Machine learning models for predicting diagnosis or prognosis of COVID-19: A systematic review

Clinical decision support systems (CDSS) in assistance to COVID-19 diagnosis: A scoping review on types and evaluation methods

Background and Aims

Due to the COVID-19 pandemic, a precise and reliable diagnosis of this disease is critical. The use of clinical decision support systems (CDSS) can help facilitate the diagnosis of COVID-19. This scoping review aimed to investigate the role of CDSS in diagnosing COVID-19.

Methods

We searched four databases (Web of Science, PubMed, Scopus, and Embase) using three groups of keywords related to CDSS, COVID-19, and diagnosis. To collect data from studies, we utilized a data extraction form that consisted of eight fields. Three researchers selected relevant articles and extracted data using a data collection form. To resolve any disagreements, we consulted with a fourth researcher.

Results

A search of the databases retrieved 2199 articles, of which 68 were included in this review after removing duplicates and irrelevant articles. The studies used nonknowledge-based CDSS (n = 52) and knowledge-based CDSS (n = 16). Convolutional Neural Networks (CNN) (n = 33) and Support Vector Machine (SVM) (n = 8) were employed to design the CDSS in most of the studies. Accuracy (n = 43) and sensitivity (n = 35) were the most common metrics for evaluating CDSS.

Conclusion

CDSS for COVID-19 diagnosis have been developed mainly through machine learning (ML) methods. The greater use of these techniques can be due to their availability of public data sets about chest imaging. Although these studies indicate high accuracy for CDSS based on ML, their novelty and data set biases raise questions about replacing these systems as clinician assistants in decision-making. Further studies are needed to improve and compare the robustness and reliability of nonknowledge-based and knowledge-based CDSS in COVID-19 diagnosis.

Prognosing the risk of COVID-19 death through a machine learning-based routine blood panel: A retrospective study in Brazil

Background:

Despite an extensive network of primary care availability, Brazil has suffered profoundly during the COVID-19 pandemic, experiencing the greatest sanitary collapse in its history. Thus, it is important to understand phenotype risk factors for SARS-CoV-2 infection severity in the Brazilian population in order to provide novel insights into the pathogenesis of the disease.

Objective:

This study proposes to predict the risk of COVID-19 death through machine learning, using blood biomarkers data from patients admitted to two large hospitals in Brazil.

Methods:

We retrospectively collected blood biomarkers data in a 24-h time window from 6,979 patients with COVID-19 confirmed by positive RT-PCR admitted to two large hospitals in Brazil, of whom 291 (4.2%) died and 6,688 (95.8%) were discharged. We then developed a large-scale exploration of risk models to predict the probability of COVID-19 severity, finally choosing the best performing model regarding the average AUROC. To improve generalizability, for each model five different testing scenarios were conducted, including two external validations.

Results:

We developed a machine learning-based panel composed of parameters extracted from the complete blood count (lymphocytes, MCV, platelets and RDW), in addition to C-Reactive Protein, which yielded an average AUROC of 0.91 ± 0.01 to predict death by COVID-19 confirmed by positive RT-PCR within a 24-h window.

Conclusion:

Our study suggests that routine laboratory variables could be useful to identify COVID-19 patients under higher risk of death using machine learning. Further studies are needed for validating the model in other populations and contexts, since the natural history of SARS-CoV-2 infection and its consequences on the hematopoietic system and other organs is still quite recent.

Stress and Burnout Related to Electronic Health Record Use among Healthcare Providers during the COVID-19 Pandemic in Saudi Arabia: A Preliminary National Randomized Survey

Healthcare providers' burnout may potentially have a negative impact on patient care. The use of the electronic health record (EHR) increases the burden for healthcare providers (HCPs), particularly during the coronavirus disease-2019 (COVID-19) pandemic. This study assessed the stress and burnout related to the use of EHRs and health information technology (HIT) tools among HCPs during COVID-19 in Saudi Arabia. We used a self-developed survey tool. It consisted of five sections; demographics and professional data, experience using EHR, effects of EHR use, use of EHR and technology tools during COVID-19, and health and wellbeing. The survey link was emailed to a random sample of HCPs registered with a national scientific regulatory body. Univariate, bivariate, and multivariate analyses were performed to measure the association between burnout and study variables. A total of 182 participants completed the survey. 50.5% of participants reported a presence of HIT-related stress, and 40.1% reported a presence of burnout. The variables independently associated with burnout were providing tertiary level of care, working with COVID-19 suspected cases, dissatisfaction with EHRs, and agreement with the statement that using EHRs added frustration to the workday. Further research that explores possible solutions is warranted to minimize burnout among HCPs, especially during infectious outbreaks.

Prediction of COVID Criticality Score with Laboratory, Clinical and CT Images using Hybrid Regression Models

BACKGROUND AND OBJECTIVE

Rapid and precise diagnosis of COVID-19 is very critical in hotspot regions. The main aim of this proposed work is to investigate the baseline, laboratory and CT features of COVID-19 affected patients of two groups (Early and Critical stages). The detection model for COVID-19 is built depending upon the manifestations that define the severity of the disease.

METHODS

The CT scan images are fed into the various deep learning, machine learning and hybrid learning models to mine the necessary features and predict CT Score. The predicted CT score along with other clinical, laboratory and CT scan image features are then passed to train the various Regression models for predicting the COVID Criticality (CC) Score. These baseline, laboratory and CT features of COVID-19 are reduced using Statistical analysis and Univariate logistic regression analysis.

RESULTS

When analysing the prediction of CT scores using images alone, AlexNet+Lasso yields better outcome with regression score of 0.9643 and RMSE of 0.0023 when compared with Decision tree (RMSE of 0.0034; Regression score of 0.9578) and GRU (RMSE of 0.1253; regression score of 0.9323). When analysing the prediction of CC scores using CT scores and other baseline, laboratory and CT features, VGG-16+Linear Regression yields better results with regression score of 0.9911 and RMSE of 0.0002 when compared with Linear SVR (RMSE of 0.0006; Regression score of 0.9911) and LSTM (RMSE of 0.0005; Regression score of 0.9877). The correlation analysis is performed to identify the significance of utilizing other features in prediction of CC Score. The correlation coefficient of CT scores with actual value is 0.93 and 0.92 for Early stage group and Critical stage group respectively. The correlation coefficient of CC scores with actual value is 0.96 for Early stage group and 0.95 for Critical stage group.The classification of COVID-19 patients are carried out with the help of predicted CC Scores.

CONCLUSIONS

This proposed work is carried out in the motive of helping radiologists in faster categorization of COVID patients as Early or Severe staged using CC Scores. The automated prediction of COVID Criticality Score using our diagnostic model can help radiologists and physicians save time for carrying out further treatment and procedures.