Tomato plants are highly affected by diverse diseases. A timely and accurate diagnosis plays an important role to prevent the quality of crops. Recently, deep learning (DL), specifically convolutional neural networks (CNNs), have achieved extraordinary results in many applications, including the classification of plant diseases. This work focused on fine-tuning based on the comparison of the state-of-the-art architectures: AlexNet, GoogleNet, Inception V3, Residual Network (ResNet) 18, and ResNet 50. An evaluation of the comparison was finally performed. The dataset used for the experiments is contained by nine different classes of tomato diseases and a healthy class from PlantVillage. The models were evaluated through a multiclass statistical analysis based on accuracy, precision, sensitivity, specificity, F-Score, area under the curve (AUC), and receiving operating characteristic (ROC) curve. The results present significant values obtained by the GoogleNet technique, with 99.72% of AUC and 99.12% of sensitivity. It is possible to conclude that this significantly success rate makes the GoogleNet model a useful tool for farmers in helping to identify and protect tomatoes from the diseases mentioned.
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is responsible for the coronavirus disease (COVID-19), a highly contagious infectious disease that has caused many deaths worldwide. Despite global efforts, it continues to cause great losses, and leaving multiple unknowns that we must resolve in order to face the pandemic more effectively. One of the questions that has arisen recently is what happens, after recovering from COVID-19. For this reason, the objective of this study is to identify the risk of presenting persistent symptoms in recovered from COVID-19. This case-control study was conducted in one state of Mexico. Initially the data were obtained from the participants, through a questionnaire about symptoms that they had at the moment of the interview. Initially were captured the collected data, to make a dataset. After the pre-processed using the R project tool to eliminate outliers or missing data. Obtained finally a total of 219 participants, 141 recovered and 78 controls. It was used confidence level of 90% and a margin of error of 7%. From results it was obtained that all symptoms have an associated risk in those recovered. The relative risk of the selected symptoms in the recovered patients goes from 3 to 22 times, being infinite for the case of dyspnea, due to the fact that there is no control that presents this symptom at the moment of the interview, followed by the nausea and the anosmia with a RR of 8.5. Therefore, public health strategies must be rethought, to treat or rehabilitate, avoiding chronic problems in patients recovered from COVID-19.
Preeclampsia (PE) is a pregnancy complex disease, distinguished by high blood pressure and proteinuria, diagnosed after the 20th gestation week. Depending on the values of blood pressure, urine protein concentrations, symptomatology, and onset of disease there is a wide range of phenotypes, from mild forms developing predominantly at the end of pregnancy to severe forms developing in the early stage of pregnancy. In the worst cases severe forms of PE could lead to systemic endothelial dysfunction, eclampsia, and maternal and/or fetal death. Worldwide the fetal morbidity and mortality related to PE is calculated to be around 8% of the total pregnancies. PE still being an enigma regarding its etiology and pathophysiology, in general a deficient trophoblast invasion during placentation at first stage of pregnancy, in combination with maternal conditions are accepted as a cause of endothelial dysfunction, inflammatory alterations and appearance of symptoms. Depending on the PE multifactorial origin, several in vitro, in vivo, and in silico models have been used to evaluate the PE pathophysiology as well as to identify or test biomarkers predicting, diagnosing or prognosing the syndrome. This review focuses on the most common models used for the study of PE, including those related to placental development, abnormal trophoblast invasion, uteroplacental ischemia, angiogenesis, oxygen deregulation, and immune response to maternal-fetal interactions. The advances in mathematical and computational modeling of metabolic network behavior, gene prioritization, the protein-protein interaction network, the genetics of PE, and the PE prediction/classification are discussed. Finally, the potential of these models to enable understanding of PE pathogenesis and to evaluate new preventative and therapeutic approaches in the management of PE are also highlighted. Impact statement This review is important to the field of preeclampsia (PE), because it provides a description of the principal in vitro, in vivo, and in silico models developed for the study of its principal aspects, and to test emerging therapies or biomarkers predicting the syndrome before their evaluation in clinical trials. Despite the current advance, the field still lacking of new methods and original modeling approaches that leads to new knowledge about pathophysiology. The part of in silico models described in this review has not been considered in the previous reports.
Oral health represents an essential component in the quality of life of people, being a determinant factor in general health since it may affect the risk of suffering other conditions, such as chronic diseases. Oral diseases have become one of the main public health problems, where dental caries is the condition that most affects oral health worldwide, occurring in about 90% of the global population. This condition has been considered a challenge because of its high prevalence, besides being a chronic but preventable disease which can be caused depending on the consumption of certain nutritional elements interacting simultaneously with different factors, such as socioeconomic factors. Based on this problem, an analysis of a set of 189 dietary and demographic determinants is performed in this work, in order to find the relationship between these factors and the oral situation of a set of subjects. The oral situation refers to the presence and absence/restorations of caries. The methodology is performed constructing a dense artificial neural network (ANN), as a computer-aided diagnosis tool, looking for a generalized model that allows for classifying subjects. As validation, the classification model was evaluated through a statistical analysis based on a cross validation, calculating the accuracy, loss function, receiving operating characteristic (ROC) curve and area under the curve (AUC) parameters. The results obtained were statistically significant, obtaining an accuracy ≃ 0.69 and AUC values of 0.69 and 0.75. Based on these results, it is possible to conclude that the classification model developed through the deep ANN is able to classify subjects with absence of caries from subjects with presence or restorations with high accuracy, according to their demographic and dietary factors.
Among the current challenges of the Smart City, traffic management and maintenance are of utmost importance. Road surface monitoring is currently performed by humans, but the road surface condition is one of the main indicators of road quality, and it may drastically affect fuel consumption and the safety of both drivers and pedestrians. Abnormalities in the road, such as manholes and potholes, can cause accidents when not identified by the drivers. Furthermore, human-induced abnormalities, such as speed bumps, could also cause accidents. In addition, while said obstacles ought to be signalized according to specific road regulation, they are not always correctly labeled. Therefore, we developed a novel method for the detection of road abnormalities (i.e., speed bumps). This method makes use of a gyro, an accelerometer, and a GPS sensor mounted in a car. After having the vehicle cruise through several streets, data is retrieved from the sensors. Then, using a cross-validation strategy, a genetic algorithm is used to find a logistic model that accurately detects road abnormalities. The proposed model had an accuracy of 0.9714 in a blind evaluation, with a false positive rate smaller than 0.018, and an area under the receiver operating characteristic curve of 0.9784. This methodology has the potential to detect speed bumps in quasi real-time conditions, and can be used to construct a real-time surface monitoring system.
Depression is a mental disorder characterized by recurrent sadness and loss of interest in the enjoyment of the positive aspects of life, in addition to fatigue, causing inability to perform daily activities, which leads to a loss of quality of life. To monitor depression (unipolar and bipolar patients), traditional methods rely on reports from patients; nevertheless, bias is commonly present in them. To overcome this problem, Ecological Momentary Assessment (EMA) reports have been widely used, which include data of the behavior, feelings and other types of activities recorded almost in real time through the use of portable devices and smartphones containing motion sensors. In this work a methodology was proposed to detect depressive subjects from control subjects based in the data of their motor activity, recorded by a wearable device, obtained from the “Depresjon” database. From the motor activity signals, the extraction of statistical features was carried out to subsequently feed a random forest classifier. Results show a sensitivity value of 0.867, referring that those subjects with presence of depression have a degree of 86.7% of being correctly classified, while the specificity shows a value of 0.919, referring that those subjects with absence of depression have a degree of 91.9% of being classified with a correct response, using the motor activity signal provided from the wearable device. Based on these results, it is concluded that the motor activity allows distinguishing between the two classes, providing a preliminary and automated tool to specialists for the diagnosis of depression.
Major Depression Disease has been increasing in the last few years, affecting around 7 percent of the world population, but nowadays techniques to diagnose it are outdated and inefficient. Motor activity data in the last decade is presented as a better way to diagnose, treat and monitor patients suffering from this illness, this is achieved through the use of machine learning algorithms. Disturbances in the circadian rhythm of mental illness patients increase the effectiveness of the data mining process. In this paper, a comparison of motor activity data from the night, day and full day is carried out through a data mining process using the Random Forest classifier to identified depressive and non-depressive episodes. Data from Depressjon dataset is split into three different subsets and 24 features in time and frequency domain are extracted to select the best model to be used in the classification of depression episodes. The results showed that the best dataset and model to realize the classification of depressive episodes is the night motor activity data with 99.37% of sensitivity and 99.91% of specificity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.