Aplicación de procesamiento digital a imágenes de baja resolución para mejorar la detección del cáncer de mama mediante redes neuronales

Patiño Callejas, Juan Sebastian

Aplicación de procesamiento digital a imágenes de baja resolución para mejorar la detección del cáncer de mama mediante redes neuronales

dc.contributor.advisor	Duitama Leal, Alejandro
dc.contributor.author	Patiño Callejas, Juan Sebastian
dc.date.accessioned	2024-09-06T20:27:49Z
dc.date.available	2024-09-06T20:27:49Z
dc.date.issued	2024-06
dc.description.abstract	El Carcinoma Ductal Invasivo (CDI) es una de las principales causas de morbilidad y mortalidad en mujeres, representando entre el 70 % y el 80 % de los casos de cáncer de mama. La detección inicial de este tipo de cáncer se realiza mediante mamografías. Cuando estas imágenes sugieren la presencia de una anomalía, se procede a un diagnóstico más preciso a través de biopsias. Los diagnósticos se basan en imágenes histológicas de alta resolución disponibles en centros especializados ubicados en grandes ciudades, lo que limita su acceso en regiones remotas. Además, su interpretación requiere la experiencia del radiólogo y patólogo, lo que puede resultar en una alta tasa de falsos positivos. Esto conlleva a exámenes adicionales que pueden ser invasivos, incrementando el estrés de los pacientes y los costos del sistema de salud. Para abordar esta limitación, se investigó la implementación de técnicas de procesamiento digital en imágenes histológicas de baja resolución utilizando redes neuronales para la detección del cáncer de mama. Se presenta un modelo que emplea imágenes de baja resolución (50x50 píxeles y 72 ppi) y redes neuronales convolucionales (CNN). Durante la investigación se exploraron diversas técnicas de procesamiento de imágenes basadas en color, bordes y umbrales.
dc.description.abstractenglish	Invasive Ductal Carcinoma (IDC) is one of the leading causes of morbidity and mortality in women, representing between 70% and 80% of breast cancer cases. Initial detection of this type of cancer is performed through mammograms. When these images suggest the presence of an anomaly, a more precise diagnosis is made through biopsies. Diagnoses are based on high-resolution histological images available in specialized centers located in large cities, limiting their access in remote regions. Additionally, their interpretation requires the expertise of radiologists and pathologists, which can result in a high rate of false positives. This leads to additional examinations that may be invasive, increasing patient stress and healthcare costs. To address this limitation, the implementation of digital processing techniques in low-resolution histological images using neural networks was investigated for breast cancer detection. A model employing low-resolution images (50x50 pixels and 72 ppi) and convolutional neural networks (CNN) is presented. Various image processing techniques based on color, edges, and thresholds were explored during the research.
dc.identifier.uri	https://hdl.handle.net/20.500.12495/12969
dc.language.iso	es
dc.relation.references	H. Sung, J. Ferlay, R. L. Siegel et al., «Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries, » CA: A Cancer Journal for Clinicians, vol. 71, págs. 209-249, 2020.
dc.relation.references	Y. Xu, M. Gong, Y. Wang, Y. Yang, S. Liu y Q. Zeng, «Global trends and forecasts of breast cancer incidence and deaths,» Scientific Data, vol. 10, n.o 1, pág. 334, 2023. DOI: 10.1038/s41597- 023- 02253- 5. dirección: https://doi.org/10.1038/s41597-023-02253-5.
dc.relation.references	S. B. Edge, A. C. S. American Joint Committee on Cancer et al., AJCC cancer staging handbook: from the AJCC cancer staging manual. Springer, 2010, vol. 19.
dc.relation.references	T.-Q. H. Ho, M. C. Bissell, K. Kerlikowske et al., «Cumulative probability of false-positive results after 10 years of screening with digital breast tomosynthesis vs digital mammography,» JAMA Network Open, vol. 5, n.o 3, e222440-e222440, 2022.
dc.relation.references	J. Chaki, «Cancer Data Pre-Processing Techniques,» en Current Applications of Deep Learning in Cancer Diagnostics, 1st Edition. CRC Press, 2023, pág. 15, ISBN: 9781003277002
dc.relation.references	L. Tsochatzidis, L. Costaridou e I. Pratikakis, «Deep Learning for Breast Cancer Diagnosis from Mammograms— A Comparative Study,» Journal of Imaging, vol. 5, pág. 37, 2019. DOI: 10.3390/jimaging5030037. dirección: https://doi.org/10.3390/jimaging5030037.
dc.relation.references	M. M. Srikantamurthy, V. P. S. Rallabandi, D. B. Dudekula et al., «Classification of benign and malignant subtypes of breast cancer histopathology imaging using hybrid CNN-LSTM based transfer learning,» BMC Medical Imaging, vol. 23, pág. 19, 2023. DOI: 10.1186/ s12880- 023- 00964- 0. dirección: https : / / doi . org/10. 1186/s12880-023-00964-0.
dc.relation.references	S. Karimi Jafarbigloo y H. Danyali, «Nuclear atypia grading in breast cancer histopathological images based on CNN feature extraction and LSTM classification,» CAAI Transactions on Intelligence Technology, vol. 6, n.o 4, págs. 426-439, 2021.
dc.relation.references	N. Antropova, B. Huynh, H. Li y M. L. Giger, «Breast lesion classification based on dynamic contrastenhanced magnetic resonance images sequences with long short-term memory networks,» Journal of Medical Imaging, vol. 6, n.o 1, págs. 011 002-011 002, 2019.
dc.relation.references	L. Liu, W. Feng, C. Chen et al., «Classification of breast cancer histology images using MSMV-PFENet,» Scientific Reports, vol. 12, pág. 17 447, 2022. DOI: 10. 1038/s41598-022-22358- y. dirección: https://doi.org/ 10.1038/s41598-022-22358-y.
dc.relation.references	R. Yan, F. Ren, Z. Wang et al., «A hybrid convolutional and recurrent deep neural network for breast cancer pathological image classification,» en 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), IEEE, 2018, págs. 957-962
dc.relation.references	J. Arevalo, F. A. González, R. Ramos-Pollán, J. L. Oliveira y M. A. Guevara Lopez, «Representation learning for mammography mass lesion classification with convolutional neural networks,» Computer Methods and Programs in Biomedicine, vol. 127, págs. 248-257, 2016, ISSN: 0169-2607. DOI: https : / / doi . org / 10 . 1016 / j . cmpb . 2015 . 12 . 014. dirección: https : / / www . sciencedirect . com / science / article / pii / S0169260715300110.
dc.relation.references	G. Carneiro, J. Nascimento y A. P. Bradley, «Unregistered multiview mammogram analysis with pre-trained deep learning models,» en International conference on medical image computing and computer-assisted intervention, Springer, 2015, págs. 652-660.
dc.relation.references	Z. Jiao, X. Gao, Y. Wang y J. Li, «A deep feature based framework for breast masses classification,» Neurocomputing, vol. 197, págs. 221-231, 2016.
dc.relation.references	R. Sanyal, M. Jethanandani y R. Sarkar, «DAN : Breast Cancer Classification from High-Resolution Histology Images Using Deep Attention Network,» en Innovations in Computational Intelligence and Computer Vision, M. K. Sharma, V. S. Dhaka, T. Perumal, N. Dey y J. M. R. S. Tavares, eds., Singapore: Springer Singapore, 2021, págs. 319-326.
dc.relation.references	E. Rakha, M. Toss y C. Quinn, «Specific cell differentiation in breast cancer: a basis for histological classification,» Journal of Clinical Pathology, vol. 75, n.o 2, págs. 76-84, 2022, ISSN: 0021-9746. DOI: 10 . 1136/jclinpath- 2021- 207487. eprint: https://jcp.bmj. com/content/75/2/76.full.pdf. dirección: https://jcp.bmj. com/content/75/2/76.
dc.relation.references	P. A. Kenny, G.-Y. Lee, C. A. Myers et al., «The morphologies of breast cancer cell lines in threedimensional assays correlate with their profiles of gene expression,» Molecular Oncology, vol. 1, n.o 1, págs. 84-96, 2007. DOI: 10.1016/j.molonc.2007.02.004. dirección: https://doi.org/10.1016/j.molonc.2007.02. 004.
dc.relation.references	G. Botti, M. Cantile, F. Collina et al., «Morphological and pathological features of basal-like breast cancer, » Translational Cancer Research, vol. 8, n.o Suppl 5, 2019, ISSN: 2219-6803. dirección: https : / / tcr . amegroups.org/article/view/30356.
dc.relation.references	E. Alizadeh, J. Castle, A. Quirk, C. D. Taylor, W. Xu y A. Prasad, «Cellular morphological features are predictive markers of cancer cell state,» Computers in Biology and Medicine, vol. 126, pág. 104 044, 2020, ISSN: 0010-4825. DOI: https://doi.org/10.1016/j.compbiomed. 2020 . 104044. dirección: https : / /www. sciencedirect . com/science/article/pii/S0010482520303759.
dc.relation.references	S. Syed, J. Schober, A. Blanco y S. P. Zustiak, «Morphological adaptations in breast cancer cells as a function of prolonged passaging on compliant substrates, » PLOS ONE, vol. 12, n.o 11, e0187853, 2017. DOI: 10.1371/journal.pone.0187853. dirección: https: //doi.org/10.1371/journal.pone.0187853.
dc.relation.references	Y. Shen, N. Wu, J. Phang et al., «An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization,» Medical Image Analysis, vol. 68, pág. 101 908, 2021, ISSN: 1361-8415. DOI: https : / / doi . org / 10 . 1016 / j . media . 2020 . 101908. dirección: https : / /www. sciencedirect . com/science/article/pii/S1361841520302723.
dc.relation.references	I. Aifuwa, B. C. Kim, P. Kamat et al., «Senescent stroma induces nuclear deformations in cancer cells via the inhibition of RhoA/ROCK/myosin II-based cytoskeletal tension,» PNAS Nexus, vol. 2, n.o 1, pgac270, dic. de 2022.
dc.relation.references	S. Tang, C. Jing, Y. Jiang et al., «The effect of image resolution on convolutional neural networks in breast ultrasound,» Heliyon, vol. 9, n.o 8, e19253, 2023, ISSN: 2405-8440. DOI: https : / / doi . org / 10 . 1016 / j . heliyon . 2023 . e19253. dirección: https : / /www. sciencedirect . com/science/article/pii/S2405844023064617.
dc.relation.references	D. Murcia-Gómez, I. Rojas-Valenzuela y O. Valenzuela, «Impact of Image Preprocessing Methods and Deep Learning Models for Classifying Histopathological Breast Cancer Images,» Applied Sciences, vol. 12, n.o 22, 2022, ISSN: 2076-3417. dirección: https://www. mdpi.com/2076-3417/12/22/11375.
dc.relation.references	R. C. Gonzalez y R. E. Woods, Digital Image Processing. 2008.
dc.relation.references	A. Ranjbaran, A. H. A. Hassan, M. Jafarpour y B. Ranjbaran, «A Laplacian based image filtering using switching noise detector,» SpringerPlus, vol. 4, pág. 119, 2015. DOI: 10.1186/ s40064 - 015- 0846- 5. dirección: https://doi.org/10.1186/s40064-015-0846-5.
dc.relation.references	M. Saii y A. Mia, «Lung Detection and Segmentation Using Marker Watershed and Laplacian Filtering,» International Journal of Biomedical Engineering and Clinical Science, vol. 1, n.o 2, págs. 29-42, 2015. DOI: 10.11648/j.ijbecs.20150102.12. dirección: https://doi. org/10.11648/j.ijbecs.20150102.12.
dc.relation.references	S. Chen, L. Yao y B. Chen, «A parameterized logarithmic image processing method with Laplacian of Gaussian filtering for lung nodule enhancement in chest radiographs,» Medical & Biological Engineering & Computing, vol. 54, págs. 1793-1806, 2016. DOI: 10. 1007/s11517-016-1469-x. dirección: https://doi.org/10. 1007/s11517-016-1469-x.
dc.relation.references	M. Padlia y J. Sharma, «Fractional Sobel Filter Based Brain Tumor Detection and Segmentation Using Statistical Features and SVM,» en Nanoelectronics, Circuits and Communication Systems, ép. Lecture Notes in Electrical Engineering, V. Nath y J. Mandal, eds., vol. 511, Springer, Singapore, 2019, págs. 155-164. DOI: 10.1007/978-981-13-0776-8_15. dirección: https: //doi.org/10.1007/978-981-13-0776-8_15
dc.relation.references	D. Sharma y G. Jindal, «Identifying lung cancer using image processing techniques,» en International Conference on Computational Techniques and Artificial Intelligence (ICCTAI), Citeseer, vol. 17, 2011, págs. 872-880.
dc.relation.references	M. Sonka, V. Hlavac y R. Boyle, Image Processing, Analysis, and Machine Vision. 2014
dc.relation.references	N. Bnouni, H. B. Amor, I. Rekik, M. S. Rhim, B. Solaiman y N. E. B. Amara, «Boosting CNN Learning by Ensemble Image Preprocessing Methods for Cervical Cancer Segmentation,» en 2021 18th International Multi-Conference on Systems, Signals Devices (SSD), 2021, págs. 264-269.
dc.relation.references	W. Setiawan, M. M. Suhadi, Y. D. Pramudita et al., «Histopathology of lung cancer classification using convolutional neural network with gamma correction,» Commun. Math. Biol. Neurosci., vol. 2022, Article-ID, 2022.
dc.relation.references	D. Okuboyejo y O. O. Olugbara, «Segmentation of melanocytic lesion images using gamma correction with clustering of keypoint descriptors,» Diagnostics, vol. 11, n.o 8, pág. 1366, 2021.
dc.relation.references	S. M. Pizer et al., Adaptive Histogram Equalization and Its Variations. 1987.
dc.relation.references	N. Kharel, A. Alsadoon, P. Prasad y A. Elchouemi, «Early diagnosis of breast cancer using contrast limited adaptive histogram equalization (CLAHE) and Morphology methods,» en 2017 8th International Conference on Information and Communication Systems (ICICS), IEEE, 2017, págs. 120-124.
dc.relation.references	A. Vidyarthi, J. Shad, S. Sharma y P. Agarwal, «Classification of breast microscopic imaging using hybrid CLAHE-CNN deep architecture,» en 2019 Twelfth International Conference on Contemporary Computing (IC3), IEEE, 2019, págs. 1-5.
dc.relation.references	N. Otsu, «A Threshold Selection Method from Gray- Level Histograms,» 1979.
dc.relation.references	A. Renjith, P. Manjula y P. Mohan Kumar, «Brain tumour classification and abnormality detection using neuro-fuzzy technique and Otsu thresholding,» Journal of medical engineering & technology, vol. 39, n.o 8, págs. 498-507, 2015.
dc.relation.references	K. Mittal, A. Shekhar, P. Singh y M. Kumar, «Brain tumour extraction using otsu based threshold segmentation, » International Journal of Advanced Research in Computer Science and Software Engineering, vol. 7, n.o 4, 2017.
dc.relation.references	A. S. Chanchlani, V. M. Thakare y V. M. Wadhai, «Performance Analysis of Image Thresholding Methods on Brain Cancer MRI for Segmentation.,» Grenze International Journal of Engineering & Technology (GIJET), vol. 7, n.o 1, 2021.
dc.relation.references	R. Widodo, T. Badriyah, I. Syarif y W. Sandhika, «SEGMENTATION OF LUNG CANCER IMAGE BASED ON CYTOLOGIC EXAMINATION USING THRESHOLDING METHOD,» Jurnal Ilmiah Kursor, vol. 12, n.o 1, págs. 41-48, 2023
dc.relation.references	A. Cruz-Roa, A. Basavanhally, F. González et al., «Automatic detection of invasive ductal carcinoma in whole slide images with convolutional neural networks,» en Medical Imaging 2014: Digital Pathology, SPIE, vol. 9041, 2014, pág. 904 103
dc.rights	Attribution-NoDerivatives 4.0 Internacional
dc.rights.accessrights	info:eu-repo/semantics/openAccess
dc.rights.accessrights	http://purl.org/coar/access_right/c_abf2
dc.rights.local	Acceso abierto
dc.rights.uri	http://creativecommons.org/licenses/by-nd/4.0/
dc.subject	Cancer de mama
dc.subject	Preprocesamiento
dc.subject	Imagen baja calidad
dc.subject.keywords	Breast cancer
dc.subject.keywords	Preprocessing
dc.subject.keywords	Low quality image
dc.title	Aplicación de procesamiento digital a imágenes de baja resolución para mejorar la detección del cáncer de mama mediante redes neuronales
dc.title.translated	Application of digital processing to low-resolution images for improved breast cancer detection using neural networks