, the model trained on natural images still showed remarkable performance, especially considering it was trained with a reduced fraction of labeled data, highlighting its potential for use in resource-constrained scenarios. (Texto tomado de la fuente)eng
dc.description.abstractEl análisis de imágenes digitales de láminas de histopatología desempeña un papel crucial en la investigación y el diagnóstico del cáncer, incluido el cáncer de próstata. Sin embargo, adquirir suficientes datos anotados para entrenar modelos de aprendizaje profundo en este dominio suele ser desafiante debido a la escasez de patólogos, la experiencia requerida de ellos para la anotación precisa y al consumo de tiempo del proceso. Este estudio propone un enfoque de aprendizaje auto-supervisado con SimCLR (Simple Contrastive Learning of Representations) para análisis histopatológico, permitiendo aprender representaciones visuales a partir de datos no anotados. Estas representaciones se utilizan luego para entrenar un clasificador supervisado con pocos datos etiquetados, facilitando la gradación del cáncer de próstata con alta generalización. En este trabajo se utilizó el conjunto de datos para valoración del grado del cáncer de próstata (Prostate cANcer graDe Assessment - PANDA) [1], que comprende imágenes histopatológicas de muestras de tejido prostático junto con anotaciones de expertos correspondientes para la evaluación del grado de cáncer. Se evaluó diferentes escenarios y configuraciones variando la cantidad de datos anotados, el aprendizaje no supervisado de la representación de las imágenes de conjuntos de datos no anotados de un dominio general (imágenes naturales) o específico (histopatología). En nuestro enfoque propuesto basado en SimCLR, se demuestra la efectividad de las técnicas de auto-supervisión en aprender representaciones significativas a partir de datos no anotados. El enfoque aprovecha las estructuras inherentes y los patrones presentes en las imágenes histopatológicas para aprender representaciones ricas, las cuales luego pueden ser ajustadas (fine-tuned) en un pequeño conjunto de datos anotados para tareas específicas posteriores. En el enfoque propuesto basado en SimCLR se evaluó en la tarea de valoración del grado de cáncer de próstata utilizando un número limitado de muestras anotadas. Los resultados experimentales muestran la capacidad del modelo auto-supervisado para generalizar datos no vistos y lograr un rendimiento competitivo en comparación con enfoques supervisados entrenados en conjuntos de datos anotados más grandes. En particular, se observó un mejor desempeño cuando el modelo fue preentrenado directamente con imágenes histopatológicas, alcanzando con solo el 15 % de datos notados precisiones de 0.87 en tejido estromal, 0.79 en tejido sano y 0.66 en Gleason 5. En contraste, al partir de un modelo preentrenado con imágenes naturales (STL-10), los resultados fueron ligeramente inferiores en esas mismas clases: 0.84, 0.75 y 0.64 respectivamente. A pesar de la diferencia, el enfoque con imágenes naturales también mostró un desempeño destacado, especialmente considerando que se entrenó con una fracción reducida de datos anotados, lo cual evidencia su potencial en escenarios con recursos limitados. (Texto tomado de la fuente)spa
dc.description.degreelevelMaestría
dc.description.degreenameMagíster en Ingeniería de Sistemas y Computaciónspa
dc.description.researchareaSistemas Inteligentes
dc.description.researchareaIntelligent Systems
dc.format.extentx, 56 páginas
dc.format.mimetypeapplication/pdf
dc.identifier.instnameUniversidad Nacional de Colombiaspa
dc.identifier.reponameRepositorio Institucional Universidad Nacional de Colombiaspa
dc.identifier.repourlhttps://repositorio.unal.edu.co/spa
dc.identifier.urihttps://repositorio.unal.edu.co/handle/unal/88707
dc.language.isoeng
dc.publisherUniversidad Nacional de Colombia
dc.publisher.branchUniversidad Nacional de Colombia - Sede Bogotá
dc.publisher.facultyFacultad de Ingeniería
dc.publisher.placeBogotá, Colombia
dc.publisher.programBogotá - Ingeniería - Maestría en Ingeniería - Ingeniería de Sistemas y Computación
dc.relation.referencesW. Bulten, K. Kartasalo, P. C. Chen, P. Ström, H. Pinckaers, K. Nagpal, Y. Cai, D. F. Steiner, H. van Boven, R. Vink, C. Hulsbergen-van de Kaa, J. van der Laak, M. B. Amin, A. J. Evans, T. van der Kwast, R. Allan, P. A. Humphrey, H. Grönberg, H. Samaratunga, B. Delahunt, T. Tsuzuki, T. Häkkinen, L. Egevad, M. Demkin, S. Dane, F. Tan, M. Valkonen, G. S. Corrado, L. Peng, C. H. Mermel, P. Ruusuvuori, G. Litjens, and M. Eklund. Artificial intelligence for diagnosis and gleason grading of prostate cancer: the panda challenge. Nature Medicine, 28, 2021.
dc.relation.referencesAltaklinik. Gleason score in prostate cancer. https://www.altaklinik.com/ prostate/prostate-cancer/gleason-score/, 2025. Accessed: 2025-07-21.
dc.relation.referencesA. Chaudhary. Understanding simclr: A simple framework for contrastive learning. https://amitness.com/posts/simclr/, 2020. Accessed: 2025-04-09.
dc.relation.referencesP. Y. Wicaksana. Self-supervised pre-training with simclr, March 26 2023.
dc.relation.referencesJ. R. Stark, S. Perner, M. J. Stampfer, J. A. Sinnott, S. Finn, A. S. Eisenstein, J. Ma, M. Fiorentino, T. Kurth, M. Loda, M. L. Gonzalez, C. A. Andriole, E. A. Platz, P. W. Kantoff, and L. A. Mucci. Gleason score and lethal prostate cancer: Does 3 + 4 = 4 + 3? Journal of Clinical Oncology, 27(21):3459–3464, 2009.
dc.relation.referencesK. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016.
dc.relation.referencesR. L. Siegel, K. D. Miller, and A. Jemal. Cancer statistics, 2020. CA: A Cancer Journal for Clinicians, 70(1):7–30, 2020.
dc.relation.referencesWorld Cancer Research Fund. Prostate cancer statistics. https://www.wcrf.org/, 2022.
dc.relation.referencesAmerican Cancer Society. Key statistics for prostate cancer. https://www.cancer. org/, 2024.
dc.relation.referencesColombia fact sheet – globocan 2022. https://gco.iarc.fr/today/data/ factsheets/populations/170-colombia-fact-sheets.pdf, 2022. Accessed: 2025- 04-08.
dc.relation.referencesJ. I. Epstein, L. Egevad, M. B. Amin, B. Delahunt, J. R. Srigley, and P. A. Humphrey. The 2014 international society of urological pathology (isup) consensus conference on gleason grading of prostatic carcinoma: Definition of grading patterns and proposal for a new grading system. Am J Surg Pathol, 40(2):244–252, 2016.
dc.relation.referencesB. Williams, H. Lee, and D. M. Grzybicki. The global shortage of pathologists: the power of telepathology. Journal of Pathology Informatics, 9:1–4, 2018.
dc.relation.referencesJ. A. Mejía, A. Torres, and P. Rodríguez. Pathologist workforce in latin america: A survey-based study of training and distribution. Diagnostic Pathology, 15(1):1–8, 2020.
dc.relation.referencesD. C. Castro, I. Walker, and B. Glocker. Causality matters in medical imaging. arXiv, 2019.
dc.relation.referencesT. Chen, S. Kornblith, M. Norouzi, and G. Hinton. A simple framework for contrastive learning of visual representations. arXiv preprint arXiv:2002.05709, 2020.
dc.relation.referencesT. A. Ozkan, A. T. Eruyar, O. O. Cebeci, O. Memik, L. Ozcan, and I. Kuskonmaz. Interobserver variability in gleason histological grading of prostate cancer: a multicenter study. Scandinavian Journal of Urology, 50(6):420–424, 2016.
dc.relation.referencesWilliam C. Allsbrook Jr, Kent A. Mangold, Matthew H. Johnson, Ronald B. Lane, Charles G. Lane, and Jonathan I. Epstein. Interobserver and intraobserver reproduci- bility of gleason grading of prostatic carcinoma: general pathologist. Human Pathology, 32(1):81–88, 2001.
dc.relation.referencesC. Galindo, J. Fonseca, and A. Rojas. Workforce characteristics and geographical distribution of pathologists in colombia: A nationwide study. Revista Colombiana de Patología, 2024. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC11400992/.
dc.relation.referencesA. Rojas, L. Pérez, and E. Valdez. The latin american pathologist workforce: Distri- bution, access, and training challenges. The Lancet Regional Health – Americas, 2023. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11025004/.
dc.relation.referencesY. LeCun, Y. Bengio, and G. Hinton. Deep learning. Nature, 521(7553):436–444, 2015.
dc.relation.referencesG. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Setio, F. Ciompi, M. Ghafoorian, J. A. W. M. van der Laak, B. van Ginneken, and C. I. Sánchez. A survey on deep learning in medical image analysis. Medical Image Analysis, 42:60–88, 2017.
dc.relation.referencesD. Komura and S. Ishikawa. Machine learning methods for histopathological image analysis. Computational and Structural Biotechnology Journal, 16:34–42, 2018.
dc.relation.referencesW. Bulten, H. Pinckaers, H. van Boven, R. Vink, T. de Bel, B. van Ginneken, J. van der Laak, C. Hulsbergen-van de Kaa, and G. Litjens. Automated deep-learning system for gleason grading of prostate cancer using biopsies: a diagnostic study. The Lancet Oncology, 21:233–241, 2020.
dc.relation.referencesM. Y. Lu, D. F. K. Williamson, T. Y. Chen, R. J. Chen, M. Barbieri, and F. Mahmood. Data-efficient and weakly supervised computational pathology on whole-slide images. Nature Biomedical Engineering, 5(6):555–570, 2021.
dc.relation.referencesK. Nagpal, D. Foote, Y. Liu, P. C. Chen, E. Wulczyn, F. Tan, T. G. Olsen, N. Olson, J. L. Smith, A. Mohtashamian, J. H. Wren, E. Wong, G. S. Corrado, P. Q. Nelson, J. D. Hipp, D. F. Steiner, C. H. Mermel, and M. C. Stumpe. Development and validation of a deep learning algorithm for improving gleason scoring of prostate cancer. npj Digital Medicine, 2(1):48, 2019.
dc.relation.referencesL. Betancourth, F. González, and A. Cruz. Self-supervised learning for histopathological image analysis using limited annotated data. Presented at the II Simposio de Patología Digital e Inteligencia Artificial (REDPAT), 2024. Universidad de los Llanos, August 30, 2024.
dc.relation.referencesJ. D. Bancroft and C. Layton. Theory and Practice of Histological Techniques. Elsevier, 8th edition, 2019.
dc.relation.referencesM. K. K. Niazi, A. V. Parwani, and M. N. Gurcan. Digital pathology and artificial intelligence. The Lancet Oncology, 20(5):e253–e261, 2019.
dc.relation.referencesP. Rawla. Epidemiology of prostate cancer. World Journal of Oncology, 10(2):63–89, 2019.
dc.relation.referencesH. B. Carter, P. C. Albertsen, M. J. Barry, A. M. Etzioni, R. M. Freedland, M. S. Kirkby, P. R. Carroll, M. J. Roobol, D. F. Penson, C. M. Chu, J. W. Davis, R. S. Sandhu, J. M. McNaughton-Collins, and P. H. Carroll. Genetic predisposition to prostate cancer. The Lancet Oncology, 23(4):542–550, 2022.
dc.relation.referencesP. A. Humphrey. Histological variants of prostatic carcinoma and their significance. Modern Pathology, 32(1):16–34, 2019.
dc.relation.referencesInternational Society of Urological Pathology. Consensus on gleason grading of prostatic carcinoma. Modern Pathology, 27(9):1221–1230, 2014.
dc.relation.referencesG. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Setio, F. Ciompi, M. Ghafoorian, J. A. W. M. van der Laak, B. van Ginneken, and C. I. Sánchez. Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis. Scientific Reports, 6:26286, 2016.
dc.relation.referencesM. Balkenhol, D. Tellez, C. P. H. Vreuls, J. A. W. M. van der Laak, and F. Ciompi. Deep learning assisted mitotic counting for breast cancer. Laboratory Investigation, 99:1596–1606, 2019.
dc.relation.referencesA. Janowczyk and A. Madabhushi. Deep learning for digital pathology image analysis: A comprehensive tutorial with selected use cases. Journal of Pathology Informatics, 7:29, 2016.
dc.relation.referencesG. Campanella, M. G. Hanna, L. Geneslaw, A. Miraflor, V. Werneck Krauss Silva, K. J. Busam, E. Brogi, V. E. Reuter, D. S. Klimstra, and T. J. Fuchs. Clinical-grade computational pathology using weakly supervised deep learning on whole-slide images. Nature Medicine, 25:1301–1309, 2019.
dc.relation.referencesH. Pinckaers, L. van Huis, J. van der Laak, and G. Litjens. Detection of prostate cancer in whole-slide images through end-to-end training with image-level labels. IEEE Transactions on Medical Imaging, 40:1817–1826, 2021.
dc.relation.referencesL. Egevad, B. Delahunt, J. R. Srigley, and D. M. Berney. Standardization of gleason grading among 337 european pathologists. Histopathology, 62:247–256, 2013.
dc.relation.referencesP. Ström, K. Kartasalo, H. Olsson, M. Smedmark, L. Egevad, F. Strand, M. Lindskog, K. Lövgren, F. Westergren, M. Hägglund, J. H. N. Palm, A. Andrén Olin, P. J. Helminen, and M. Eklund. Artificial intelligence for diagnosis and grading of prostate cancer in biopsies: a population-based, diagnostic study. The Lancet Oncology, 21:222–232, 2020.
dc.relation.referencesJ. van der Laak, G. Litjens, and F. Ciompi. Deep learning in histopathology: the path to the clinic. Nature Medicine, 27:775–784, 2021.
dc.relation.referencesK. Bera, K. A. Schalper, D. L. Rimm, V. Velcheti, and A. Madabhushi. Artificial intelligence in digital pathology—new tools for diagnosis and precision oncology. Nature Reviews Clinical Oncology, 16:703–715, 2019.
dc.relation.referencesF. Doshi-Velez and B. Kim. Towards a rigorous science of interpretable machine learning. arXiv preprint, arXiv:1702.08608, 2017.
dc.relation.referencesS. Belharbi, A. Hervieu, N. Betrouni, M. Cordonnier, H. Delingette, C. González, C. De- valland, É. Stindel, N. Albarghach, H. Benoît-Cattin, and N. Ayache. Deep interpretable classification and weakly-supervised segmentation of histology images via max-min un- certainty. IEEE Transactions on Medical Imaging, 41:702–714, 2021.
dc.relation.referencesC. Chen, O. Li, X. Tao, A. Barnett, J. Su, and C. Rudin. This looks like that: Deep learning for interpretable image recognition. Advances in Neural Information Processing Systems, 32, 2019.
dc.relation.referencesS. Shen, S. Han, D. R. Aberle, A. A. T. Bui, and W. J. Hsu. An interpretable deep hierarchical semantic convolutional neural network for lung nodule malignancy classification. Expert Systems with Applications, 128:84–95, 2019.
dc.relation.referencesN. Rieke, J. Hancox, W. Li, F. Milletari, H. R. Roth, S. Albarqouni, S. Bakas, M. N. Galtier, B. Landman, K. H. Maier-Hein, S. Ourselin, M. Cardoso, A. Criminisi, B. A. Patiño, D. A. Oyarzun, R. A. Schultz, S. K. Warfield, D. S. Marcus, D. Xu, A. Shetty, A. K. Sinha, C. R. McGuinness, A. S. Rajpurkar, B. A. Bender, and D. S. Kermany. Federated learning for medical imaging: a review of the state-of-the-art. Medical Image Analysis, 64:101742, 2020.
dc.relation.referencesM. Ochi, D. Komura, and S. Ishikawa. Pathology foundation models. AI in Medicine, 2024.
dc.relation.referencesJ. Ma, Z. Guo, F. Zhou, Y. Zhang, S. Gu, S. Xu, C. Xu, J. Tan, Y. Gong, C. Liu, Y. Li, H. Wang, S. Xu, W. Liu, J. Xie, L. Zhang, Z. Zheng, Y. Yu, J. Fan, S. Yu, L. Wang, L. Zhang, J. Li, H. Huang, H. Yang, Y. Chen, X. Chen, T. Xu, D. Zeng, Y. Wang, S. Yan, J. Xie, X. Yu, Y. Zhu, and J. Feng. Towards a generalizable pathology foundation model via unified knowledge distillation. arXiv preprint, arXiv:2401.04567, 2024.
dc.relation.referencesP. Meseguer, R. del Amor, A. Colomer, R. González-García, A. O. Fernández, A. Hervella, J. Muñoz-Barrutia, J. M. Salas-Gonzalez, A. Ortega, A. Turiel, R. Molina, A. R. Martín, M. A. Gómez, R. Vicente, D. M. J. Tax, M. C. López, and A. Garcia-Garcia. Foundation models for slide-level cancer subtyping in digital pathology. arXiv preprint, 2024. arXiv:2403.08086.
dc.relation.referencesM. Alber, S. Tietz, J. Dippel, T. Homeyer, D. Rueckert, F. Springer, T. Zerbe, J. Lotz, J. H. Moltz, F. J. Theis, A. Madabhushi, A. Walch, F. Klauschen, J. Neumann, L. Hou, and M. J. Schlegel. A novel pathology foundation model by mayo clinic, charité, and aignostics. arXiv preprint, 2025. arXiv:2403.20045.
dc.relation.referencesJ. Guo, S. Lu, C. Xu, Y. Zhang, J. Yu, X. Hu, H. Li, Q. Zhou, J. Xie, J. Zhang, H. Wang, C. Liu, X. Chen, D. Zhang, Q. Feng, L. Yang, and Y. Lu. Assessment of cell nuclei ai foundation models in kidney pathology. AI in Pathology, 2024.
dc.relation.referencesG. Campanella, S. C. Chen, R. Verma, T. A. Alshalalfa, J. D. Reuter, T. Sedarsky, R. C. Young, B. Kim, A. Miraflor, M. G. Hanna, and T. J. Fuchs. A clinical benchmark of public self-supervised pathology foundation models. Nature Biomedical Engineering, 2024.
dc.relation.referencesS. Hua, F. Yan, T. Shen, Z. Zhang, B. Wang, C. Lin, H. Li, Z. Yu, J. Zhao, X. Zhu, Y. Song, Y. Wang, D. Zeng, W. Liu, and Y. Zhu. Pathoduet: Foundation models for pathological slide analysis of h&e and ihc stains. arXiv preprint, 2023. arXiv:2311.16450.
dc.relation.referencesW. Bulten, H. Pinckaers, H. van Boven, R. Vink, T. de Bel, B. van Ginneken, J. van der Laak, C. Hulsbergen-van de Kaa, and G. Litjens. Artificial intelligence assistance significantly improves gleason grading of prostate biopsies by pathologists. Modern Pathology, 34(3):490–501, 2021.
dc.relation.referencesE. Arvaniti, N. Fricker, M. Moret, N. Rupp, T. Hermanns, C. Fankhauser, N. Wey, P. J. Wild, J. H. Rüschoff, and M. Claassen. Automated gleason grading of prostate cancer tissue microarrays via deep learning. The Lancet Digital Health, 1(3):e102–e111, 2019.
dc.relation.referencesS. Toledo-Cortés, D. H. Useche, H. Müller, and F. A. González. Grading prostate cancer diagnostic images with deep quantum ordinal regression. Computers in Biology and Medicine, 145:105472, 2022.
dc.relation.referencesR. Chen, L. Zhao, J. Wang, Y. Liu, F. Li, H. Xu, X. Zhou, T. Zhang, and Y. Zhang. Self-supervised learning for patch-based prostate cancer diagnosis. Nature Biomedical Engineering, 7:34–48, 2023.
dc.relation.referencesS. Medina. Wisdom: Weakly-supervised interpretable prostate cancer grading on wsis. Master’s thesis, Universidad Nacional de Colombia, 2024.
dc.relation.referencesA. Obeid, S. Boumaraf, A. Sohail, T. Hassan, S. Javed, J. Dias, M. Bennamoun, and N. Werghi. Advancing histopathology with deep learning under data scarcity: A decade in review. arXiv preprint arXiv:2410.19820, 2023.
dc.relation.referencesA. Obeid, S. Boumaraf, A. Sohail, T. Hassan, S. Javed, J. Dias, M. Bennamoun, and N. Werghi. A survey on deep learning tools dealing with data scarcity: definitions, challenges, solutions, tips, and applications. Journal of Big Data, 10(1):1–50, 2023.
dc.relation.referencesS. Shurrab and R. Duwairi. Self-supervised learning methods and applications in medical imaging analysis: A survey. Expert Systems with Applications, 165:113765, 2021.
dc.relation.referencesK. Kartasalo, W. Bulten, B. Delahunt, L. Egevad, H. Grönberg, J. Oksanen, P. Ruusuvuori, H. Samaratunga, P. Ström, and M. Eklund. Artificial intelligence for diagnosis and gleason grading of prostate cancer in biopsies—current status and next steps. European Urology Focus, 7(4):687–691, 2021.
dc.relation.referencesK. He, H. Fan, Y. Wu, S. Xie, and R. Girshick. Momentum contrast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9729–9738, 2020.
dc.relation.referencesJ. B. Grill, F. Strub, F. Altché, C. Tallec, P. H. Richemond, E. Buchatskaya, C. Doersch, B. A. Pires, Z. D. Guo, M. G. Azar, B. Piot, K. Kavukcuoglu, R. Munos, and M. Valko. Bootstrap your own latent: A new approach to self-supervised learning. In Advances in Neural Information Processing Systems (NeurIPS), volume 33, pages 21271–21284, 2020.
dc.relation.referencesA. Coates and A. Y. Ng. An analysis of single-layer networks in unsupervised feature learning. In Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, pages 215–223, 2011.
dc.relation.referencesQ. Yao, M. Wang, Y. Liu, X. Li, Z. Tang, Z. Li, and J. Huang. Self-supervised learning for large-scale histopathological image classification with contrastive predictive coding. arXiv preprint arXiv:2011.13971, 2020.
dc.relation.referencesO. Ciga, T. Xu, and A. L. Martel. Self-supervised contrastive learning for digital histopathology. arXiv preprint arXiv:2011.13971, 2020.
dc.relation.referencesA. Krizhevsky. Learning multiple layers of features from tiny images. Technical report, University of Toronto, 2009.
dc.relation.referencesG. Litjens, H. Pinckaers, K. Kartasalo, M. Maggie, M. Eklund, P. Ruusuvuori, P. Ström, S. Dane, and W. Bulten. Prostate cancer grade assessment (panda) challenge. https: //kaggle.com/competitions/prostate-cancer-grade-assessment, 2020. Kaggle.
dc.relation.referencesL. Zhang, C. Wang, H. Yu, M. Yang, S. Song, J. Zhang, Y. Zheng, J. Li, Y. Wang, D. Zhao, Q. Liang, X. Wu, Y. Gao, Z. Xie, H. Zhang, and Y. Qiao. Uni: Unlocking the potential of self-supervised learning for histopathology with over 100 million image patches. Nature Medicine, 2024.
dc.relation.referencesR. Chen, A. Arnaout, D. J. Steiner, S. Ghaffari Laleh, Y. Wang, C. Ni, Y. Liu, T. D. C. Nguyen, R. T. Chen, Y. Gao, C. Liu, A. Ankit, T. J. Fuchs, M. G. Hanna, and A. Yala. Conch: Contrastive learning from captions for histopathology. Communications Medicine, 2024.
dc.relation.referencesS. Stein, A. Khanna, T. Sedarsky, P. Natarajan, C. A. Gutierrez, K. Yeung, K. Deng, A. Mahmood, M. G. Hanna, and T. J. Fuchs. Virchow: Foundation model for clinicalgrade computational pathology. Nature Medicine, 2024.
dc.relation.referencesGoogle Research. Path foundation model for histopathology. https://huggingface. co/google/path-foundation, 2024.
dc.rights.accessrightsinfo:eu-repo/semantics/openAccess
dc.rights.licenseAtribución-NoComercial 4.0 Internacional
dc.rights.urihttp://creativecommons.org/licenses/by-nc/4.0/
dc.subject.bneHistología patológicaspa
dc.subject.bneHistology, pathologicaleng
dc.subject.ddc000 - Ciencias de la computación, información y obras generales::003 - Sistemas
dc.subject.ddc000 - Ciencias de la computación, información y obras generales::006 - Métodos especiales de computación
dc.subject.ddc610 - Medicina y salud::616 - Enfermedades
dc.subject.decsProcesamiento de imagen asistido por computadorspa
dc.subject.decsImage processing, computer-assistedeng
dc.subject.decsInterpretación de imagen asistida por computadorspa
dc.subject.decsImage interpretation, computer-assistedeng
dc.subject.decsNeoplasias de la próstata -- Diagnóstico por imagenspa
dc.subject.decsProstatic neoplasms -- Diagnostic imagingeng
dc.subject.decsClasificación del tumor -- Métodosspa
dc.subject.decsNeoplasm grading -- Methodseng
dc.subject.decsTécnicas histológicasspa
dc.subject.decsHistological techniqueseng
dc.subject.decsNeoplasias de la próstata -- Clasificaciónspa
dc.subject.decsProstatic neoplasms -- Classificationeng
dc.subject.otherAprendizaje supervisado (Aprendizaje automático)spa
dc.subject.otherSupervised learning (Machine learning)eng
dc.subject.otherAprendizaje automático (Inteligencia artificial)spa
dc.subject.otherMachine learning (Artificial intelligence)eng
dc.subject.proposalHistopathologyeng
dc.subject.proposalProstate cancereng
dc.subject.proposalCancer gradingeng
dc.subject.proposalDeep learningeng
dc.subject.proposalSelf-supervised learningeng
dc.subject.proposalHistopatologíaspa
dc.subject.proposalCáncer de próstataspa
dc.subject.proposalGradación de cáncerspa
dc.subject.proposalAprendizaje automáticospa
dc.subject.proposalAprendizaje auto-supervisadospa
dc.subject.wikidataAprendizaje autosupervisadospa
dc.subject.wikidataSelf-supervised learningeng
dc.titleSelf-supervised learning for histopathological image analysis using limited annotated dataeng
dc.title.translatedModelo de aprendizaje auto-supervisado para el análisis de imágenes histopatológicas a partir de pocos datos anotadosspa
dc.typeTrabajo de grado - Maestría
dc.type.coarhttp://purl.org/coar/resource_type/c_bdcc
dc.type.coarversionhttp://purl.org/coar/version/c_ab4af688f83e57aa
dc.type.contentText
dc.type.driverinfo:eu-repo/semantics/masterThesis
dc.type.redcolhttp://purl.org/redcol/resource_type/TM
dc.type.versioninfo:eu-repo/semantics/acceptedVersion
dcterms.audience.professionaldevelopmentEstudiantes
dcterms.audience.professionaldevelopmentInvestigadores
dcterms.audience.professionaldevelopmentPúblico general
oaire.accessrightshttp://purl.org/coar/access_right/c_abf2

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
Landneyker_Betancourth.pdf
Tamaño:
5.78 MB
Formato:
Adobe Portable Document Format
Descripción:
Tesis de Maestría en Ingeniería de Sistemas y Computación

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
5.74 KB
Formato:
Item-specific license agreed upon to submission
Descripción: