Your browser doesn't support javascript.
loading
Show: 20 | 50 | 100
Results 1 - 20 de 170
Filter
2.
Rev. cuba. oftalmol ; 34(2): e1152, 2021.
Article in Spanish | LILACS, CUMED | ID: biblio-1341465

ABSTRACT

El presente trabajo aborda el uso de la inteligencia artificial en la cirugía de catarata y la incursión de Cuba en este campo. La inteligencia artificial tiene como objetivo dotar a un agente con la capacidad de tomar decisiones correctas. Dentro de los campos de la inteligencia artificial se encuentra el aprendizaje de máquinas cuyo propósito es entrenar a las computadoras para aprender de un conjunto de datos las decisiones que han de tomar, dada una situación específica. Uno de los métodos más utilizados para el entrenamiento y el aprendizaje de máquinas es el desarrollo de redes neuronales artificiales. Desde un enfoque social, se explica cómo la influencia sobre el resultado visual que puede lograrse con esta tecnología repercute en el individuo y la sociedad, y se resaltan las ventajas y las desventajas de su utilización(AU)


The study addresses the use of artificial intelligence in cataract surgery and Cuba's incorporation into this field. The purpose of artificial intelligence is to develop agents with the ability to take appropriate decisions. One of the branches of artificial intelligence is machine learning, whose aim is to train computers to draw from a set of data the decisions to be taken in response to a specific situation. One of the most common methods in machine training and learning is the development of artificial neural networks. A social explanation is provided of the effect of the visual outcomes obtained by this technology on the individual and society, highlighting the advantages and disadvantages of its use(AU)


Subject(s)
Humans , Artificial Intelligence , Cataract Extraction/methods , Machine Learning
3.
São Paulo; s.n; 2021. 212 p.
Thesis in Portuguese | LILACS | ID: biblio-1293376

ABSTRACT

Objetivo: Esta tese é apresentada no formato de quatro artigos científicos, que estão articulados em torno do objetivo geral, que foi analisar o declínio da mobilidade funcional e a mortalidade em idosos residentes no município de São Paulo. O primeiro artigo identificou os fatores associados ao declínio da mobilidade funcional em idosos ao longo de 15 anos de acompanhamento. O segundo analisou a sobrevida dos idosos por 10 anos de acordo com a mobilidade funcional por meio de regressões de Cox, tendo como desfecho o óbito por todas as causas e as principais causas específicas. O terceiro artigo testou a performance de algoritmos de machine learning na predição de óbito por causas específicas, utilizando como preditores testes de desempenho físico, variáveis de saúde e características sociodemográficas. O quarto artigo analisou a performance dos algoritmos de machine learning para predizer declínio funcional em tarefas de mobilidade. Métodos: Os dados utilizados nas análises dos quatro artigos foram provenientes do Estudo Saúde, Bem-Estar e Envelhecimento (SABE), de múltiplas coortes e representativo para a população de residentes do município de São Paulo com idade igual ou superior a 60 anos. No primeiro artigo, foi realizada uma análise seriada de regressões multinível das quatro ondas do estudo, ocorridas em 2000, 2006, 2010 e 2015, com o objetivo de analisar a prevalência de limitação na mobilidade e as diferenças entre as ondas. Foram ajustadas, também, regressões logísticas separadas para cada onda, para analisar os fatores associados ao declínio da mobilidade. O segundo artigo utilizou regressões de Cox para analisar o tempo até o óbito por todas as causas e por causas específicas, de acordo com a condição de mobilidade, avaliada por dois testes (velocidade da marcha e o teste de levantar e sentar 5 vezes). Para o terceiro artigo, aprovado para publicação na revista Age and Ageing, foi realizada uma predição multinomial com cinco categorias: óbito por doenças do aparelho circulatório, óbito por doenças do aparelho respiratório, óbito por neoplasias, óbito por outras causas específicas e não óbito. Algoritmos preditivos de machine learning foram treinados em 70% da amostra, e em seguida testados nos 30% restantes. No quarto artigo foram utilizados algoritmos de machine learning para predizer a dificuldade na realização de tarefas de mobilidade, como caminhar, subir escadas, agachar e ajoelhar e carregar objetos. Nos dois últimos artigos, a capacidade preditiva dos modelos foi testada por meio da área abaixo da curva ROC, além de outras métricas como a sensibilidade e especificidade. Resultados: O primeiro artigo encontrou um aumento da prevalência de limitação na mobilidade após o ano 2000, mesmo após o ajuste por outros fatores. Foram também verificadas associações do declínio da mobilidade com condições crônicas de saúde e aspectos socioeconômicos. O segundo artigo identificou que o teste de levantar e sentar apresentou associação mais consistente com a mortalidade (HR=1.03, IC95%1.00-1.05) do que a velocidade da marcha. Além disso, indivíduos com imobilidade apresentaram um risco aumentado de morrer por todas as causas (HR=1.71, IC95%1.21-2.42) e por doenças do aparelho circulatório (HR=2.14, IC95%1.25-3.65). No terceiro artigo, o desfecho em que os algoritmos apresentaram melhor poder preditivo foi a mortalidade por doenças do aparelho respiratório (AUC-ROC=0.89). Os algoritmos com melhor desempenho foram o light gradient boosted machine e extreme gradient boosting. No quarto artigo, o random forest foi o algoritmo com melhor performance e os desfechos com as melhores performances preditivas foram a dificuldade de agachar e ajoelhar (AUC-ROC: 0.81) e carregar pesos (AUC-ROC: 0.80). Conclusão: Os resultados da tese trazem novas evidências acerca do declínio da mobilidade funcional e mortalidade de pessoas idosas no Brasil. Além disso, demonstrou que algoritmos preditivos de machine learning podem ser ferramentas importantes para o rastreio de idosos em risco de desfechos negativos e o estabelecimento de medidas preventivas personalizadas.


Objective: This thesis is presented in the format of four articles articulated around the general objective, which was to analyze the functional mobility decline and mortality in older residents from the municipality of São Paulo. The first article identified the associated factors of the decline in functional mobility in older adults over a 15-year follow-up. The second analyzed the 10-years mortality from all-cause and the main specific causes of death in older individuals according to functional mobility by Cox regression models. The third article tested the performance of machine learning algorithms in predicting death from specific causes, using physical performance tests, health and sociodemographic features as predictors. The fourth article analyzed the performance of machine learning algorithms to predict functional decline in mobility tasks. Methods: The data used in the four articles analysis were from the Health, Well-Being, and Aging (SABE) Study, characterized by multiple cohorts and by a representative sample of the older residents from the municipality of São Paulo, aged 60 years and over. In the first article, we performed serial analysis of multilevel regressions of the four waves of the study, collected in 2000, 2006, 2010, and 2015, with the aim of analyzing the prevalence of limitation in mobility and the differences between the waves. Separate logistic regressions were also adjusted for each wave, to analyze the factors associated with the mobility decline. The second article used Cox regressions to analyze the time to all-cause and specific-cause of death, according to the mobility condition, assessed by two tests (gait speed and the 5 times chair stand test). For the third article, approved for publication in the journal Age and Ageing, a multinomial prediction was made with five categories: death from diseases of the circulatory system, death from diseases of the respiratory system, death from neoplasms, death from other specific causes and non-death. Predictive machine learning algorithms were trained in 70% of the sample, and then tested in the remaining 30%. In the fourth article, machine learning algorithms were used to predict the difficulty in performing mobility tasks, such as walking, climbing stairs, crouching and kneeling and carrying objects. In the last two articles, the predictive performance of the models was tested using the area under ROC curve, in addition to other metrics, such as sensitivity and specificity. Results: The first article found an increase in the prevalence of mobility limitations after the year 2000, even after adjusting for other factors. Associations between the decline in mobility with chronic health conditions and socioeconomic aspects were also verified. The second article identified that the stand-up test presented a more consistent association with mortality (HR = 1.03, 95% CI 1.00-1.05) than gait speed. In addition, individuals with immobility had an increased risk of dying from all causes (HR = 1.71, 95% CI 1.21-2.42) and from diseases of the circulatory system (HR = 2.14, 95% CI 1.25-3.65). In the third article, the outcome which the best performance of the algorithms was mortality from diseases of the respiratory system (AUC-ROC = 0.89). The algorithms with best performance were the light gradient boosted machine and extreme gradient boosting. In the fourth article, the random forest presented the best performance and the outcomes with the best predictions were the difficulty of stooping, crouching or kneeling (AUC-ROC: 0.81) and carrying weights (AUC-ROC: 0.80). Conclusion: The results bring new evidence about the decline in functional mobility and mortality of older people in Brazil. In addition, it demonstrated that predictive machine learning algorithms could be important tools to screen older adults at risk of poor outcomes and could help to assign personalized preventive interventions.


Subject(s)
Humans , Aged , Aged, 80 and over , Population Dynamics , Mortality , Mobility Limitation , Health of the Elderly , Machine Learning
4.
Rev. saúde pública (Online) ; 55: 23, 2021. tab, graf
Article in English | LILACS, BBO | ID: biblio-1280613

ABSTRACT

ABSTRACT OBJECTIVE To predict the risk of absence from work due to morbidities of teachers working in early childhood education in the municipal public schools, using machine learning algorithms. METHODS This is a cross-sectional study using secondary, public and anonymous data from the Relação Anual de Informações Sociais, selecting early childhood education teachers who worked in the municipal public schools of the state of São Paulo between 2014 and 2018 (n = 174,294). Data on the average number of students per class and number of inhabitants in the municipality were also linked. The data were separated into training and testing, using records from 2014 to 2016 (n = 103,357) to train five predictive models, and data from 2017 to 2018 (n = 70,937) to test their performance in new data. The predictive performance of the algorithms was evaluated using the value of the area under the ROC curve (AUROC). RESULTS All five algorithms tested showed an area under the curve above 0.76. The algorithm with the best predictive performance (artificial neural networks) achieved 0.79 of area under the curve, with accuracy of 71.52%, sensitivity of 72.86%, specificity of 70.52%, and kappa of 0.427 in the test data. CONCLUSION It is possible to predict cases of sickness absence in teachers of public schools with machine learning using public data. The best algorithm showed a better result of the area under the curve when compared with the reference model (logistic regression). The algorithms can contribute to more assertive predictions in the public health and worker health areas, allowing to monitor and help prevent the absence of these workers due to morbidity.


RESUMO OBJETIVO Predizer o risco de ausência laboral decorrente de morbidades dos docentes que atuam na educação infantil na rede pública municipal, com o uso de algoritmos de machine learning. MÉTODOS Trata-se de um estudo transversal utilizando dados secundários, públicos e anônimos da Relação Anual de Informações Sociais, selecionando professores da educação infantil que atuaram na rede pública municipal do estado de São Paulo entre 2014 e 2018 (n = 174.294). Foram também vinculados dados da média de alunos por turma e número de habitantes no município. Os dados foram separados em treinamento e teste, utilizando os registros de 2014 a 2016 (n = 103.357) para treinar cinco modelos preditivos e os dados de 2017 a 2018 (n = 70.937) para testar seus desempenhos em dados novos. A performance preditiva dos algoritmos foi avaliada por meio do valor da área abaixo da curva ROC (AUROC). RESULTADOS Todos os cinco algoritmos testados apresentaram área abaixo da curva acima de 0,76. O algoritmo com melhor performance preditiva (redes neurais artificiais) obteve 0,79 de área abaixo da curva, com acurácia de 71,52%, sensibilidade de 72,86%, especificidade de 70,52% e kappa de 0,427 nos dados de teste. CONCLUSÃO É possível predizer casos de afastamentos por morbidade em docentes da rede pública com machine learning usando dados públicos. O melhor algoritmo apresentou melhor resultado da área abaixo da curva quando comparado ao modelo de referência (regressão logística). Os algoritmos podem contribuir para predições mais assertivas na área da saúde pública e da saúde do trabalhador, permitindo acompanhar e ajudar a prevenir afastamentos por morbidade desses trabalhadores.


Subject(s)
Humans , Child, Preschool , Absenteeism , Machine Learning , Schools , Brazil , Cross-Sectional Studies , ROC Curve
5.
Braz. arch. biol. technol ; 64(spe): e21210217, 2021. tab, graf
Article in English | LILACS | ID: biblio-1285562

ABSTRACT

Abstract Robotic Process Automation (RPA) is one of the several important techniques currently available for companies in search of performance improvement. The step forward in RPA is its association with Artificial Intelligence for more skilled robots. This scenario is not different in Power Distribution Utilities, in which a multitude of complex processes must be executed over different data sources. Making such situation even more complex, these processes are frequently regulated and subject to audit by external bodies. However, an old question remains: what should be robotized and what should be done by humans? This paper aims at partially answering the question in the context of data analysis tasks used for making decisions in complex processes. The research development is conducted based on an Artificial Intelligence methodology incorporated into one software robot (RPA) which acquires data automatically, treats and analyzes these data, helping the human professional take decisions in the process. It is applied to a real case process that is important for validating the research. Four approaches are tested in the data analysis, but only two are really used. The robot analyzes a series of information from an energy consumption meter. The detection of possible behavior deviations in the meter data is made by comparison with its data series. The robot is capable of prioritizing the detected occurrences in the energy consumption data, indicating to the human operator the most critical situations that require attention. The association of Artificial Intelligence and RPA is viable and can really apport important benefits to the company and teams, valuing human work and bringing more efficiency to the processes.


Subject(s)
Robotics/methods , Artificial Intelligence , Energy Supply , Energy Consumption , Machine Learning
6.
Braz. oral res. (Online) ; 35: e094, 2021. graf
Article in English | LILACS, BBO | ID: biblio-1285723

ABSTRACT

Abstract Artificial intelligence (AI) is a general term used to describe the development of computer systems which can perform tasks that normally require human cognition. Machine learning (ML) is one subfield of AI, where computers learn rules from data, capturing its intrinsic statistical patterns and structures. Neural networks (NNs) have been increasingly employed for ML complex data. The application of multilayered NN is referred to as "deep learning", which has been recently investigated in dentistry. Convolutional neural networks (CNNs) are mainly used for processing large and complex imagery data, as they are able to extract image features like edges, corners, shapes, and macroscopic patterns using layers of filters. CNN algorithms allow to perform tasks like image classification, object detection and segmentation. The literature involving AI in dentistry has increased rapidly, so a methodological guidance for designing, conducting and reporting studies must be rigorously followed, including the improvement of datasets. The limited interaction between the dental field and the technical disciplines, however, remains a hurdle for applicable dental AI. Similarly, dental users must understand why and how AI applications work and decide to appraise their decisions critically. Generalizable and robust AI applications will eventually prove helpful for clinicians and patients alike.


Subject(s)
Humans , Artificial Intelligence , Deep Learning , Neural Networks, Computer , Dentistry , Machine Learning
7.
Braz. arch. biol. technol ; 64: e21200483, 2021. tab, graf
Article in English | LILACS | ID: biblio-1345495

ABSTRACT

Abstract Agriculture, the backbone of every country, has been an emerging field of research, particularly in the recent past. The soil type and environment are critical factors that drive agriculture, especially in terms of crop prediction. To determine which crops grow best in certain types of soil and environment, the characteristics of the latter are to be ascertained. In the past, farmers picked suitable crops for cultivation, based on first-hand experience. Today, however, identifying appropriate crops for particular areas has become a difficult proposition. The application of machine learning techniques to agriculture is an emerging field of research that helps predicts crops for easy cultivation and improved productivity. In this work, a comparative analysis is undertaken using several classifiers like the k-Nearest Neighbor (kNN), Naïve Bayes (NB), Decision Tree (DT), Support Vector Machines (SVM), Random Forests (RF) and Bagging to help suggest the most suitable cultivable crop(s), based on soil and environmental characteristics, for a specific piece of land. The algorithms are trained with the training data and subsequently tested with the soil and climate-based test dataset. The results of all the approaches are evaluated to identify the best classification techniques. Experimental results show that the bagging method outclasses others with respect to all performance metrics.


Subject(s)
Crops, Agricultural , Agriculture , Environment , Machine Learning
8.
Braz. arch. biol. technol ; 64: e21200758, 2021. tab, graf
Article in English | LILACS | ID: biblio-1339312

ABSTRACT

Abstract Infertility is becoming a growing issue in almost all countries. Assisted Reproductive Technologies (ART) are recent development in treating infertility that give hope to the infertile couples. However, the pregnancy rates achieved with the aid of ART is considerably low, as success in ART is not only based on the treatment but also on many other controllable and uncontrollable biological, social, and environmental features. High expenditures and painful process of ART cycles are the two major barriers for opting for ART. Moreover, ART treatments are not covered by any health insurance schemes. Computational prediction models could be used to improve the success rate by predicting the treatment outcome, before the start of an ART cycle. This may suggest the couples and the doctors to decide on the next course of action i.e. either to opt for ART or opt for correcting determinants or quit the ART. With the intension to improve the success rate of ART by providing decision support system to the physicians as well to the patients before entering into the treatment this research work proposes a dynamic model for ART outcome prediction using Machine Learning (ML) techniques. The proposed dynamic model is partially implemented with the help of an ensemble of heterogeneous incremental classifier and its performance is compared with state-of-art classifiers such as Naïve Bayes (NB), Random Forest (RF), K-star etc.,using ART dataset. Performance of the model is evaluated with various metrics such as accuracy, Precision Recall Curve (PRC), Receiver Operating Characteristic (ROC), F-Measure etc., However, ROC cure area is taken as the chief metric. Evaluation results shows that the model achieves the performance with the ROC area value of 94.1 %.


Subject(s)
Reproductive Techniques, Assisted/instrumentation , Machine Learning/trends , Forecasting , Infertility/therapy
9.
São Paulo; s.n; 2021. 176 p.
Thesis in Portuguese | LILACS | ID: biblio-1178443

ABSTRACT

Introdução: A avaliação do consumo alimentar permite gerar conhecimento sobre a alimentação de indivíduos e populações, além de identificar os determinantes e tendências no consumo. Com ela é possível planejar ações, orientar serviços e implementar políticas públicas de saúde adequadas as necessidades da população. Com o apoio da tecnologia é possível automatizar algumas etapas do processo de análise de dados, com redução do tempo e recursos necessários, especialmente em grandes grupos. Entretanto, em países como o Brasil, ainda são escassas as aplicações de algoritmos de machine learning na avaliação da dieta. Objetivo: Aplicar algoritmos de machine learning na avaliação do consumo alimentar de servidores públicos em um grande estudo brasileiro. Métodos: Este estudo analisou transversalmente os dados da linha de base do Estudo Longitudinal de Saúde do Adulto (ELSA-Brasil). A partir destes dados, para explorar e classificar padrões alimentares, foi utilizado o algoritmo de cluster - K-Means. Na sequência, quatro algoritmos preditivos - Support Vector Machines (SVM), Decision Trees (DT), Naïve Bayes (NB), K-Nearest Neighbours (Knn) - foram aplicados incluindo variáveis demográficas, socioeconômicas e clínicas para predizer padrões alimentares. Adicionalmente, Sistemas de Recomendações foram construídos com algoritmos de Filtragem Colaborativa Baseada em Usuário e Itens (UBCF / IBCF) para o aconselhamento personalizado de dieta. As análises foram realizadas com a utilização do ambiente R. Resultados: Dois padrões alimentares foram derivados na amostra. O primeiro padrão, rotulado como "Padrão Ocidental", no qual os participantes apresentaram ingestões médias superiores para cereais refinados, feijões, carnes vermelhas e processadas, leite e produtos lácteos com alto teor de gorduras e bebidas adoçadas, quando comparados aqueles incluídos no outro padrão. O segundo padrão, rotulado como "Padrão Prudente", os participantes apresentaram consumo superior de frutas, vegetais, cereais integrais, aves, peixes, leite e produtos lácteos com redução de gorduras. Para a construção dos Sistemas de Recomendações foi fixado o limite de cinco itens, por participante, para evitar recomendações extensas e inespecíficas sobre a dieta (precisão entre 90% [IBCF] e 91% [UBCF]). Conclusão: Através da aplicação de algoritmos de machine learning foi possível realizar a análise de dados sobre o consumo, predizer padrões e personalizar recomendações sobre a dieta. Com o apoio das técnicas utilizadas, é possível subsidiar profissionais na gestão e no planejamento de ações de educação alimentar e nutricional personalizadas.


Introduction: The evaluation of food consumption allows generating knowledge about the diet of individuals and populations, in addition to identifying the determinants and trends in consumption. With it is possible to plan actions, guide services and implement public health policies appropriate to the needs of the population. With the support of technology, it is possible to automate some stages of the data analysis process, reducing the time and resources needed, especially in large groups. However, in countries like Brazil, the applications of machine learning algorithms in diet assessment are still scarce. Objective: Apply machine learning algorithms in the evaluation of food consumption by public servants in a large Brazilian study. Methods: This study cross-sectionally analyzed the baseline data from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). From these data, to explore and classify dietary patterns, the cluster algorithm K-Means was used. Next, four predictive algorithms - Support Vector Machines (SVM), Decision Trees (DT), Naïve Bayes (NB), K-Nearest Neighbors (Knn) - were applied including demographic, socioeconomic and clinical variables to predict dietary patterns. Additionally, Recommendation Systems were built with User- and Items-Based Collaborative Filtering algorithms (UBCF / IBCF) for personalized diet advice. The analyzes were performed using the environment R. Results: Two dietary patterns were derived in the sample. The first pattern, labeled as "Western Pattern", in which the participants had higher average intakes for refined cereals, beans, red and processed meats, milk and dairy products with a high fat content and sweetened drinks, when compared to those included in the other pattern. The second pattern, labeled "Prudent Pattern", participants showed a higher consumption of fruits, vegetables, whole grains, poultry, fish, milk and dairy products with reduced fats. For the construction of the Recommender Systems, a limit of five items was set, per participant, to avoid extensive and unspecific recommendations on the diet (accuracy between 90% [IBCF] and 91% [UBCF]). Conclusion: Through the application of machine learning algorithms, it was possible to perform data analysis on consumption, predict patterns and personalize diet recommendations. With the support of the techniques used, it is possible to subsidize professionals in the management and planning of personalized food and nutrition education actions.


Subject(s)
Diet , Nutritional Epidemiology , Feeding Behavior , Machine Learning , Data Analysis , Cluster Analysis
10.
Article in Chinese | WPRIM | ID: wpr-879287

ABSTRACT

Lung diseases such as lung cancer and COVID-19 seriously endanger human health and life safety, so early screening and diagnosis are particularly important. computed tomography (CT) technology is one of the important ways to screen lung diseases, among which lung parenchyma segmentation based on CT images is the key step in screening lung diseases, and high-quality lung parenchyma segmentation can effectively improve the level of early diagnosis and treatment of lung diseases. Automatic, fast and accurate segmentation of lung parenchyma based on CT images can effectively compensate for the shortcomings of low efficiency and strong subjectivity of manual segmentation, and has become one of the research hotspots in this field. In this paper, the research progress in lung parenchyma segmentation is reviewed based on the related literatures published at domestic and abroad in recent years. The traditional machine learning methods and deep learning methods are compared and analyzed, and the research progress of improving the network structure of deep learning model is emphatically introduced. Some unsolved problems in lung parenchyma segmentation were discussed, and the development prospect was prospected, providing reference for researchers in related fields.


Subject(s)
COVID-19 , Humans , Lung/diagnostic imaging , Machine Learning , SARS-CoV-2 , Tomography, X-Ray Computed
11.
Article in Chinese | WPRIM | ID: wpr-879272

ABSTRACT

The peak period of cardiovascular disease (CVD) is around the time of awakening in the morning, which may be related to the surge of sympathetic activity at the end of nocturnal sleep. This paper chose 140 participants as study object, 70 of which had occurred CVD events while the rest hadn't during a two-year follow-up period. A two-layer model was proposed to investigate whether hypnopompic heart rate variability (HRV) was informative to distinguish these two types of participants. In the proposed model, the extreme gradient boosting algorithm (XGBoost) was used to construct a classifier in the first layer. By evaluating the feature importance of the classifier, those features with larger importance were fed into the second layer to construct the final classifier. Three machine learning algorithms, i.e., XGBoost, random forest and support vector machine were employed and compared in the second layer to find out which one can achieve the highest performance. The results showed that, with the analysis of hypnopompic HRV, the XGBoost+XGBoost model achieved the best performance with an accuracy of 84.3%. Compared with conventional time-domain and frequency-domain features, those features derived from nonlinear dynamic analysis were more important to the model. Especially, modified permutation entropy at scale 1 and sample entropy at scale 3 were relatively important. This study might have significance for the prevention and diagnosis of CVD, as well as for the design of CVD-risk assessment system.


Subject(s)
Algorithms , Cardiovascular Diseases , Heart Rate , Humans , Machine Learning , Sleep
12.
Article in Chinese | WPRIM | ID: wpr-879198

ABSTRACT

Drug combination is a common clinical phenomenon. However, the scientific implementation of drug combination is li-mited by the weak rational evaluation that reflects its clinical characteristics. In order to break through the limitations of existing evaluation tools, examining drug-to-drug and drug-to-target action characteristics is proposed from the physical, chemical and biological perspectives, combining clinical multicenter case resources, domestic and international drug interaction public facilities with the aim of discovering the common rules of drug combination. Machine learning technology is employed to build a system for evaluating and predicting the rationality of clinical drug combinations based on "drug characteristics-repository information-artificial intelligence" strategy, which will be debugged and validated in multi-center clinical practice, with a view to providing new ideas and technical references for the safety and efficacy of clinical drug use.


Subject(s)
Artificial Intelligence , Drug Combinations , Machine Learning , Technology
13.
Arq. neuropsiquiatr ; 78(12): 789-796, Dec. 2020. tab, graf
Article in English | LILACS | ID: biblio-1142372

ABSTRACT

ABSTRACT Introduction: Magnetic resonance imaging (MRI) is the most important tool for diagnosis and follow-up in multiple sclerosis (MS). The discrimination of relapsing-remitting MS (RRMS) from secondary progressive MS (SPMS) is clinically difficult, and developing the proposal presented in this study would contribute to the process. Objective: This study aimed to ensure the automatic classification of healthy controls, RRMS, and SPMS by using MR spectroscopy and machine learning methods. Methods: MR spectroscopy (MRS) was performed on a total of 91 participants, distributed into healthy controls (n=30), RRMS (n=36), and SPMS (n=25). Firstly, MRS metabolites were identified using signal processing techniques. Secondly, feature extraction was performed based on MRS Spectra. N-acetylaspartate (NAA) was the most significant metabolite in differentiating MS types. Lastly, binary classifications (healthy controls-RRMS and RRMS-SPMS) were carried out according to features obtained by the Support Vector Machine algorithm. Results: RRMS cases were differentiated from healthy controls with 85% accuracy, 90.91% sensitivity, and 77.78% specificity. RRMS and SPMS were classified with 83.33% accuracy, 81.81% sensitivity, and 85.71% specificity. Conclusions: A combined analysis of MRS and computer-aided diagnosis may be useful as a complementary imaging technique to determine MS types.


RESUMO Introdução: A ressonância magnética é a ferramenta mais importante para o diagnóstico e acompanhamento na EM. A transição da EM recorrente-remitente (EMRR) para a EM progressiva secundária (EMPS) é clinicamente difícil e seria importante desenvolver a proposta apresentada neste estudo a fim de contribuir com o processo. Objetivo: o objetivo deste estudo foi garantir a classificação automática de grupo controle saudável, EMRR e EMPS usando a RM com espectroscopia e métodos de aprendizado de máquina. Métodos: Os exames de RM com espectroscopia foram realizados em um total de 91 amostras com grupo controle saudável (n=30), EMRR (n=36) e EMPS (n=25). Em primeiro lugar, os metabólitos da RM com espectroscopia foram identificados usando técnicas de processamento de sinal. Em segundo lugar, a extração de recursos foi realizada a partir do MRS Spectra. O NAA foi determinado como o metabólito mais significativo na diferenciação dos tipos de MS. Por fim, as classificações binárias (Healthy Control Group-RRMS e RRMS-SPMS) foram realizadas de acordo com as características obtidas por meio do algoritmo Support Vector Machine. Resultados: Os casos de EMRR e do grupo de controle saudável foram diferenciados entre si com 85% de acerto, 90,91% de sensibilidade e 77,78% de especificidade, respectivamente. A EMRR e a EMPS foram classificadas com 83,33% de acurácia, 81,81% de sensibilidade e 85,71% de especificidade, respectivamente. Conclusões: Uma análise combinada de RM com espectroscopia e abordagem de diagnóstico auxiliado por computador pode ser útil como uma técnica de imagem complementar na determinação dos tipos de EM.


Subject(s)
Humans , Multiple Sclerosis, Chronic Progressive/diagnostic imaging , Multiple Sclerosis, Relapsing-Remitting/diagnostic imaging , Multiple Sclerosis , Magnetic Resonance Imaging , Magnetic Resonance Spectroscopy , Machine Learning
14.
RECIIS (Online) ; 14(1): 150-166, jan.-mar. 2020. ilus, tab, graf
Article in Portuguese | LILACS | ID: biblio-1087302

ABSTRACT

A internet das coisas e o aprendizado de máquina são temas emergentes na área da saúde com potencial para otimizar a área e criar um sistema de saúde inteligente em virtude do envelhecimento da população. Este artigo analisa a produção científica do período de 2009 a 2019 a respeito da internet das coisas e do aprendizado de máquina na área da saúde. Utiliza metodologia bibliométrica em 1.353 artigos recuperados na base de dados Web of Science. Constata um crescimento da produção científica sobre o tema, sendo os Estados Unidos o principal polo de pesquisa na área. Identifica os autores mais produtivos e com maior impacto, periódicos mais produtivos, colaboração entre países e palavras-chave utilizadas, bem como suas relações. Incentiva que novas pesquisas explorem as abordagens identificadas no estudo.


The internet of things and machine learning are emerging issues with the potential to optimize the health field and create an intelligent health system due to the aging population. This article analyzes the scientific production of the period from 2009 to 2019 regarding the internet of things and machine learning in the health area. It uses bibliometric methodology in 1.353 articles retrieved from the Web of Science database. It notes an increase in scientific production on the subject, the United States being the main research center in this area. It identifies the most productive and influential authors, the most productive journals, collaboration between countries and keywords used, as well as their relations. It encourages new research to explore the approaches identified in the study.


La internet de las cosas y el aprendizaje de máquinas son temas emergentes en el área de la salud con potencial para optimizar el área y crear un sistema de salud inteligente en virtud del envejecimiento de la población. Este artículo analiza la producción científica del período de 2009 hasta 2019 respecto a internet de las cosas y del aprendizaje de máquina en el área de la salud. Utiliza metodología bibliométrica en 1.353 artículos recuperados en la base de datos Web of Science. Constata un crecimiento de la producción científica sobre el tema, siendo los Estados Unidos el principal polo de investigación en el área. Identifica a los autores más productivos y con mayor impacto, periódicos más productivos, colaboración entre países y palabras clave utilizadas, así como sus relaciones. Estimula a que nuevas investigaciones exploren los enfoques identificados en el estudio.


Subject(s)
Humans , Technology , Health Systems , Artificial Intelligence , Internet , Scientific and Technical Activities , Bibliometrics , Scientific and Technical Publications , Electronic Health Records , Machine Learning
15.
Article in English | WPRIM | ID: wpr-816627

ABSTRACT

Machine learning (ML) applications have received extensive attention in endocrinology research during the last decade. This review summarizes the basic concepts of ML and certain research topics in endocrinology and metabolism where ML principles have been actively deployed. Relevant studies are discussed to provide an overview of the methodology, main findings, and limitations of ML, with the goal of stimulating insights into future research directions. Clear, testable study hypotheses stem from unmet clinical needs, and the management of data quality (beyond a focus on quantity alone), open collaboration between clinical experts and ML engineers, the development of interpretable high-performance ML models beyond the black-box nature of some algorithms, and a creative environment are the core prerequisites for the foreseeable changes expected to be brought about by ML and artificial intelligence in the field of endocrinology and metabolism, with actual improvements in clinical practice beyond hype. Of note, endocrinologists will continue to play a central role in these developments as domain experts who can properly generate, refine, analyze, and interpret data with a combination of clinical expertise and scientific rigor.


Subject(s)
Artificial Intelligence , Cooperative Behavior , Data Accuracy , Endocrinology , Machine Learning , Metabolism , Osteoporosis , Thyroid Gland
16.
Article in English | WPRIM | ID: wpr-786208

ABSTRACT

No abstract available.


Subject(s)
Machine Learning
20.
Article in Chinese | WPRIM | ID: wpr-828158

ABSTRACT

The outbreak of pneumonia caused by novel coronavirus (COVID-19) at the end of 2019 was a major public health emergency in human history. In a short period of time, Chinese medical workers have experienced the gradual understanding, evidence accumulation and clinical practice of the unknown virus. So far, National Health Commission of the People's Republic of China has issued seven trial versions of the "Guidelines for the Diagnosis and Treatment of COVID-19". However, it is difficult for clinicians and laymen to quickly and accurately distinguish the similarities and differences among the different versions and locate the key points of the new version. This paper reports a computer-aided intelligent analysis method based on machine learning, which can automatically analyze the similarities and differences of different treatment plans, present the focus of the new version to doctors, reduce the difficulty in interpreting the "diagnosis and treatment plan" for the professional, and help the general public better understand the professional knowledge of medicine. Experimental results show that this method can achieve the topic prediction and matching of the new version of the program text through unsupervised learning of the previous versions of the program topic with an accuracy of 100%. It enables the computer interpretation of "diagnosis and treatment plan" automatically and intelligently.


Subject(s)
Betacoronavirus , China , Coronavirus Infections , Diagnosis , Therapeutics , Humans , Machine Learning , Pandemics , Pneumonia, Viral , Diagnosis , Therapeutics , Practice Guidelines as Topic
SELECTION OF CITATIONS
SEARCH DETAIL