Your search

The area of clinical decision support systems (CDSS) is facing a boost in research and development with the increasing amount of data in clinical analysis together with new tools to support patient care. This creates a vibrant and challenging environment for the medical and technical staff. This chapter presents a discussion about the challenges and trends of CDSS considering big data and patient-centered constraints. Two case studies are presented in detail. The first presents the development of a big data and AI classification system for maternal and fetal ambulatory monitoring, composed by different solutions such as the implementation of an Internet of Things sensors and devices network, a fuzzy inference system for emergency alarms, a feature extraction model based on signal processing of the fetal and maternal data, and finally a deep learning classifier with six convolutional layers achieving an F1-score of 0.89 for the case of both maternal and fetal as harmful. The system was designed to support maternal–fetal ambulatory premises in developing countries, where the demand is extremely high and the number of medical specialists is very low. The second case study considered two artificial intelligence approaches to providing efficient prediction of infections for clinical decision support during the COVID-19 pandemic in Brazil. First, LSTM recurrent neural networks were considered with the model achieving R2=0.93 and MAE=40,604.4 in average, while the best, R2=0.9939, was achieved for the time series 3. Second, an open-source framework called H2O AutoML was considered with the “stacked ensemble” approach and presented the best performance followed by XGBoost. Brazil has been one of the most challenging environments during the pandemic and where efficient predictions may be the difference in saving lives. The presentation of such different approaches (ambulatory monitoring and epidemiology data) is important to illustrate the large spectrum of AI tools to support clinical decision-making.

The gold standard to detect SARS-CoV-2 infection consider testing methods based on Polymerase Chain Reaction (PCR). Still, the time necessary to confirm patient infection can be lengthy, and the process is expensive. On the other hand, X-Ray and CT scans play a vital role in the auxiliary diagnosis process. Hence, a trusted automated technique for identifying and quantifying the infected lung regions would be advantageous. Chest X-rays are two-dimensional images of the patient’s chest and provide lung morphological information and other characteristics, like ground-glass opacities (GGO), horizontal linear opacities, or consolidations, which are characteristics of pneumonia caused by COVID-19. But before the computerized diagnostic support system can classify a medical image, a segmentation task should usually be performed to identify relevant areas to be analyzed and reduce the risk of noise and misinterpretation caused by other structures eventually present in the images. This chapter presents an AI-based system for lung segmentation in X-ray images using a U-net CNN model. The system’s performance was evaluated using metrics such as cross-entropy, dice coefficient, and Mean IoU on unseen data. Our study divided the data into training and evaluation sets using an 80/20 train-test split method. The training set was used to train the model, and the evaluation test set was used to evaluate the performance of the trained model. The results of the evaluation showed that the model achieved a Dice Similarity Coefficient (DSC) of 95%, Cross entropy of 97%, and Mean IoU of 86%.

COVID-19 is a respiratory disorder caused by CoronaVirus and SARS (SARS-CoV2). WHO declared COVID-19 a global pandemic in March 2020 and several nations’ healthcare systems were on the verge of collapsing. With that, became crucial to screen COVID-19-positive patients to maximize limited resources. NAATs and antigen tests are utilized to diagnose COVID-19 infections. NAATs reliably detect SARS-CoV-2 and seldom produce false-negative results. Because of its specificity and sensitivity, RT-PCR can be considered the gold standard for COVID-19 diagnosis. This test’s complex gear is pricey and time-consuming, using skilled specialists to collect throat or nasal mucus samples. These tests require laboratory facilities and a machine for detection and analysis. Deep learning networks have been used for feature extraction and classification of Chest CT-Scan images and as an innovative detection approach in clinical practice. Because of COVID-19 CT scans’ medical characteristics, the lesions are widely spread and display a range of local aspects. Using deep learning to diagnose directly is difficult. In COVID-19, a Transformer and Convolutional Neural Network module are presented to extract local and global information from CT images. This chapter explains transfer learning, considering VGG-16 network, in CT examinations and compares convolutional networks with Vision Transformers (ViT). Vit usage increased VGG-16 network F1-score to 0.94.

This chapter describes an AUTO-ML strategy to detect COVID on chest X-rays utilizing Transfer Learning feature extraction and the AutoML TPOT framework in order to identify lung illnesses (such as COVID or pneumonia). MobileNet is a lightweight network that uses depthwise separable convolution to deepen the network while decreasing parameters and computation. AutoML is a revolutionary concept of automated machine learning (AML) that automates the process of building an ML pipeline inside a constrained computing framework. The term “AutoML” can mean a number of different things depending on context. AutoML has risen to prominence in both the business world and the academic community thanks to the ever-increasing capabilities of modern computers. Python Optimised ML Pipeline (TPOT) is a Python-based ML tool that optimizes pipeline efficiency via genetic programming. We use TPOT builds models for extracted MobileNet network features from COVID-19 image data. The f1-score of 0.79 classifies Normal, Viral Pneumonia, and Lung Opacity.

The use of computational tools for medical image processing are promising tools to effectively detect COVID-19 as an alternative to expensive and time-consuming RT-PCR tests. For this specific task, CXR (Chest X-Ray) and CCT (Chest CT Scans) are the most common examinations to support diagnosis through radiology analysis. With these images, it is possible to support diagnosis and determine the disease’s severity stage. Computerized COVID-19 quantification and evaluation require an efficient segmentation process. Essential tasks for automatic segmentation tools are precisely identifying the lungs, lobes, bronchopulmonary segments, and infected regions or lesions. Segmented areas can provide handcrafted or self-learned diagnostic criteria for various applications. This Chapter presents different techniques applied for Chest CT Scans segmentation, considering the state of the art of UNet networks to segment COVID-19 CT scans and a segmentation experiment for network evaluation. Along 200 epochs, a dice coefficient of 0.83 was obtained.

Following the World Health Organization proclaims a pandemic due to a disease that originated in China and advances rapidly across the globe, studies to predict the behavior of epidemics have become increasingly popular, mainly related to COVID-19. The critical point of these studies is to discuss the disease's behavior and the progression of the virus's natural course. However, the prediction of the actual number of infected people has proved to be a difficult task, due to a wide range of factors, such as mass testing, social isolation, underreporting of cases, among others. Therefore, the objective of this work is to understand the behavior of COVID-19 in the state of Ceará to forecast the total number of infected people and to aid in government decisions to control the outbreak of the virus and minimize social impacts and economics caused by the pandemic. So, to understand the behavior of COVID-19, this work discusses some forecast techniques using machine learning, logistic regression, filters, and epidemiologic models. Also, this work brings a new approach to the problem, bringing together data from Ceará with those from China, generating a hybrid dataset, and providing promising results. Finally, this work still compares the different approaches and techniques presented, opening opportunities for future discussions on the topic. The study obtains predictions with R2 score of 0.99 to short-term predictions and 0.93 to long-term predictions.