Supervised Deep Learning Techniques for Image Description: A Systematic Review.

Automatic image description, also known as image captioning, aims to describe the elements included in an image and their relationships. This task involves two research fields: computer vision and natural language processing; thus, it has received much attention in computer science. In this review paper, we follow the Kitchenham review methodology to present the most relevant approaches to image description methodologies based on deep learning. We focused on works using convolutional neural networks (CNN) to extract the characteristics of images and recurrent neural networks (RNN) for automatic sentence generation. As a result, 53 research articles using the encoder-decoder approach were selected, focusing only on supervised learning. The main contributions of this systematic review are: (i) to describe the most relevant image description papers implementing an encoder-decoder approach from 2014 to 2022 and (ii) to determine the main architectures, datasets, and metrics that have been applied to image description.

Toward a machine learning model for a primary diagnosis of Guillain-Barré syndrome subtypes.

Guillain-Barré Syndrome (GBS) is a neurological disorder affecting people of any age and sex, mainly damaging the peripheral nervous system. GBS is divided into several subtypes, in which only four are the most common, demanding different treatments. Identifying the subtype is an expensive and time-consuming task. Early GBS detection is crucial to save the patient's life and not aggravate the disease. This work aims to provide a primary screening tool for GBS subtypes fast and efficiently without complementary invasive methods, based only on clinical variables prospected in consultation, taken from clinical history, and based on risk factors. We conducted experiments with four classifiers with different approaches, five different filters for feature selection, six wrappers, and One versus All (OvA) classification. For the experiments, we used a data set that includes 129 records of Mexican patients and 26 clinical representative variables. Random Forest filter obtained the best results in each classifier for the diagnosis of the four subtypes, in the same way, this filter with the SVM classifier achieved the best result (0.6840). OvA with SVM classifier reached a balanced accuracy of 0.8884 for the Miller-Fisher (MF) subtype.

A Predictive Model for Guillain-Barré Syndrome Based on Ensemble Methods.

Nowadays, Machine Learning methods have proven to be highly effective on the identification of various types of diseases, in the form of predictive models. Guillain-Barré syndrome (GBS) is a potentially fatal autoimmune neurological disorder that has barely been studied with computational techniques and few predictive models have been proposed. In a previous study, single classifiers were successfully used to build a predictive model. We believe that a predictive model is imperative to carry out adequate treatment in patients promptly. We designed three classification experiments: (1) using all four GBS subtypes, (2) One versus All (OVA), and (3) One versus One (OVO). These experiments use a real-world dataset with 129 instances and 16 relevant features. Besides, we compare five state-of-the-art ensemble methods against 15 single classifiers with 30 independent runs. Standard performance measures were used to obtain the best classifier in each experiment. Derived from the experiments, we conclude that Random Forest showed the best results in four GBS subtypes classification, no ensemble method stood out over the rest in OVA classification, and single classifiers outperformed ensemble methods in most cases in OVO classification. This study presents a novel predictive model for classification of four subtypes of Guillain-Barré syndrome. Our model identifies the best method for each classification case. We expect that our model could assist specialized physicians as a support tool and also could serve as a basis to improved models in the future.

The Latin American laws of correct nutrition: Review, unified interpretation, model and tools.

BACKGROUND: The "Laws of Correct Nutrition": the Law of Quantity, the Law of Quality, the Law of Harmony and the Law of Adequacy, provide the basis of a proper diet, i.e. one that provides the body with the energy required and nutrients it needs for daily activities and maintenance of vital functions. For several decades, these Laws have been the basis of nourishing menus designed in Latin America; however, they are stated in a colloquial language, which leads to differences in interpretation and ambiguities for non-experts and even experts in the field. METHODS: We present a review of the different interpretations of the Laws and describe a consensus. We represent concepts related to nourishing menu design employing the Unified Modeling Language (UML). We formalize the Laws using the Object Constraint Language (OCL). We design a nourishing menu for a particular user through enforcement of the Laws. RESULTS: We designed a domain model with the essential elements to plan a nourishing menu and we expressed the necessary constraints to make the model׳s behavior conform to the four Laws. We made a formal verification and validation of the model via USE (UML-based Specification Environment) and we developed a software prototype for menu design under the Laws. CONCLUSION: Diet planning is considered as an art but consideration should be given to the need for a set of strict rules to design adequate menus. Thus, we model the "Laws of Nutrition" as a formal basis and standard framework for this task.