Por favor, use este identificador para citar o enlazar este ítem: https://repositorio.uti.edu.ec//handle/123456789/3429
Registro completo de metadatos
Campo DC Valor Lengua/Idioma
dc.contributor.authorZapata, Lorena-
dc.contributor.authorChalco, Lorena-
dc.contributor.authorAguilar, Lenin-
dc.contributor.authorPimbosa, Esmeralda-
dc.contributor.authorRamírez-Morales, Iván-
dc.contributor.authorHidalgo, Jairo-
dc.contributor.authorYandún, Marco-
dc.contributor.authorArias-Flores, Hugo-
dc.contributor.authorGuevara-Maldonado, César-
dc.date.accessioned2022-06-30T14:52:13Z-
dc.date.available2022-06-30T14:52:13Z-
dc.date.issued2020-
dc.identifier.urihttps://link.springer.com/chapter/10.1007/978-3-030-20454-9_8-
dc.identifier.urihttp://repositorio.uti.edu.ec//handle/123456789/3429-
dc.description.abstractCytological diagnosis is useful in the practical context compared to the histopathology, since it can classify pathologies among the cutaneous masses, the samples can be collected easily without anesthetizing the patient, at very low cost. However, an experimented veterinarian performs the cytological diagnosis in approximately 25 min. Artificial intelligence is being used for the diagnosis of many pathologies in human medicine, the experience gained by years of work in the area of work allow to issue correct diagnoses, this experience can be trained in an intelligent system. In this work, we collected a total of 1500 original cytologic images, performed some preliminary tests and also propose a deep learning based approach for image analysis and classification using convolutional neural networks (CNN). To adjust the parameters of the classification model, we recommend to perform a random and grid search will be applied, modifying the batch size of images for training, the number of layers, the learning speed and the selection of three optimizers: Adadelta, RMSProp and SGD. The performance of the classifiers will be evaluated by measuring the accuracy and two loss functions: cross-categorical entropy and mean square error. These metrics will be evaluated in a set of images different from those with which the model was trained (test set). By applying this model, an image classifier can be generated that efficiently identifies a cytology diagnostic in a short time and with an optimal detection rate. This is the first approach for the development of a more complex model of skin mass detection in all its types. © 2020, Springer Nature Switzerland AG.es
dc.language.isoenges
dc.publisherAdvances in Intelligent Systems and Computing. Volume 965, Pages 83 - 91. AHFE International Conference on Human Factors in Artificial Intelligence and Social Computing, the AHFE International Conference on Human Factors, Software, Service and Systems Engineering, and the AHFE International Conference of Human Factors in Energy, 2019. Washington D.C.24 . July 2019 through 28 July 2019es
dc.rightsopenAccesses
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/es
dc.titleDetection of Cutaneous Tumors in Dogs Using Deep Learning Techniqueses
dc.typearticlees
Aparece en las colecciones: Artículos Científicos Indexados

Ficheros en este ítem:
No hay ficheros asociados a este ítem.


Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons Creative Commons