Abstract The suffix -ístico has an atypical form in Spanish because it has three syllables and is proparoxytone. Furthermore, its segmentation can be ambiguous, since it can be described as the result of a root + -ístico, or as a base ending in -ista + -’ico. According to these characteristics of -ístico, this suffix fails to follow the principles established by the preference theory of natural morphology, specifically two of them: the optimal shape of units and the morphotactic transparency. Because of this, -ístico seems unnatural from this approach. In this paper, I suggest an analysis of -ístico as a result of a combination of a base ending in -ista + -’ico, following the concept of morphological solidarity, which I will define in this paper also (Beniers 1996). These affixes are on a solidary relationship, from which it follows that it is unnecessary to observe a term derived with -ista in the process of using -’ico on it.

Resumen El sufijo -ístico es anómalo en español en su forma porque tiene tres sílabas y es esdrújulo. Además, su segmentación puede ser ambigua, ya que se puede describir como resultado de una raíz + -ístico, o como una base terminada en -ista + -´ico. De acuerdo con sus características, este sufijo incumple los principios de la teoría de las preferencias de la morfología natural, particularmente dos: la forma óptima de la unidad y la transparencia morfotáctica. Por lo anterior, el sufijo -ístico resulta antinatural según este enfoque. En este artículo, propongo un análisis de -ístico como el resultado de la combinación de una base terminada en -ista + -ico, siguiendo la noción de solidaridad morfológica que también explico en este trabajo (Beniers 1996). Estos afijos se encuentran en una relación solidaria, de la que se sigue que no es necesario observar una palabra derivada con -ista en el proceso para emplear el sufijo -’ico sobre ella.

Chronic sinusitis is a common disease that significantly affects quality of life. To treat chronic sinusitis, functional endoscopic sinus surgery (FESS) is frequently considered. FESS alleviates chronic sinusitis symptoms by restoring natural sinus drainage. Otolaryngologists rely on computed tomography (CT) reports to establish whether FESS is appropriate. To enhance sinus CT reports and improve decision-making, a segmentation model is developed. Both 2D and 3D segmentation models were compared, with the 3D model achieving marginally better results. The model accurately segments the sinus system, including the nasal cavity, achieving a mean Dice coefficient of 0.889 ± 0.028. The resulting 3D visualisation of the segmented sinus system enables quick identification of opacified regions, helping otolaryngologists to make informed decisions about the appropriateness of FESS. This automated approach reduces the time required to compile reports, improves the precision of clinical evaluations, and ultimately enhances patient care

Chroniese sinusitis is 'n algemene siekte wat lewenskwaliteit aansienlik beïnvloed. Om chroniese sinusitis te behandel, word funksionele endoskopiese sinuschirurgie (FESS) gereeld oorweeg. FESS verlig chroniese sinusitis simptome deur die herstel van natuurlike sinus dreinering. Otolaryngoloë maak staat op rekenaartomografie (CT) verslae om vas te stel of FESS toepaslik is. Om sinus-CT-verslae te verbeter en besluitneming te verbeter, word 'n segmenteringsmodel ontwikkel. Beide 2D- en 3D-segmenteringsmodelle is vergelyk, met die 3D-model wat marginaal beter resultate behaal het. Die model segmenteer die sinusstelsel akkuraat, insluitend die neusholte, en bereik 'n gemiddelde Dobbelsteenkoëffisiënt van 0.889 ± 0.028. Die gevolglike 3D-visualisering van die gesegmenteerde sinusstelsel maak vinnige identifikasie van ondeursigtige streke moontlik, wat otolaryngoloë help om ingeligte besluite oor die toepaslikheid van FESS te neem. Hierdie outomatiese benadering verminder die tyd wat nodig is om verslae saam te stel, verbeter die akkuraatheid van kliniese evaluasies en verbeter uiteindelik pasiëntsorg

Resumen: Esta investigación aborda la diversidad de rasgos en el mercado minorista mexicano, buscando identificar atributos comunes fundamentales entre los consumidores de tiendas de retail. El objetivo es analizar las variaciones en las preferencias de distintos tipos de consumidores con respecto a las estrategias de precios implementadas por un supermercado y cómo estas contribuyen a una experiencia transformadora. Se destacan cinco segmentos de consumidores prósperos: i) myopic switcher light, ii) strategic switchers light, iii) myopic switchers heavy, iv) strategic loyal heavy, y v) myopic loyal light. La aplicación del enfoque de segmentación permite a los tomadores de decisiones discernir estrategias más aceptadas en cada segmento, facilitando la comprensión de cómo abordar efectivamente las características distintivas de los consumidores. Este enfoque informado proporciona un marco valioso para especialistas y tomadores de decisiones de marketing, permitiéndoles adaptar estrategias específicas según las preferencias diferenciadas de los consumidores en el mercado minorista mexicano. La investigación destaca la viabilidad de adoptar estrategias de precios basadas en la segmentación de mercado en distintas regiones, ofreciendo hitos prácticos para eficientizar la implementación de tácticas de precios en el contexto retail en México.

Abstract: The present research addresses the diversity of traits in the Mexican retail market, aiming to find fundamental common attributes among consumers in retail stores. Aim: To examine differences in preferences among various consumer segments regarding pricing strategies implemented by a supermarket and their role in fostering a transformative experience. Five prosperous consumer segments are highlighted: i) myopic switcher light, ii) strategic switchers light, iii) myopic switchers heavy, iv) strategic loyal heavy, and v) myopic loyal light. The application of the segmentation approach allows decision-makers to discern more accepted strategies in each segment, easing an understanding of how to effectively address the distinctive characteristics of consumers. This informed approach supplies a valuable framework for specialists and marketing decision-makers, enabling them to adapt specific strategies grounded regarding the diverse preferences exhibited by consumers in the Mexican retail sector. The research highlights the feasibility of adopting market segmentation-based pricing strategies in various regions, providing practical milestones to streamline the implementation of pricing tactics in the Mexican retail context.

Autonomous electric vehicles use camera sensors for vision-based steering control and detecting both roads and objects. In this study, road and object detection are combined, utilizing the YOLOv8x-seg model trained for 200 epochs, achieving the lowest segmentation loss at 0.53182. Simulation tests demonstrate accurate road and object detection, effective object distance measurement, and real-time road identification for steering control, successfully keeping the vehicle on track with an average object distance measurement error of 2.245 m. Route planning for autonomous vehicles is crucial, and the A-Star algorithm is employed to find the optimal route. In real-time tests, when an obstacle is placed between nodes 6 and 7, the A-Star algorithm can reroute from the original path (5, 6, 7, 27, and 28) to a new path (5, 6, 9, 27, and 28). This study demonstrates the vital role of sensor fusion in autonomous vehicles by integrating various sensors. This study focuses on sensor fusion for object-road detection and path planning using the A* algorithm. Real-time tests in two different scenarios demonstrate the successful integration of sensor fusion, enabling the vehicle to follow planned routes. However, some route nodes remain unreachable, requiring occasional driver intervention. These results demonstrate the feasibility of sensor fusion with diverse tasks in third-level autonomous vehicles.

Resumen Este estudio tiene como objetivo determinar si el consumo de noticias, la confianza en los medios, el interés por las noticias y la afinidad política están correlacionados con el segmento generacional del mercado, el cual se mide por la edad de la audiencia digital. Además, se busca verificar si la edad está asociada con la atención que las audiencias digitales prestan a los periodistas y a las marcas de noticias. La investigación se basa en una encuesta realizada por YouGov a 2005 personas en México en el año 2022 por encargo del Instituto Reuters para el Estudio del Periodismo de la Universidad de Oxford como parte de su Informe de Noticias Digitales anual. Para esto, el método estadístico utilizado fue la prueba de independencia chi-cuadrado entre variables cualitativas. Los resultados demostraron que la edad se correlaciona con los hábitos de consumo, la confianza en los medios, el interés por las noticias y la afinidad política de las audiencias digitales en México. Por otra parte, la edad se relaciona con el hecho de que las audiencias digitales se centren en los periodistas o en las marcas de noticias en América Latina. Este estudio profundiza en las características de las audiencias digitales de México y hace un aporte desde una perspectiva de economía, gestión y marketing de medios. También ayuda a las organizaciones de medios a evaluar la estrategia de mercado, la segmentación del mercado y los mercados objetivo en función de las características de las audiencias de noticias digitales en México.

Resumo Este estudo tem como objetivo determinar se o consumo de notícias, a confiança na mídia, o interesse por notícias e a afinidade política estão correlacionados com o segmento geracional do mercado, que é medido pela idade do público digital. Além disso, busca verificar se a idade está associada à atenção que o público digital dá aos jornalistas e às marcas de notícias. A pesquisa se baseia em uma pesquisa realizada pela YouGov com 2.005 pessoas no México em 2022, encomendada pelo Instituto Reuters para o Estudo do Jornalismo da Universidade de Oxford como parte de seu Relatório de Notícias Digitais anual. Para isso, o método estatístico usado foi o teste qui-quadrado de independência entre variáveis qualitativas. Os resultados mostraram que a idade se correlaciona com os hábitos de consumo, com a confiança na mídia, com o interesse em notícias e com a afinidade política do público digital no México. Por sua vez, a idade está relacionada ao fato de as audiências digitais se concentrarem em jornalistas ou marcas de notícias na América Latina. Este estudo aprofunda nas características das audiências digitais no México e faz uma contribuição a partir de uma perspectiva de economia, gestão e marketing da mídia. Ele também ajuda as organizações de mídia a avaliar a estratégia de mercado, a segmentação de mercado e os mercados-alvo com base nas características das audiências de notícias digitais no México.

Abstract This study aims to determine whether news consumption, media trust, interest in news, and political affinity are correlated with the generational segment of the market, measured by the age of the digital audience. Furthermore, this study seeks to verify whether age is associated with digital audiences' paying attention to journalists or news brands. This research is based on a survey of 2,005 people in Mexico by YouGov in 2022, commissioned by the Reuters Institute for the Study of Journalism at the University of Oxford as part of its annual Digital News Report. The statistical method used was the chi-square test of independence between qualitative variables. The results showed that age correlates with consumption habits, media trust, interest in news, and citizens' political affinity for digital audiences in Mexico. Moreover, age relates to whether digital audiences focus on journalists or news brands in Latin America. This study delves into the characteristics of digital audiences in Mexico and contributes from a media economics, media management, and media marketing perspective. It also helps media organizations assess their market strategy, market segmentation, and target markets based on the characteristics of digital news audiences in Mexico.

Resumen: Este estudio se lleva a cabo para identificar los principales riesgos en los proyectos de construcción en Palestina desde la perspectiva de los contratistas. Una encuesta con cuestionario es la herramienta utilizada para lograr los objetivos del estudio. Se identificaron y clasificaron como factores externos e internos, 41 factores de riesgo críticos en los proyectos asociados con la duración del proyecto. Un mapa de riesgos para las causas investigadas se creó en base a su gravedad percibida y frecuencia de ocurrencia. Los hallazgos revelaron 29 causas en la zona roja del mapa de riesgos, 11 en la zona amarilla y 1 en la zona verde. Los factores de riesgo más graves son: pagos atrasados por parte del propietario, cierre por parte de las autoridades, mala supervisión del sitio, segmentación de las gobernaciones palestinas, pagos tardíos a subcontratistas, órdenes de cambio por ajustes frecuentes del propietario, falta de mano de obra calificada, fecha de finalización irrealista de los proyectos, condiciones climáticas y rehacer el trabajo debido a incumplimiento de los documentos del contrato.

Abstract: This study is conducted to identify the main risks in construction projects in Palestine from the contractors’ perspective. A questionnaire survey is the tool used to achieve the objectives of the study. 41 critical risk factors in projects associated with project duration were identified and classified as external and internal factors. The risk map for the investigated causes is created based on their perceived severity and frequency of occurrence. The findings revealed 29 causes in the red zone of the risk map, 11 in the yellow zone and 1 in the green zone. The top severe risk factors are: delayed payments by the owner, closure by the authorities, poor site supervision, segmentation of the Palestinian governorates, late payments to subcontractors, change orders by frequent adjustments of the owner, lack of qualified labors, unrealistic completion date of projects, weather conditions, and rework due to noncompliance with contract documents.

ABSTRACT Introduction: Breast cancer remains one of the most prevalent cancers globally, specifically the most common in females. The use of artificial intelligence promises to contribute to early diagnosis through imaging. Previously, the landscape and evolution of this scientific production have not been described. Methods: Cross-sectional bibliometric study using Scopus as the data source. The bibliometrix package in R was employed for calculating bibliometric indicators and visualizing the results. Results: 1292 documents published between 1989 and 2024 were selected. 75.3% (n=973) were articles with primary data, followed by 16.2% (n=209) corresponding to reviews. An international collaboration rate of 26.5% was identified, with an annual production growth of 10.78%. It was observed that risk classification through screening, digital breast tomosynthesis, transfer learning, segmentation, and feature selection were the most commonly used keywords. In the last five years, deep learning and mammography have been the most popular topics. International collaboration has been led by the United States, China, and the United Kingdom. Conclusion: A notable growth in global research on the use of artificial intelligence in breast cancer imaging for detection was identified, particularly since the 2010s, primarily through the publication of articles with primary data. The relationship between artificial intelligence and imaging for breast cancer diagnosis has focused on risk and prediction.

RESUMEN Introducción: El cáncer de mama sigue siendo uno de los cánceres más frecuentes a nivel global, específicamente, el más frecuente en el sexo femenino. El uso de inteligencia artificial promete contribuir al diagnóstico precoz, a través de la imagenología. Previamente, no se ha descrito el panorama y avance de esta producción científica. Métodos: Estudio bibliométrico de corte transversal, que usó Scopus como fuente de datos. Se utilizó el paquete bibliometrix de R para el cálculo de indicadores bibliométricos y visualización de los resultados. Resultados: Se seleccionaron 1292 documentos, publicados entre 1989 y 2024. El 75,3% (n=973) fueron artículos con datos primarios, seguido de un 16,2% (n=209) correspondiente a revisiones. Se identificó una colaboración internacional del 26,5%, y un crecimiento anual de la producción del 10,78%. Se observó que, la clasificación de riesgo por screening, tomosíntesis digital de la mama, aprendizaje por transferencia, segmentación y selección por características, son las palabras clave más comúnmente usadas. En los últimos cinco años, el aprendizaje profundo y la mamografía, han sido los temas con mayor popularidad. La colaboración internacional, ha sido liderada por Estados Unidos, China y Reino Unido. Conclusiones: Se identificó un crecimiento notable en la investigación global sobre el uso de inteligencia artificial en imagenología para la detección de cáncer de mama, marcado a partir de la década del 2010, esencialmente por medio de publicación de artículos con datos primarios. La relación entre inteligencia artificial e imagenología para diagnóstico de cáncer de mama, se ha centrado en riesgo y predicción.

Abstract Vegetarianism is a growing food practice. This subject has characteristically been a habit for individuals, but it has not commonly been practiced by families. The objective of this study is to identify the difficulties in the food field experienced by families with children under 12 who follow vegetarian diets, based on their lived experience. A qualitative study based on lived experience was carried out to identify these difficulties using a semi-structured interview, which was answered by both the caregivers and some children. Sixteen families were interviewed, from which the analysis categories of “social and professional support” and “food alternatives” were derived. As a result of “social and professional support,” segmentation and exclusion were identified for having this type of diet, especially towards children; judgments regarding the upbringing of vegetarian infants come from family members and outsiders; and there is also little support from health personnel. In relation to “vegetarian alternatives,” families perceive it as a challenge to get their children to eat food that has been prepared at home and that provides the required nutrients. They generally cook everything at home, especially food for their children. There is also difficulty with vegetarian options because some are considered high-cost. It is concluded that having a vegetarian diet can cause segmentation and that it becomes more complex when families with children are involved.

Resumen El vegetarianismo es una práctica alimentaria que va en aumento. Es un tema que se ha caracterizado a nivel individual, pero poco familiar. El objetivo de este estudio es identificar las dificultades en el ámbito alimentario experimentadas por familias con hijos menores de 12 años que siguen dietas vegetarianas, a partir de su experiencia vivida. Se llevó a cabo un estudio cualitativo fundamentado en la experiencia vivida, con el fin de identificar las dificultades usando una entrevista semiestructurada, la cual fue respondida tanto por los cuidadores como por algunos niños. Se entrevistaron 16 familias, de las cuales se derivaron las categorías de análisis “apoyo social y profesional” y “alternativas alimentarias”. Como resultados del “apoyo social y profesional”, se identificó segmentación o exclusión por tener este tipo de alimentación, especialmente con los niños/niñas; los juicios respecto a la crianza de infantes vegetarianos provienen de familiares y personas externas; y también se identifica poco apoyo del personal de salud. En relación con las “alternativas vegetarianas”, las familias perciben como un desafío que se consuma lo preparado y que proporcione los nutrientes que se requieren; generalmente preparan todo en casa, especialmente lo que consumen los niños; también se evidencia dificultad con las opciones vegetarianas porque algunas se consideran de alto costo. Se concluye que tener una dieta vegetariana puede provocar segmentación y que se hace más complejo cuando se trata de familias con infantes.

Abstract: Computer vision, for decades, has been involved in solving problems in everyday life, under the implementation of different computational methods, that have evolved over time. Feature extraction, along with other computer techniques, is considered a way to develop computer vision systems; currently, plays an important role, considered a complex task, allowing to obtain essential descriptors of the segmented images, differentiating particular characteristics between different classes, even when they share similarity with each other, guaranteeing the delivery of information not redundant to classification algorithms. Likewise, in this work, a computer vision system has been developed for the recognition of foliar damage caused by diseases and pests in tomato plants. The methodology implemented is based on four modules: preprocessing, segmentation, feature extraction, and classification; in the first module, the image is preprocessed of a color space RGB to L*a*b*; in the second module, the area interest was segmented, under the implementation of the algorithm principal component analysis PCA; in the third module, features are extracted from the area of interest, obtaining texture descriptors with the Haralick algorithm, and chromatic features through Contrast descriptors, Hu moments, Gabor characteristics, Fourier descriptors, and discrete cosine transform DCT; in the fourth module, the performance of the classification algorithms were tested, with the characteristics obtained from the previous stage, considering: SVM, Backpropagation, Logistic Regression, KNN, and Random Forests.

ABSTRACT Purpose: To evaluate the effectiveness of mapping-targeted biopsies (MTB) on the index lesion for the detection of clinically significant prostate cancer (csPCa) in transperineal fusion-image prostate biopsies. Materials and Methods: A retrospective analysis was conducted on 309 men with suspected PCa who underwent prostate biopsies at the Creu Blanca reference center in Barcelona, Spain. The Prostate Imaging-Reporting and Data System (PI-RADS v.2.1) of the magnetic resonance images (MRI) were reclassified by an expert radiologist reading of pre-biopsy biparametric MRI used for segmentation of suspected lesions. Transperineal MTB of suspicious lesions and 12-core systematic biopsies were performed using the Artemis™ platform. CsPCa was defined as International Society of Urological Pathology grade group ≥ 2. Results: CsPCa was detected in 192 men (62.1%), with detection rates of 6.3% for PI-RADS 2, 26.8% for PI-RADS 3, 87.3% for PI-RADS 4, and 93.1% for PI-RADS 5. MTB of the index lesion identified 185 csPCa (96.3%). CsPCa was detected solely in systematic biopsies in three cases (1.6%), while an additional four cases (2.1%) were identified only in the second suspected lesion. A predictive model for csPCa detection in MTB of the index lesion was developed, with an AUC of 0.918 (95% CI 0.887-0.950). Conclusions: This model had the potential to avoid 23.3% of prostate biopsies without missing additional csPCa cases. MTB of the index lesion was highly effective for identifying csPCa in fusion transperineal prostate biopsies. A developed predictive model successfully reduced the need for almost one quarter of biopsies without missing csPCa cases. Purpose mappingtargeted mapping targeted (MTB (csPCa fusionimage image Methods 30 Barcelona Spain ImagingReporting Imaging Reporting PIRADS PI RADS v.2.1 v21 v 2 1 (MRI prebiopsy pre biopsy 12core core 12 Artemis platform Results 19 62.1%, 621 62.1% , 62 (62.1%) 63 6 3 6.3 268 26 8 26.8 873 87 87.3 4 931 93 93.1 5 18 96.3%. 963 96.3% . 96 (96.3%) 1.6%, 16 1.6% (1.6%) 2.1% 21 (2.1% 0918 0 918 0.91 95% 95 (95 0.8870.950. 08870950 0.887 0.950 887 950 0.887-0.950) Conclusions 233 23 23.3 v.2. v2 62.1 (62.1% 6. 26. 87. 9 93. 96.3 (96.3% 1.6 (1.6% 2.1 (2.1 091 91 0.9 (9 8870 0.8870.950 0887095 0887 0.88 0950 0.95 88 0.887-0.950 23. v.2 62. (62.1 96. (96.3 1. (1.6 (2. 09 0. ( 0.8870.95 088709 088 0.8 095 0.887-0.95 v. (62. (96. (1. (2 0.8870.9 08870 08 0.887-0.9 (62 (96 (1 0.8870. 0.887-0. (6 0.8870 0.887-0 0.887-

Abstract Introduction: University dropout is now considered a complex phenomenon that goes beyond the number of students not enrolled and that is continuously growing, especially in the first years of study. Objective: In the present study, a prediction model combining Survival Analysis, Decision Trees, and Random Forest, under the Machine Learning philosophy, is proposed for the early diagnosis of possible factors causing dropout in university students. Methods: The proposal consists of 3 phases: the Survival Analysis that allows estimating the probability of permanence of the student (survival). Phase 2 starts from the probability value obtained in the previous phase and uses it as a response variable in the modeling process based on Decision Trees to establish survival patterns around the variables considered. Finally, in phase 3, the critical variables in the model are identified using the Random Forest. Results: The proposed methodology allowed the design of a prediction model to identify the main segmentation variables in behavior patterns of possible cases of academic dropout. Conclusion: Even though the proposal was developed considering a particular case of a Chilean university, the efficient combination of metaheuristics allows the extrapolation of the methodology to any context and academic reality. However, the conditions and needs of each institution must be considered.

Resumen Introducción: La deserción universitaria se considera actualmente como un fenómeno complejo que va más alla del número de estudiantes no matriculados, y que viene en continuo crecimiento sobre todo en los primeros años de estudio. Objetivo: En el presente estudio se propone un modelo de predicción que combina el Análisis de Supervivencia, Árboles de Decisión y Random Forest, bajo la filosofía de Machine Learning, para el diagnóstico temprano de los posibles factores de la deserción en estudiantes universitarios. Métodos: La propuesta consta de 3 fases: el Análisis de Supervivencia que permite estimar la probabilidad de permanencia del alumno (supervivencia). La fase 2 parte del valor de probabilidad obtenido en la fase anterior y lo utiliza como variable respuesta en el proceso de modelado basado en los árboles de decisión para establecer patrones de supervivencia en torno a las variables consideradas. Finalmente, en la fase 3 se identifican las variables más importantes en el modelo, utilizando Random Forest. Resultados: La metodología propuesta permitió diseñar un modelo de predicción, que identifica las principales variables de segmentación en patrones de comportamiento de posibles casos de deserción académica. Conclusión: Si bien la propuesta fue desarrollada considerando un caso particular de una universidad chilena, la eficiente combinación de la metaheurística permite la extrapolación de la metodología a cualquier contexto y realidad académica. Sin embargo, se deben considerar las condiciones y necesidades de cada institución.

Resumo Introdução: A deserção universitária é atualmente considerada um fenômeno complexo que vai além do número de estudantes não matriculados, e que vem em contínuo crescimento sobretudo nos primeiros anos de estudo. Objetivo: No presente estudo, é proposto um modelo de predição que combina Análise de Sobrevivência, Árvores de Decisão e Random Forest, sob a filosofia de Machine Learning, para o diagnóstico precoce dos possíveis fatores de deserção em estudantes universitários. Métodos: A proposta consiste em 3 fases: a Análise de Sobrevivência que permite estimar a probabilidade de permanência do aluno (sobrevivência). A fase 2 parte do valor de probabilidade obtido na fase anterior e o utiliza como variável resposta no processo de modelagem baseado em árvores de decisão para estabelecer padrões de sobrevivência em torno das variáveis consideradas. Finalmente, na fase 3, as variáveis críticas do modelo são identificadas usando Random Forest. Resultados: A metodologia proposta permitiu desenhar um modelo de previsão, que identifica as principais variáveis de segmentação em padrões de comportamento de possíveis casos de deserção acadêmica. Conclusão: Embora a proposta tenha sido desenvolvida a partir de um caso particular de uma universidade chilena, a combinação eficiente da meta-heurística permite a extrapolação da metodologia para qualquer contexto e realidade acadêmica. No entanto, devem ser consideradas as condições e necessidades de cada instituição.

Resumen La detección y el diagnóstico de diabetes e hipertensión presentan un escenario desafiante, principalmente en etapas tempranas. Estos dos son degenerativos y de graves consecuencias, provocando daños en múltiples órganos, incluidos los ojos. El oftalmólogo examina el fondo de la retina para discriminar el tipo de enfermedad. En el caso de la retinopatía, como enfermedad crónica, se evidencia en varias clases de progresión. La retinopatía diabética se clasifica en cuatro, que incluyen etapas leve, moderada, grave y proliferativa. En la etapa proliferativa, la retina produce factores vasoproliferativos que estimulan el crecimiento de nuevos vasos sanguíneos, aumentando el riesgo de pérdida de visión. En cuanto a la retinopatía hipertensiva, el método de clasificación más utilizado es la clasificación de Keith-Wagener-Barker. Similar a la retinopatía diabética, también consta de cuatro clases, que van de leves a graves. En la etapa grave, existe un mayor riesgo de desarrollar papiledema o edema retiniano. El objetivo principal de esta investigación es realizar la segmentación y clasificación de imágenes de fondo de retina con retinopatía diabética e hipertensiva. Se propuso una combinación de una red convolucional UNet y una ConvNet para la segmentación de máscara de vasos y la clasificación de retinopatía, respectivamente. El proceso de clasificación se basa en diez clases definidas, donde los valores que van del 0 al 4 representan la retinopatía diabética y los valores del 5 al 9 corresponden a la retinopatía hipertensiva. Los resultados aproximados en la segmentación fueron índices Jaccard de 74%, F1 de 85% y un Accuracy de 96%, y en la clasificación un Accuracy de 80%.

Abstract The detection and diagnosis of diabetes and hypertension present a challenging scenario, mainly in the early stages. These two are degenerative and with severe consequences, causing damage to multiple organs, including the eyes. The ophthalmologist examines the retinal fundus to discriminate between the type of disease. In the case of retinopathy, as a chronic disease, evidence in various classes of progression. Diabetic retinopathy is categorized into four, which include mild, moderate, severe, and proliferative stages. In the proliferative stage, the retina produces vasoproliferative factors that stimulate the growth of new blood vessels, increasing the risk of vision loss. As for the hypertensive retinopathy the most widely used method for classifying is the Keith-Wagener-Barker classification. Like diabetic retinopathy, it consists also of four classes, ranging from mild to severe. In the severe stage, there is an increased risk of developing papilledema or retinal edema. The main objective of this research is to perform segmentation and classification of diabetes and hypertensive retinopathy retinal fundus images. Using a combination of a convolutional network UNet and a ConvNet was proposed for vessel mask segmentation and retinopathy classification, respectively. The classification process relies on ten defined classes, where values ranging from 0 to 4 represent diabetic retinopathy, and values ranging from 5 to 9 correspond to hypertensive retinopathy. The approximate results in the segmentation were Jaccard index of 74%, F1 of 85%, and an Accuracy of 96%, and in the classification was Accuracy of 80%.

Abstract: Effective weed control in crop fields at an early stage is a crucial aspect of modern agriculture. Nonetheless, detecting and identifying these plants in environments with unpredictable conditions remain a challenging task for the agricultural industry. Thus, a two-stage deep learning-based methodology to effectively address the issue is proposed in this work. In the first stage, multi-plant image segmentation is performed, whereas regions of interest (ROIs) are classified in the second stage. In the segmentation stage, a Deep learning model, specifically a UNet-like architecture, has been used to segment the plants within an image following two approaches: resizing the image or dividing the image into patches. In the classification stage, four architectures, including ResNet101, VGG16, Xception, and MobileNetV2, have been implemented to classify different types of plants, including corn and weed plants. A large image dataset was used for training the models. After resizing the images, the segmentation network achieved a Dice Similarity Coefficient (DSC) of around 84% and a mean Intersection over Union (mIoU) of around 74%. On the other hand, when the images were divided into patches, the segmentation network achieved a mean DSC of 87.48% and a mIoU of 78.17%. Regarding the classification, the best performance was achieved by the Xception network with a 97.43% Accuracy. Then, According to the results, the proposed approach is a beneficial alternative for farmers as it offers a method for detecting crops and weeds under natural field conditions.

Abstract: Detection and segmentation of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV2 or COVID-19) is a difficult task due the different kinds of shapes, sizes and positions of the injury. Medical institutions have vast challenges because there is an urgent need for efficient tools to improve the diagnosis of COVID-19 patients. Computer tomography images (CT) are necessary for medical specialists to diagnose the patient’s condition. Nevertheless, there is a lack of both in Medical Centers, mainly in rural areas. The manual analysis of CT images is time-consuming; in addition, most images have low contrast, and it is possible to find blood vessels in the background, so the difficulty of a suitable diagnosis increases. Nowadays, deep learning methods are an alternative method to perform the detection and segmentation task. In this work, we propose a novel light model to detect and identify COVID-19 using CT images: MiniCovid-Unet. It is an improved version of U-net; main differences reside on the decoder and encoder architecture, MiniCovid-Unet needs fewer convolution layers and filters because it focuses only on COVID-19 images. Also, as a result of employing fewer parameters, it can be trained in less time, and the resulting model is light enough to be downloaded to a mobile device. In this way, it is possible to have a quick and confident diagnosis in remote areas, where there exists an absence of internet connection and medical specialists.

Abstract The contribution of anaphoric encapsulators to the information structure of the utterance has often been restricted to observations regarding their position as sentence topics (theme) or as part of the comment (rheme). However, they may appear in other positions that can be better delimited with the help of a model of discourse segmentation such as the Basel Model. I explore the different discourse functions of anaphoric encapsulators in the three main units delineated in this model: the Nucleus or propositional core of the utterance, the Frame or the information chunk preceding the Nucleus, and the Appendix or slot for background information. Besides, I claim that the possibility for an anaphoric encapsulator to express the discourse topic of (part of) a text is intimately related to its position in the Utterance and propose a tentative scale of access to discourse topicality for this type of anaphors. Keywords: anaphora, encapsulation, information structure, discourse topic, Basel Model.

Resumen El análisis del papel que juegan los encapsuladores anafóricos en la estructura informativa de los enunciados se ha limitado por lo general a observaciones relativas a la posición que ocupan como temas o bien como parte del rema de los enunciados. Sin embargo, estas anáforas aparecen en otras posiciones enunciativas que pueden delimitarse con la ayuda de un modelo de segmentación del discurso como el Modelo de Basilea. En este trabajo se exploran las diversas funciones discursivas de los encapsuladores anafóricos en las tres unidades principales establecidas por este modelo: el Núcleo o relación proposicional central del enunciado, el Marco o segmento informativo que precede al Núcleo, y el Apéndice o hueco estructural para la información secundaria. Además, se defiende que la posibilidad de que un encapsulador anafórico exprese el tópico discursivo de (parte de) un texto está íntimamente ligado a su posición en el Enunciado y se propone una escala tentativa de acceso a la topicalidad discursiva para este tipo de anáforas. Palabras clave: anáfora, encapsulación, estructura informativa, tópico discursivo, Modelo de Basilea.