Reducing the Learning Domain by Using Image Processing to Diagnose COVID-19 from X-Ray Image

Over the last months, dozens of artificial intelligence (AI) solutions for COVID-19 diagnosis based on chest X-ray image analysis have been proposed. All of them with very impressive sensitivity and specificity results. However, its generalization and translation to the clinical practice are rather challenging due to the discrepancies between domain distributions when training and test data come from different sources. Consequently, applying a trained model on a new data set may have a problem with domain adaptation leading to performance degradation. This research aims to study the impact of image pre-processing on pre-trained deep learning models to reduce the learning domain. The dataset used in this research consists of 5,000 X-ray images obtained from different sources under two categories: negative and positive COVID-19 detection. We implemented transfer learning in 3 popular convolutional neural networks (CNNs), including VGG16, VGG19, and DenseNet169. We repeated the study following the same structure for original and pre-processed images. The pre-processing method is based on the Contrast Limited Adaptive Histogram Equalization (CLAHE) filter application and image registration. After evaluating the models, the CNNs that have been trained with pre-processed images obtained an accuracy score up to 1.2% better than the unprocessed ones. Furthermore, we can observe that in the 3 CNN models, the repeated misclassified images represent 40.9% (207/506) of the original image dataset with the erroneous result. In pre-processed ones, this percentage is 48.9% (249/509). In conclusion, image processing techniques can help to reduce the learning domain for deep learning applications. © 2022 The authors and IOS Press.

Potential of hazard mapping as a tool for facing COVID-19 transmission: the geo-COVID cartographic platform

The present research analyses the epidemiological bases, the methodology approach and the utility of the Geo-Covid Cartographic Platform to face COVID-19 transmission at an intra-urban scale. Geo-Covid is based on the study of the main drawbacks and limitations of the current risk maps, and the proposed hazard mapping methodology is presented as an alternative approach with a high spatial-temporal accuracy. It is based on 1) the map of neighborhood active focuses of contagion, which are classified according to several hazard indexes, 2) the map of highly-transited areas by potential asymptomatic positives cases and 3) the map of Points of Maximum Risk of contagion. In order to test the effectiveness of the proposed methodology for mapping COVID-19 hazard and risk, it has been applied to Malaga City (Spain) during several stages of the epidemic in the city (2020 and 2021). The neighborhood focus of contagion is proposed as the basic spatial unit for the epidemiological diagnosis and the implementation of mitigation and control measures. After the analysis, it has been concluded that the proposed methodology, and thus, the maps included in the Geo-Covid Cartographic Platform allow a realistic and rigorous analysis of the spatial distribution of the epidemic in real-time.

Approach to COVID-19 pandemic management in Madrid. Chronic of a year.

After more than a year of pandemic, the international medical community has changed the perception of fear to one of respect for SARS-COV-2. This has been the consequence of the integral study of all the dimensions of the disease, from viral recombinant capacity to transmissibility, diagnosis, care and prevention. This document summarizes the main strategic lines of study and approach to the pandemic in Madrid.

Image quality assessment for closed-loop computer-assisted lung ultrasound

We describe a novel, two-stage computer assistance system for lung anomaly detection using ultrasound imaging in the intensive care setting to improve operator performance and patient stratification during coronavirus pandemics. The proposed system consists of two deep-learning-based models: a quality assessment module that automates predictions of image quality, and a diagnosis assistance module that determines the likelihood-of-anomaly in ultrasound images of sufficient quality. Our two-stage strategy uses a novelty detection algorithm to address the lack of control cases available for training the quality assessment classifier. The diagnosis assistance module can then be trained with data that are deemed of sufficient quality, guaranteed by the closed-loop feedback mechanism from the quality assessment module. Using more than 25,000 ultrasound images from 37 COVID-19-positive patients scanned at two hospitals, plus 12 control cases, this study demonstrates the feasibility of using the proposed machine learning approach. We report an accuracy of 86% when classifying between sufficient and insufficient quality images by the quality assessment module. For data of sufficient quality - as determined by the quality assessment module - the mean classification accuracy, sensitivity, and specificity in detecting COVID-19-positive cases were 0.95, 0.91, and 0.97, respectively, across five holdout test data sets unseen during the training of any networks within the proposed system. Overall, the integration of the two modules yields accurate, fast, and practical acquisition guidance and diagnostic assistance for patients with suspected respiratory conditions at pointof- care. © 2021 SPIE.

Temporary hospitals in times of the COVID pandemic. An example and a practical view.

We describe the most widely used temporary hospital in Europe during the first pandemic wave, its structure, function, and achievements. Other models of care developed during the pandemic around the world were reviewed including their capacity, total bed/ICU bed ratio and time of use. We particularly analyzed the common and differential characteristics of this type of facilities. IFEMA Exhibition Center was transformed into a temporary 1,300-bed hospital, which was in continuous operation for 42 days. A total of 3,817 people were treated, generally patients with mild to moderate COVID-19, 91% of whom had pneumonia. The average length of stay was 5 to 36 days. The most frequent comorbidities were hypertension (16.5%), diabetes mellitus (9.1%), COPD (6%), asthma (4.6%), obesity (2.9%) and dementia (1.6%). A total of 113 patients (3%) were transferred to another centers for aggravation, 19 (0.5%) were admitted to ICU and 16 patients (0.4%) died. An element of great help to reducing the overload of care in large hospitals during peaks of health emergencies could be these flexible structures capable of absorbing the excess of patients. These must be safe, breaking domestic transmission and guarantee social and emotional needs of patients. The success of these structures depends on delimitation in admission criteria taking into account the proportion of patients who may require, during admission, assistance in the critical care area.