Machine learning for hospital readmission prediction in pediatric population.

BACKGROUND AND OBJECTIVE: Pediatric readmissions are a burden on patients, families, and the healthcare system. In order to identify patients at higher readmission risk, more accurate techniques, as machine learning (ML), could be a good strategy to expand the knowledge in this area. The aim of this study was to develop predictive models capable of identifying children and adolescents at high risk of potentially avoidable 30-day readmission using ML. METHODS: Retrospective cohort study was carried out with 9,080 patients under 18 years old admitted to a tertiary university hospital. Demographic, clinical, and biochemical data were collected from electronic databases. We randomly divided the dataset into training (75 %) and testing (25 %), applied downsampling, repeated cross-validation with five folds and ten repetitions, and the hyperparameter was optimized of each technique using a grid search via racing with ANOVA models. We applied six ML classification algorithms to build the predictive models, including classification and regression tree (CART), random forest (RF), gradient boosting machine (GBM), extreme gradient boosting (XGBoost), decision tree and logistic regression (LR). The area under the receiver operating curve (AUC), sensitivity, specificity, Youden's J-index and accuracy were used to evaluate the performance of each model. RESULTS: The avoidable 30-day hospital readmissions rate was 9.5 %. Some algorithms presented similar AUC, both in the dataset training and in the dataset testing, such as XGBoost, RF, GBM and CART. Considering the Youden's J-index, the algorithm that presented the best index was XGBoost with bagging imputation, with AUC of 0.814 (J-index of 0.484). Cancer diagnosis, age, red blood cells, leukocytes, red cell distribution width and sodium levels, elective admission, and multimorbidity were the most important characteristics to classify between readmission and non-readmission groups. CONCLUSION: Machine learning approaches, especially XGBoost, can predict potentially avoidable 30-day pediatric hospital readmission into tertiary assistance. If implemented in the computer hospital system, our model can help in the early and more accurate identification of patients at readmission risk, targeting health strategic interventions.

Measuring water holding capacity in pork meat images using deep learning.

Water holding capacity (WHC) plays an important role when obtaining a high-quality pork meat. This attribute is usually estimated by pressing the meat and measuring the amount of water expelled by the sample and absorbed by a filter paper. In this work, we used the Deep Learning (DL) architecture named U-Net to estimate water holding capacity (WHC) from filter paper images of pork samples obtained using the press method. We evaluated the ability of the U-Net to segment the different regions of the WHC images and, since the images are much larger than the traditional input size of the U-Net, we also evaluated its performance when we change the input size. Results show that U-Net can be used to segment the external and internal areas of the WHC images with great precision, even though the difference in the appearance of these areas is subtle.

An Artificial Intelligence-as-a-Service Architecture for deep learning model embodiment on low-cost devices: A case study of COVID-19 diagnosis.

Coronavirus Disease-2019 (COVID-19) causes Severe Acute Respiratory Syndrome-Corona Virus-2 (SARS-CoV-2) and has opened several challenges for research concerning diagnosis and treatment. Chest X-rays and computed tomography (CT) scans are effective and fast alternatives to detect and assess the damage that COVID causes to the lungs at different stages of the disease. Although the CT scan is an accurate exam, the chest X-ray is still helpful due to the cheaper, faster, lower radiation exposure, and is available in low-incoming countries. Computer-aided diagnostic systems based on Artificial Intelligence (AI) and computer vision are an alternative to extract features from X-ray images, providing an accurate COVID-19 diagnosis. However, specialized and expensive computational resources come across as challenging. Also, it needs to be better understood how low-cost devices and smartphones can hold AI models to predict diseases timely. Even using deep learning to support image-based medical diagnosis, challenges still need to be addressed once the known techniques use centralized intelligence on high-performance servers, making it difficult to embed these models in low-cost devices. This paper sheds light on these questions by proposing the Artificial Intelligence as a Service Architecture (AIaaS), a hybrid AI support operation, both centralized and distributed, with the purpose of enabling the embedding of already-trained models on low-cost devices or smartphones. We demonstrated the suitability of our architecture through a case study of COVID-19 diagnosis using a low-cost device. Among the main findings of this paper, we point out the performance evaluation of low-cost devices to handle COVID-19 predicting tasks timely and accurately and the quantitative performance evaluation of CNN models embodiment on low-cost devices.

Validation of the HOSPITAL score as predictor of 30-day potentially avoidable readmissions in pediatric hospitalized population: retrospective cohort study.

Potentially avoidable pediatric readmissions are a burden to patients and their families. Identifying patients with higher risk of readmission could help minimize hospital costs and facilitate the targeting of care interventions. HOSPITAL score is a tool developed and widely used to predict adult patient's readmissions; however its predictive capacity for pediatric readmissions has not yet been evaluated. The aim of the study was to validate the HOSPITAL score application to predict 30-day potentially avoidable readmissions in a pediatric hospitalized population. This is a retrospective cohort study with patients under 18 years old admitted to a tertiary university hospital (n = 6,344). The HOSPITAL score was estimated for each admission. Subsequently, we classified the patients as low (0-4), intermediate (5-6), and high (7-12) risk groups. In order to estimate the discrimination power, the sensitivity, specificity, and accuracy were determined by the receiver operating characteristics (ROC) and the calibration by the Hosmer-Lemeshow goodness-of-fit. The 30-day hospital readmission was 11.70% (745). The accuracy was 0.80 (CI 95%, 0.77, 0.83), with a sensitivity of 70.96% and specificity of 78.29%, and a good calibration (p = 0.34).    Conclusion: HOSPITAL score showed a good discrimination and can be used to predict 30-day potentially avoidable readmission in a large pediatric population with different medical diagnoses. Our study validates and expands the usefulness of the HOSPITAL score as a tool to predict avoidable hospital readmissions for pediatric population. What is Known: •   Pediatric readmissions burden patients, the family network, and the health system. In addition, it influences negatively child development. •   The HOSPITAL score is one of the tools developed and widely used to identify patients at high risk of hospital readmission, but its predictive capacity for pediatric readmissions has not been yet assessed. What is New: • The HOSPITAL score showed good ability to identify a risk of 30-day potentially avoidable readmission in a pediatric population in different clinical contexts and diagnoses. • Our study expands the usefulness of the HOSPITAL score as a tool for predicting hospital readmissions for children and adolescents.

A new index for quantifying the ornamentational complexity of animals with shells.

Morphological complexity reflects the biological structure of an organism and is closely linked to its associated functions and phylogenetics. In animals with shells, ornamentation is an important characteristic of morphological complexity, and it has various functions. However, because of the variations in type, shape, density, and strength of ornamentation, a universal quantitative measure of morphological complexity for shelled animals is lacking. We propose an ornamentation index (OI) derived from 3D scanning technology and a virtual model for quantifying ornamentation complexity. This index is designed to measure the extent of folding associated with ornamentation, regardless of shape and size. Ornamentation indices were measured for 15 ammonite specimens from the Permian to Cretaceous, 2 modern bivalves, 2 gastropods from the Pliocene to the present, and a modern echinoid. Compared with other measurements, such as the fractal dimension, rugosity, and surface-volume ratio, the OI displayed superiority in quantifying ornamentational complexity. The present study demonstrates that the OI is suitable for accurately characterizing and quantifying ornamentation complexity, regardless of shape and size. Therefore, the OI is potentially useful for comparing the ornamentational complexity of various organisms and can be exploited to provide further insight into the evolution of conchs. Ultimately, the OI can enhance our understanding of morphological evolution of shelled organisms, for example, whether shell ornaments simplify under ocean acidification or extinction, and how predation pressure is reflected in ornamentation complexity.

Optimizing a Deep Residual Neural Network with Genetic Algorithm for Acute Lymphoblastic Leukemia Classification.

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer worldwide, and it is characterized by the production of immature malignant cells in the bone marrow. Computer vision techniques provide automated analysis that can help specialists diagnose this disease. Microscopy image analysis is the most economical method for the initial screening of patients with ALL, but this task is subjective and time-consuming. In this study, we propose a hybrid model using a genetic algorithm (GA) and a residual convolutional neural network (CNN), ResNet-50V2, to predict ALL using microscopy images available in ALL-IDB dataset. However, accurate prediction requires suitable hyperparameters setup, and tuning these values manually still poses challenges. Hence, this paper uses GA to find the best hyperparameters that lead to the highest accuracy rate in the models. Also, we compare the performance of GA hyperparameter optimization with Random Search and Bayesian optimization methods. The results show that GA optimization improves the accuracy of the classifier, obtaining 98.46% in terms of accuracy. Additionally, our approach sheds new perspectives on identifying leukemia based on computer vision strategies, which could be an alternative for applications in a real-world scenario.

Automatic Segmentation of Bone Canals in Histological Images.

The literature provides many works that focused on cell nuclei segmentation in histological images. However, automatic segmentation of bone canals is still a less explored field. In this sense, this paper presents a method for automatic segmentation approach to assist specialists in the analysis of the bone vascular network. We evaluated the method on an image set through sensitivity, specificity and accuracy metrics and the Dice coefficient. We compared the results with other automatic segmentation methods (neighborhood valley emphasis (NVE), valley emphasis (VE) and Otsu). Results show that our approach is proved to be more efficient than comparable methods and a feasible alternative to analyze the bone vascular network.

Medical image retrieval and analysis by Markov random fields and multi-scale fractal dimension.

Many Content-based Image Retrieval (CBIR) systems and image analysis tools employ color, shape and texture (in a combined fashion or not) as attributes, or signatures, to retrieve images from databases or to perform image analysis in general. Among these attributes, texture has turned out to be the most relevant, as it allows the identification of a larger number of images of a different nature. This paper introduces a novel signature which can be used for image analysis and retrieval. It combines texture with complexity extracted from objects within the images. The approach consists of a texture segmentation step, modeled as a Markov Random Field process, followed by the estimation of the complexity of each computed region. The complexity is given by a Multi-scale Fractal Dimension. Experiments have been conducted using an MRI database in both pattern recognition and image retrieval contexts. The results show the accuracy of the proposed method in comparison with other traditional texture descriptors and also indicate how the performance changes as the level of complexity is altered.

Color texture classification using shortest paths in graphs.

Color textures are among the most important visual attributes in image analysis. This paper presents a novel method to analyze color textures by modeling a color image as a graph in two different and complementary manners (each color channel separately and the three color channels altogether) and by obtaining statistical moments from the shortest paths between specific vertices of this graph. Such an approach allows to create a set of feature vectors, which were extracted from VisTex, USPTex, and TC00013 color texture databases. The best classification results were 99.07%, 96.85%, and 91.54% (LDA with leave-one-out), 87.62%, 66.71%, and 88.06% (1NN with holdout), and 98.62%, 96.16%, and 91.34% (LDA with holdout) of success rate (percentage of samples correctly classified) for these three databases, respectively. These results prove that the proposed approach is a powerful tool for color texture analysis to be explored.