Knowledge Distillation in Video-Based Human Action Recognition: An Intuitive Approach to Efficient and Flexible Model Training.

Training a model to recognize human actions in videos is computationally intensive. While modern strategies employ transfer learning methods to make the process more efficient, they still face challenges regarding flexibility and efficiency. Existing solutions are limited in functionality and rely heavily on pretrained architectures, which can restrict their applicability to diverse scenarios. Our work explores knowledge distillation (KD) for enhancing the training of self-supervised video models in three aspects: improving classification accuracy, accelerating model convergence, and increasing model flexibility under regular and limited-data scenarios. We tested our method on the UCF101 dataset using differently balanced proportions: 100%, 50%, 25%, and 2%. We found that using knowledge distillation to guide the model's training outperforms traditional training without affecting the classification accuracy and while reducing the convergence rate of model training in standard settings and a data-scarce environment. Additionally, knowledge distillation enables cross-architecture flexibility, allowing model customization for various applications: from resource-limited to high-performance scenarios.

Virtual reality as a non-medical tool in the treatment of anxiety, pain, and perception of time in children in the maintenance phase of acute lymphoblastic leukemia treatment.

Introduction: Management of pediatric cancer patients involves invasive procedures such as punctures, injections, catheter placements, and chemotherapy which can generate fatigue, nausea, vomiting, anxiety, and pain. Virtual Reality (VR) is a nonpharmacological intervention classified as a cognitive-behavioral method to relieve symptoms. Methods: We designed a crossover protocol and included 20 patients between 9 and 12 years old; ten were male. All patients had acute lymphoblastic leukemia diagnosis and were treatedwith St. Jude's XV protocol in the maintenance phase. Pain and anxiety were measured with validated scales in the pediatric population. Results: Although we used a small group of patients, we found statistical difference in the reduction of anxiety and perception of time. Discussion: These results open a window to non-pharmacological treatments and show a strategy to improve quality of life in children inside the hospital.

Not only a therapeutic target; mTOR in Hodgkin lymphoma and acute lymphoblastic leukemia.

Introduction: The mechanistic/mammalian target of rapamycin (mTOR) is a serine/threonine kinase, which is downregulated or upregulated and is implicated in different types of cancer including hematologic neoplasms, skin prostate, and head and neck cancer. Aim: The aim of this study was to explore the current knowledge of mTOR signaling in acute lymphoblastic leukemia and Hodgkin lymphoma. Methods: A systematic review was performed according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, searching PubMed, Discovery Service for National Autonomous University of Mexico, Registro Nacional de Instituciones y Empresas Científicas y Tecnológicas (RENIECYT), and Scientific Electronic Library Online (SciELO) from 1994 to 2023. A total of 269 papers were identified for acute lymphoblastic leukemia, but based on specific criteria, 15 were included; for Hodgkin lymphoma, 110 papers were identified, but 5 were included after manual searching. Results: A total of 20 papers were evaluated, where mTOR activity is increased in patients with Hodgkin lymphoma and acute lymphoblastic leukemia by different molecular mechanisms. Conclusions: mTOR activity is increased in patients with both hematologic neoplasms and NOTCH; interleukin 4, 7, and 9, and nuclear proteins have been studied for their role in the activation of mTOR signaling.

Biografía mediática de la apropiación de medios de comunicación y TIC en I.E. en Cartago - Colombia / Media biography of the appropriation of media and ICT in I.E., in Cartago, Colombia / Biografia da mídia e apropriação das TIC em I.E. em Cartago - Colômbia

RESUMEN Se presentan resultados del proyecto de investigación, mapa de conocimiento del uso de medios de comunicación en instituciones educativas de básica y media realizado en Cartago, Colombia en la I.E. (Institución educativa) GABO, la I.E Ciudad Cartago y la I.E Antonio Holguín Garcés. La metodología aplicada fue cualitativa y se utiliza la Investigación Acción Participativa para la construcción de biografías mediáticas a modo de síntesis de resultados. Se identifica una desarticulación entre las políticas públicas e iniciativas del estado a nivel nacional con las dinámicas regionales, un interés particular de la Secretaria de Educación municipal en las acciones por la calidad educativa a nivel local y una postura positiva, crítica y significativa en el papel de las TIC y los medios de comunicación en los procesos de aprendizaje en los estudiantes.

ABSTRACT Results of the research project are presented, map of knowledge of the use of media in basic and medium educational institutions carried out in Cartago, Colombia, in I.E GABO, I.E Ciudad Cartago and I.E Antonio Holguín Garcés. The methodology applied was qualitative and Participatory Action Research is used for the construction of Media Biographies as a synthesis of results. A disarticulation is identified between public policies and initiatives of the state at the national level with regional dynamics, a particular interest of the municipal education secretary in actions for educational quality at the local level and a positive, critical posture and significant in the role of ICT and the media in student learning processes.

RESUMO São apresentados os resultados do projecto de investigação, mapa de conhecimento do uso dos media nas instituições de ensino básico e secundário realizado em Cartago, Colômbia, na I.E. (Instituição Educativa) GABO, na I.E. Ciudad Cartago e na I.E. Antonio Holguín Garcés. A metodologia aplicada foi qualitativa e utilizou a Investigação de Acção Participativa para a construção de biografias dos meios de comunicação como síntese de resultados. É identificada uma desarticulação entre políticas públicas e iniciativas estatais a nível nacional com dinâmica regional, um interesse particular da Secretaria Municipal de Educação nas acções para a qualidade educativa a nível local e uma posição positiva, crítica e significativa sobre o papel das TIC e dos media nos processos de aprendizagem dos estudantes.

Regional and state-level patterns of type 2 diabetes prevalence in Mexico over the last three decades.

AIMS: We aimed to characterize and illustrate the regional and state-level change in type 2 diabetes (T2D) prevalence in Mexico between 1990 and 2017. METHODS: We conducted an ecological and secondary analysis using data from the Global Burden of Disease study on T2D prevalence of the adult Mexican population. We estimated the absolute increase and annual growth rate of T2D prevalence between 1990 and 2017, stratified by age group and region. RESULTS: Nationally, between 1990 and 2017, the prevalence of T2D in Mexico increased from 9.5% to 14.3%. The highest increase in T2D prevalence was observed in the East and Southcentral regions, with the lowest absolute change in T2D prevalence observed in Northern states. The highest average annual growth rate in T2D prevalence was observed in Southern Mexico, in the three Southern states with the lowest human development index, and among individuals ages 15-49 years across all regions, compared to those 50 years and older. CONCLUSIONS: The prevalence of T2D in Mexico has increased substantially over the past three decades, with a clear shift in T2D prevalence from Northern to Southern states and a faster increase occurring in Southern Mexico among younger adults and in areas with lower economic resources.

Aberrant immunophenotypes in acute lymphoblastic leukemia / Inmunofenotipos aberrantes en la leucemia linfoblástica aguda

Abstract Acute lymphoblastic leukemia (ALL) is the most prevalent hematologic neoplasia worldwide. To classify leukemia, we analyzed the immunophenotypic characteristics in the neoplastic cells obtained with antibodies by cell flow cytometry or immunohistochemistry. The aberrant immunophenotypes are antigen expression patterns that differ from the normal hematopoietic maturation process, which can include some different lineage antigens such as myeloid antigens in ALL or asynchronous expression of antigens. These aberrant immunophenotypes have been studied as prognostic factors and residual disease markers. In this review, some aspects of aberrant immunophenotypes are addressed, including definition, epidemiology, and potential uses.

Resumen La leucemia linfoblástica aguda es la neoplasia hematológica más prevalente en el mundo. Para clasificar la leucemia se utilizan características inmunofenotípicas en las células neoplásicas que se pueden observar con anticuerpos en la citometría de flujo o mediante inmunohistoquímica. Los inmunofenotipos aberrantes son los patrones de expresión de los antígenos que difieren del proceso normal de maduración hematopoyética, y pueden incluir antígenos de linajes diferentes, como antígenos mieloides en la leucemia linfoblástica aguda, o la expresión asincrónica de antígenos. Estos inmunofenotipos aberrantes se han estudiado como factores de pronóstico y como marcadores de enfermedad mínima residual. En esta revisión se abordan algunos aspectos de los inmunofenotipos aberrantes, incluyendo su definición, epidemiología y usos potenciales.

Aberrant immunophenotypes in acute lymphoblastic leukemia.

Acute lymphoblastic leukemia (ALL) is the most prevalent hematologic neoplasia worldwide. To classify leukemia, we analyzed the immunophenotypic characteristics in the neoplastic cells obtained with antibodies by cell flow cytometry or immunohistochemistry. The aberrant immunophenotypes are antigen expression patterns that differ from the normal hematopoietic maturation process, which can include some different lineage antigens such as myeloid antigens in ALL or asynchronous expression of antigens. These aberrant immunophenotypes have been studied as prognostic factors and residual disease markers. In this review, some aspects of aberrant immunophenotypes are addressed, including definition, epidemiology, and potential uses.

La leucemia linfoblástica aguda es la neoplasia hematológica más prevalente en el mundo. Para clasificar la leucemia se utilizan características inmunofenotípicas en las células neoplásicas que se pueden observar con anticuerpos en la citometría de flujo o mediante inmunohistoquímica. Los inmunofenotipos aberrantes son los patrones de expresión de los antígenos que difieren del proceso normal de maduración hematopoyética, y pueden incluir antígenos de linajes diferentes, como antígenos mieloides en la leucemia linfoblástica aguda, o la expresión asincrónica de antígenos. Estos inmunofenotipos aberrantes se han estudiado como factores de pronóstico y como marcadores de enfermedad mínima residual. En esta revisión se abordan algunos aspectos de los inmunofenotipos aberrantes, incluyendo su definición, epidemiología y usos potenciales.

Efectos a la salud de dos bebidas carbonatadas en ratones experimentales / Health Effects of Two Carbonated Drinks in Experimental Mice

El consumo de bebidas carbonatadas es común a nivel mundial. En El Salvador es frecuente en la dieta de la población; sin embargo, algunos de sus componentes pueden ser responsables de afectar la salud. Objetivo. Evaluar los efectos a la salud de dos bebidas carbonatadas administradas continuamente durante 10 semanas a ratones experimentales. Metodología. Se utilizaron 12 ratones distribuidos en 3 grupos de 4 ratones cada uno; un control y dos experimentales, para administrar dos bebidas carbonatadas azucaradas de alto consumo dentro de la población salvadoreña por vía intragástrica. Resultados. Los chequeos clínicos presentaron alteraciones en algunos aspectos evaluados, como deshidratación, piloerección y diarrea. En peso corporal, hubo diferencias entre el grupo control y los experimentales. En la evaluación macroscópica de los órganos, los grupos tratados sufrieron irregularidades, tanto en la apariencia como en el color, aunque en su peso no existieron diferencias, a excepción del riñón derecho del grupo tratado con bebida carbonatada 1. La química sanguínea mostró únicamente diferencia en el colesterol total del grupo tratado con bebida carbonatada 2. Conclusión. La apariencia de los ratones tratados con bebidas carbonatadas mostró daños a la salud, principalmente el daño provocado en la apariencia de los órganos internos

Currently, the consumption of carbonated beverages is very common worldwide. In El Salvador it is frequent in the diet of the population; while some constituent components may be responsible for affecting health. Objective. To assess the health effects of two carbonated beverages administered continuously for 10 weeks to experimental mice. Methodology. Two carbonated beverages with high consumption sugar were chosen within the Salvadoran population. In this study 12 mice distributed in 3 groups of 4 mice each were used; one control and two experimental. The substances were administered intragastrically. Results. The clinical check-ups showed alterations in some aspects evaluated, such as dehydration, piloerection and diarrhea. In body weight, there were significant differences between the control and the experimental group. In the macroscopic evaluation of the organs, the treated groups suffered from certain irregularities, both in appearance and color, although there were no differences in weight, except for the right kidney of the group treated with carbonated beverage 1. Blood chemistry showed only

Tackling maternal mortality by improving technical efficiency in the production of primary health services: longitudinal evidence from the Mexican case.

Ensuring regular and timely access to efficient and quality health services reduces the risk of maternal mortality. Specifically, improving technical efficiency (TE) can result in improved health outcomes. To date, no studies in Mexico have explored the connection of TE with either the production of maternal health services at the primary-care level or the maternal-mortality ratio (MMR) in populations without social security coverage. The present study combined data envelopment analysis (DEA), longitudinal data and selection bias correction methods with the purpose of obtaining original evidence on the impact of TE on the MMR during the period 2008-2015. The results revealed that MMR fell 0.36% (P < 0.01) for every percentage point increase in TE at the jurisdictional level or elasticity TE-MMR. This effect proved lower in highly marginalized jurisdictions and disappeared entirely in those with low- or medium-marginalization levels. Our findings also highlighted the relevance of certain social and economic aspects in the attainment of TE by jurisdictions. This clearly demonstrates the need for comprehensive, cross-cutting policies capable of modifying the structural conditions that generate vulnerability in specific population groups. In other words, achieving an effective and sustainable reduction in the MMR requires, inter alia, that the Mexican government review and update two essential elements: the criteria behind resource allocation and distribution, and the control mechanisms currently in place for executing and ensuring accountability in these two functions.

Measuring and analysing social efficiency in the production of maternal health services in Mexico, 2008-15.

'Social efficiency' (SE) denotes the capacity of health systems to ensure equitable access to quality health services at no financial risk to users. Mexico and other low- and middle-income countries have rarely studied the performance of their health systems from an SE perspective. We propose a metric for assessing SE in the production of maternal health services in the public sector among populations without social security, analysing contextual correlates of the demand for these services. Analysis was based on administrative data collected from the 243 health jurisdictions (HJs) in Mexico for the period 2008-15. We defined production inputs as the availability of physical and human resources and social product as the unweighted sum of social sub-products, including an equitable distribution of maternal health resources, the provision of quality maternal health care and financial protection for users. We described the SE scores, the main contextual characteristics as well as those related to the demand for maternal health services. We then performed a variance decomposition analysis of the SE score by component and estimated the SE territorial concentration patterns. Finally, we identified the structural characteristics modelling SE by means of a spatial autoregressive panel data model with fixed effects by year. The SE score rose from 57.7% in 2008 to 71.9% in 2015 (P < 0.01), with its quality component accounting for the largest proportion of variance (30%). SE peaked in HJs with low social marginalization and rurality, and with service demand characterized by low parity and older populations. Different SE levels demonstrated territorial concentration patterns. Analysing SE as a metric for health system performance offers elements that contribute to the achievement of UHC as well as to the design and implementation of effective maternal health interventions intended particularly for the most socially vulnerable sectors of the population.

A spatio-temporal cluster analysis of technical efficiency in the production of outpatient maternal health services and its structural correlates in México.

OBJECTIVE: To analyze the formation of spatial clusters of technical efficiency (TE) in the production of outpatient maternal health services in México for the period 2008 to 2015. METHODS: We performed a longitudinal analysis of administrative and structural data related to the 243 Mexican health jurisdictions. We use window data envelopment analysis and spatial and econometric techniques. Structural correlates of each TE cluster obtained were identified estimating a pooled multinomial logit model. RESULTS: We observed an increase in the overall TE, accompanied by a reduction in its standard deviation. Furthermore, we identified positive TE spatial dependence both globally and locally. Multiple regression analysis showed that the maximum-performance TE cluster was composed of health jurisdictions located in the North of México and characterized by social marginalization, a reduced indigenous population, and a low demand for maternal health services. CONCLUSIONS: The identification of TE clusters can provide elements to induce decision makers to innovative in ways of allocating resources and manage their utilization. In México as other low- and middle-income countries, it is key to develop targeting strategies to implement specific health services innovations putting to the population to be served at the front of the strategy. This implies testing new modalities to strengthen primary health services, the empowerment of community participation, the training and allocation of health personnel that could respond adequately to population's demand, and the active involvement of competent state and local authorities in the assessment of the results of these innovations.

Monitoring Student Activities with Smartwatches: On the Academic Performance Enhancement.

Motivated by the importance of studying the relationship between habits of students and their academic performance, daily activities of undergraduate participants have been tracked with smartwatches and smartphones. Smartwatches collect data together with an Android application that interacts with the users who provide the labeling of their own activities. The tracked activities include eating, running, sleeping, classroom-session, exam, job, homework, transportation, watching TV-Series, and reading. The collected data were stored in a server for activity recognition with supervised machine learning algorithms. The methodology for the concept proof includes the extraction of features with the discrete wavelet transform from gyroscope and accelerometer signals to improve the classification accuracy. The results of activity recognition with Random Forest were satisfactory (86.9%) and support the relationship between smartwatch sensor signals and daily-living activities of students which opens the possibility for developing future experiments with automatic activity-labeling, and so forth to facilitate activity pattern recognition to propose a recommendation system to enhance the academic performance of each student.

Effective and Generalizable Graph-Based Clustering for Faces in the Wild.

Face clustering is the task of grouping unlabeled face images according to individual identities. Several applications require this type of clustering, for instance, social media, law enforcement, and surveillance applications. In this paper, we propose an effective graph-based method for clustering faces in the wild. The proposed algorithm does not require prior knowledge of the data. This fact increases the pertinence of the proposed method near to market solutions. The experiments conducted on four well-known datasets showed that our proposal achieves state-of-the-art results, regarding the clustering performance, also showing stability for different values of the input parameter. Moreover, in these experiments, it is shown that our proposal discovers a number of identities closer to the real number existing in the data.

A longitudinal assessment of technical efficiency in the outpatient production of maternal health services in México.

We assess technical efficiency (TE) level for Mexican Ministry of Health (MoH) primary care units. Assessment was focused on the production of adequate maternal health services defined as the coverage level of women who received timely and frequent antenatal care, and institutional and medical care during childbirth. We conducted a longitudinal analysis of administrative and socio-demographic information concerning 233 health jurisdictions for the period 2008-15. Crude TE was calculated using window data envelopment analysis (Windows-DEA). Empirical analysis included the description of several factors affecting the production of maternal health services, including the heterogeneity and trends assessment of TE among health jurisdictions. We estimated a pooled regression model with robust standard errors to identify correlates of TE and estimated adjusted performance scores. Results indicate that while the production of adequate maternal-health services and TE in health jurisdictions proved insufficient, they rose by 22% (from 40.9% to 49.8%) and 14% (from 54.3% to 62%), respectively, over time. Furthermore, variance in efficiency among production units diminished and persistent regularities were observed. Performance was highest in the Northern as opposed to the Southern and Southeastern health jurisdictions, but lowest in the most marginalized zones of the country marked by economic inequality and the presence of indigenous populations. The Mexican Health System has reached a paradoxical situation: the steady escalation of financial resources in the public health subsystem over the past 15 years has yielded sub-optimal results as regards coverage for essential maternal health interventions among the poorest. Mexican government must put in place a set of measures to guarantee efficiency in the system's performance without affecting equity gains. This necessarily involves reconsidering, and where necessary replacing, the criteria behind the allocation and distribution of resources, as well as the mechanisms for controlling how resources are used and accountability is fulfilled.

HyRA: A Hybrid Recommendation Algorithm Focused on Smart POI. Ceutí as a Study Scenario.

Nowadays, Physical Web together with the increase in the use of mobile devices, Global Positioning System (GPS), and Social Networking Sites (SNS) have caused users to share enriched information on the Web such as their tourist experiences. Therefore, an area that has been significantly improved by using the contextual information provided by these technologies is tourism. In this way, the main goals of this work are to propose and develop an algorithm focused on the recommendation of Smart Point of Interaction (Smart POI) for a specific user according to his/her preferences and the Smart POIs' context. Hence, a novel Hybrid Recommendation Algorithm (HyRA) is presented by incorporating an aggregation operator into the user-based Collaborative Filtering (CF) algorithm as well as including the Smart POIs' categories and geographical information. For the experimental phase, two real-world datasets have been collected and preprocessed. In addition, one Smart POIs' categories dataset was built. As a result, a dataset composed of 16 Smart POIs, another constituted by the explicit preferences of 200 respondents, and the last dataset integrated by 13 Smart POIs' categories are provided. The experimental results show that the recommendations suggested by HyRA are promising.

Transparent Low Electrostatic Charge Films Based on Carbon Nanotubes and Polypropylene. Homopolymer Cast Films.

Use of multi-wall carbon nanotubes (MWCNTs) in external layers (A-layers) of ABA-trilayer polypropylene films was investigated, with the purpose of determining intrinsic and extrinsic factors that could lead to antistatic behavior of transparent films. The incorporation of 0.01, 0.1, and 1 wt % of MWCTNs in the A-layers was done by dilution through the masterbatch method. Masterbatches were fabricated using isotactic polypropylene (iPP) with different melt flow indexes 2.5, 34, and 1200 g/10 min, and using different ultrasound assist methods. It was found that films containing MWCNTs show surface electrical resistivity of 1012 and 1016 Ω/sq, regardless of the iPP melt flow index (MFI) and masterbatch fabrication method. However, electrostatic charge was found to depend upon the iPP MFI, the ultrasound assist method and MWCNT concentration. A percolation electron transport mechanism was determined most likely responsible for this behavior. Optical properties for films containing MWCNTs do not show significant differences compared to the reference film at MWCNT concentrations below 0.1 wt %. However, an enhancement in brightness was observed, and it was attributed to ordered iPP molecules wrapping the MWCNTs. Bright transparent films with low electrostatic charge were obtained even for MWCNTs concentrations as low as 0.01 wt %.

Fogging Control on LDPE/EVA Coextruded Films: Wettability Behavior and Its Correlation with Electric Performance.

The transformation of fog at a non-visible water layer on a membrane of low-density polyethylene (LDPE) and ethylene-vinyl acetate (EVA) was evaluated. Nonionic surfactants of major demand in the polyolefin industry were studied. A kinetic study using a hot fog chamber showed that condensation is controlled by both the diffusion and permanency of the surfactant more than by the change of the surface energy developed by the wetting agents. The greatest permanency of the anti-fog effect of the LDPE/EVA surface was close to 3000 h. The contact angle results demonstrated the ability of the wetting agent to spread out to the surface. Complementarily, the migration of nonionic surfactants from the inside of the polymeric matrix to the surface was analyzed by Fourier transform infrared (FTIR) microscopy. Additionally, electrical measurement on the anti-fogging membrane at alternating currents and at a sweep frequency was proposed to test the conductivity and wetting ability of nonionic surfactants. We proved that the amphiphilic molecules had the ability to increase the conductivity in the polyolefin membrane. A correlation between the bulk electrical conductivity and the permanency of the fogging control on the LDPE/EVA coextruded film was found.

A Linked List-Based Algorithm for Blob Detection on Embedded Vision-Based Sensors.

Blob detection is a common task in vision-based applications. Most existing algorithms are aimed at execution on general purpose computers; while very few can be adapted to the computing restrictions present in embedded platforms. This paper focuses on the design of an algorithm capable of real-time blob detection that minimizes system memory consumption. The proposed algorithm detects objects in one image scan; it is based on a linked-list data structure tree used to label blobs depending on their shape and node information. An example application showing the results of a blob detection co-processor has been built on a low-powered field programmable gate array hardware as a step towards developing a smart video surveillance system. The detection method is intended for general purpose application. As such, several test cases focused on character recognition are also examined. The results obtained present a fair trade-off between accuracy and memory requirements; and prove the validity of the proposed approach for real-time implementation on resource-constrained computing platforms.

Metamaterial Behavior of Polymer Nanocomposites Based on Polypropylene/Multi-Walled Carbon Nanotubes Fabricated by Means of Ultrasound-Assisted Extrusion.

Metamaterial behavior of polymer nanocomposites (NCs) based on isotactic polypropylene (iPP) and multi-walled carbon nanotubes (MWCNTs) was investigated based on the observation of a negative dielectric constant (ε'). It is demonstrated that as the dielectric constant switches from negative to positive, the plasma frequency (ωp) depends strongly on the ultrasound-assisted fabrication method, as well as on the melt flow index of the iPP. NCs were fabricated using ultrasound-assisted extrusion methods with 10 wt % loadings of MWCNTs in iPPs with different melt flow indices (MFI). AC electrical conductivity (σ(AC)) as a function of frequency was determined to complement the electrical classification of the NCs, which were previously designated as insulating (I), static-dissipative (SD), and conductive (C) materials. It was found that the SD and C materials can also be classified as metamaterials (M). This type of behavior emerges from the negative dielectric constant observed at low frequencies although, at certain frequencies, the dielectric constant becomes positive. Our method of fabrication allows for the preparation of metamaterials with tunable ωp. iPP pure samples show only positive dielectric constants. Electrical conductivity increases in all cases with the addition of MWCNTs with the largest increases observed for samples with the highest MFI. A relationship between MFI and the fabrication method, with respect to electrical properties, is reported.