Table 2.
Summary of the included studies concerning the use of AI in the field of COVID-19 clinical laboratory and radiology detection.
| Authors (Year) |
n | Diagnosis Method | Input | Model/Analysis | Objective |
|---|---|---|---|---|---|
| Alouani et al. [7] (2021) |
50,146 | Real-time PCR (RT-PCR) | Fluorescent readings | Deep convolutional neural network-based software (qPCRdeepNet) https://github.com/davidalouani/qPCRdeepNet, accessed date (18 February 2024) | Detection of false positive results and improvement of test specificity, a quality assurance tool |
| Lee et al. [8] (2022) |
5810 | Real-time PCR (RT-PCR) | Fluorescence values | Long-term short memory (LSTM) | Improvement of the speed of COVID-19 RT-PCR diagnosis |
| Özbilge et al. [9] (2022) |
560 | Real-time PCR (RT-PCR) | Amplification curves | MobileNetV2 DCNN | Rapid and reliable diagnosis |
| Villarreal-González et al. [10] (2020) |
14,230 | RT-PCR | RT-PCR curves | K-neighbor classifier, support vector machine for classification (SVC), decision tree classifier, random forest classifier (RFC) | Detecting atypical profiles in PCR curves caused by contamination or artifacts |
| Alvargonzález et al. [11] (2023) |
20,418 | rRT-PCR | Ct values | Support vector machine (SVM) and neural network (NN) | Detection of a Ct pattern that is characteristic of virus variants |
| Beduk et al. [12] (2022) |
63 | Laser-scribed graphene (LSG) sensors coupled with gold nanoparticles (AuNPs) | Electrochemical sensor data | Dense neural network (DNN) | Utilization of point-of-care device as biosensing platform for new variants |
| Tschoellitsch et al. [13] (2021) |
1357 | SARS-CoV-2 RT-PCR test and blood tests | RT-PCR and blood tests results | Random forest algorithm | Prediction of SARS-CoV-2 PCR results with routine blood tests |
| Brinati et al. [14] (2020) |
279 | Routine blood tests and COVID-19 RT-PCR tests | Blood test parameters and COVID-19 RT-PCR test results | Decision tree (DT); extremely randomized trees (ETs), k-nearest neighbor (KNN) Logistic regression (LR), naïve Bayes (NB), random forest (RF), support vector machine (SVM) |
Discrimination between SARS-CoV-2 positive and negative patients |
| Yang et al. [15] (2020) |
3,356 | Routine blood tests, COVID-19 RT-PCR tests | Blood parameters, COVID-19 RT-PCR test results | Gradient boosting decision tree (GBDT), random tree (RT), logistic regression (LR), decision tree (DT) | Diagnosis of COVID-19 using the results of routine laboratory tests |
| Abayomi-Alli et al. [16] (2022) |
279 | Routine blood tests | Hematochemical values | KNN, linear SVM, RBF SVM, random forest, decision tree, neural network (multilayer perceptron), AdaBoost, extremely randomized trees (ExtraTrees), naïve Bayes, LDA, QDA, logistic regression, passive classifier, ridge classifier, and stochastic gradient descent classifier (SGDC) | Effective detection of COVID-19 using routine laboratory blood test results |
| Rocca et al. [17] (2020) |
311 | MALDI-TOF MS AND RT-PCR | Main spectra profiles | ClinPro Tools, GA/k-nearest neighbor algorithm | Identification of biomarker patterns for COVID-19 |
| Le et al. [18] (2023) |
200 | LC/MS-MS | Mass spectra | SHapley Additive exPlanations (SHAP), gradient boosted decision trees, scikit-learn v0.23.2 for random forest, stratified k-fold cross-validation, grid search | Development of an alternative diagnostic strategy for SARS-CoV-2 diagnosis |
| Rosado et al. [19] (2021) |
550 | Multiplex serological assay, RT-PCR | IgG and IgM antibody responses, RT-qPCR results | Random forest algorithm | Development of accurate serological diagnostics |
| Nachtigall et al. [20] (2020) |
3621 | MALDI-MS, RT-PCR | Mass spectra | Decision tree, DT; k-nearest neighbors, KNN; naive Bayes, NB; random forest, RF; support vector machine with a linear kernel, SVM-L; support vector machine with a radial kernel, SVM-R) | Alternative detection of SARS-CoV-2 in nasal swabs |
| Costa et al. [21] (2022) |
360 | MALDI-TOF MS | Mass spectra | Support vector machine with linear kernel (SVM-LK), support vector machine with radial basis function kernel (SVM-RK), random forest (RF) and k-nearest neighbors (K-NN), and linear discriminant analysis (LDA) | Alternative method for detection of SARS-CoV-2 in nasal swabs |
| de Fátima Cobre et al. [22] (2022) | 192 | LC-MS | Mass spectra | PLS-DA, ANNDA, XGBoostDA, SIMCA, SVM, LREG and KNN | Prediction of COVID-19 diagnosis, severity, and fatality |
| Ikponmwoba et al. [23] (2022) |
20 | SERS | Spectra | Gaussian process classifier (GPC), k-fold cross-validation | Predictive diagnosis of COVID-19 in biological samples |