Prediction of disease progression in patients with COVID-19 by artificial intelligence assisted lesion quantification

Quantitative evaluation of CT scan images to determinate the prognosis of COVID-19 patient using deep learning

The purpose of this research is to evaluate the accuracy of AI-assisted quantification in comparison to conventional CT parameters reviewed by a radiologist in predicting the severity, progression, and clinical outcome of disease. The current study is a cross-sectional study that was conducted on patients with the diagnosis of COVID-19 and underwent a pulmonary CT scan between August 23th, 2021 to December 21th, 2022. The initial CT scan on admission was used for imaging analysis. The presence of ground glass opacity (GGO), and consolidation were visually evaluated. CT severity score was calculated according to a semi-quantitative method. In addition, AI based quantification of GGO and consolidation volume were also performed. 291 patients (mean age: 64.7 ± 7; 129 males) were included. GGO + consolidation was more frequently revealed in progress-to-severe group whereas pure GGO was more likely to be found in non-severe group. Compared to non-severe group, patients in progress-to-severe group had larger GGO volume percentage (40.6%± 11.9%versus 21.7%± 8.8%, p ˂0.001) as well as consolidation volume percentage (4.8% ± 2% versus 1.9% ± 1%, p < 0.001). Among imaging parameters, consolidation volume percentage and the largest area under curve (AUC) in discriminating non-severe from progress-to-severe group (AUC = 0.91, p < 0.001). According to multivariate regression, consolidation volume was the strongest predictor for disease progression. In conclusion, the consolidation volume measured on the initial chest CT was the most accurate predictor of disease progression, and a larger consolidation volume was associated with a poor clinical outcome. In patients with COVID-19, AI-assisted lesion quantification was useful for risk stratification and prognosis evaluation.

Tools and factors predictive of the severity of COVID-19

The outbreak of the novel coronavirus infection caused worldwide confusion. The problem with this infection is that it causes severe illness in some patients, resulting in a high rate of death if appropriate treatment is not given. If patients with severe illness that requires treatment are appropriately identified, treatment can be focused on these patients. However, in the early days of the COVID-19 outbreak, the inability to predict and diagnose the disease led to hospitals being overwhelmed. Therefore, various methods for the diagnosis of severe disease were developed early on, and various methods are still being investigated to predict high-risk patients. The currently available prediction methods are divided into those that predict the onset of severe disease and those used to determine the severity of the disease. Specifically, the main methods include genetic factors, serum humoral factors, laboratory tests, and diagnostic imaging. Since each of these factors has different features, using them in combination is likely to be advantageous.

Computed tomography severity score as a predictor of disease severity and mortality in COVID-19 patients: A systematic review and meta-analysis

BACKGROUND

Prediction of outcomes in severe COVID-19 patients using chest computed tomography severity score (CTSS) may enable more effective clinical management and early, timely ICU admission. We conducted a systematic review and meta-analysis to determine the predictive accuracy of the CTSS for disease severity and mortality in severe COVID-19 subjects.

METHODS

The electronic databases PubMed, Google Scholar, Web of Science, and the Cochrane Library were searched to find eligible studies that investigated the impact of CTSS on disease severity and mortality in COVID-19 patients between 7 January 2020 and 15 June 2021. Two independent authors looked into the risk of bias using the Quality in Prognosis Studies (QUIPS) tool.

RESULTS

Seventeen studies involving 2788 patients reported the predictive value of CTSS for disease severity. The pooled sensitivity, specificity, and summary area under the curve (sAUC) of CTSS were 0.85 (95% CI 0.78-0.90, I² =83), 0.86 (95% CI 0.76-0.92, I² =96) and 0.91 (95% CI 0.89-0.94), respectively. Six studies involving 1403 patients reported the predictive values of CTSS for COVID-19 mortality. The pooled sensitivity, specificity, and sAUC of CTSS were 0.77 (95% CI 0.69-0.83, I² = 41), 0.79 (95% CI 0.72-0.85, I² = 88), and 0.84 (95% CI 0.81-0.87), respectively.

DISCUSSION

Early prediction of prognosis is needed to deliver the better care to patients and stratify them as soon as possible. Because different CTSS thresholds have been reported in various studies, clinicians are still determining whether CTSS thresholds should be used to define disease severity and predict prognosis.

CONCLUSION

Early prediction of prognosis is needed to deliver optimal care and timely stratification of patients.  CTSS has strong discriminating power for the prediction of disease severity and mortality in patients with COVID-19.

Increased lactate dehydrogenase reflects the progression of COVID-19 pneumonia on chest computed tomography and predicts subsequent severe disease

Chest computed tomography (CT) is effective for assessing the severity of coronavirus disease 2019 (COVID-19). However, the clinical factors reflecting the disease progression of COVID-19 pneumonia on chest CT and predicting a subsequent exacerbation remain controversial. We conducted a retrospective cohort study of 450 COVID-19 patients. We used an automated image processing tool to quantify the COVID-19 pneumonia lesion extent on chest CT at admission. The factors associated with the lesion extent were estimated by a multiple regression analysis. After adjusting for background factors by propensity score matching, we conducted a multivariate Cox proportional hazards analysis to identify factors associated with severe disease after admission. The multiple regression analysis identified, body-mass index (BMI), lactate dehydrogenase (LDH), C-reactive protein (CRP), and albumin as continuous variables associated with the lesion extent on chest CT. The standardized partial regression coefficients for them were 1.76, 2.42, 1.54, and 0.71. The multivariate Cox proportional hazards analysis identified LDH (hazard ratio, 1.003; 95% confidence interval, 1.001-1.005) as a factor independently associated with the development of severe COVID-19 pneumonia. Increased serum LDH at admission may be useful in real-world clinical practice for the simple screening of COVID-19 patients at high risk of developing subsequent severe disease.

Serum gasdermin D levels are associated with the chest computed tomography findings and severity of COVID-19

BACKGROUND

The role of programmed cell death, especially pyroptosis and apoptosis, in unfavorable immune responses in COVID-19 remains to be elucidated.

METHODS

We conducted a cross-sectional analysis to investigate the association between the serum gasdermin D (GSDMD) levels, a pyroptotic marker, and caspase-cleaved cytokeratin 18 fragment (M30), an apoptotic marker, and the clinical status and abnormal chest computed tomography (CT) findings in patients with COVID-19.

RESULTS

In this study, 46 patients diagnosed with COVID-19 were divided into the following three groups according to the disease severity: mild to moderate group (n = 10), severe group (n = 14), and critical group (n = 22). The serum GSDMD levels were higher in the critical group than in the mild to moderate group (P = 0.016). In contrast, serum M30 levels were lower in the critical group than in the severe group (P = 0.048). Patients who required mechanical ventilation or died had higher serum GSDMD levels than those who did not (P = 0.007). Area of consolidation only and of ground glass opacity plus consolidation positively correlated with serum GSDMD levels (r = 0.56, P < 0.001 and r = 0.53, P < 0.001, respectively).

CONCLUSION

Higher serum GSDMD levels are associated with critical respiratory status and the consolidation area on chest CT in patients with COVID-19, suggesting that excessive activation of pyroptosis may affect the clinical manifestations in patients with COVID-19.

Leukocyte glucose index as a novel biomarker for COVID-19 severity

The severity of coronavirus disease 2019 (COVID-19) quickly progresses with unfavorable outcomes due to the host immune response and metabolism alteration. Hence, we hypothesized that leukocyte glucose index (LGI) is a biomarker for severe COVID-19. This study involved 109 patients and the usefulness of LGI was evaluated and compared with other risk factors to predict COVID 19 severity. LGI was identified as an independent risk factor (odds ratio [OR] = 1.727, 95% confidence interval [CI]: 1.026-3.048, P = 0.041), with an area under the curve (AUC) of 0.749 (95% CI: 0.642-0.857, P < 0.0001). Interestingly, LGI was a potential risk factor (OR = 2.694, 95% CI: 1.575-5.283, Pcorrected < 0.05) for severe COVID-19 in female but not in male patients. In addition, LGI proved to be a strong predictor of the severity in patients with diabetes (AUC = 0.915 (95% CI: 0.830-1), sensitivity = 0.833, and specificity = 0.931). The AUC of LGI, together with the respiratory rate (LGI + RR), showed a considerable improvement (AUC = 0.894, 95% CI: 0.835-0.954) compared to the other biochemical and respiratory parameters analyzed. Together, these findings indicate that LGI could potentially be used as a biomarker of severity in COVID-19 patients.

Application of artificial intelligence in COVID-19 medical area: a systematic review

Background

Coronavirus disease 2019 (COVID-19) has caused a large-scale global epidemic, impacting international politics and the economy. At present, there is no particularly effective medicine and treatment plan. Therefore, it is urgent and significant to find new technologies to diagnose early, isolate early, and treat early. Multimodal data drove artificial intelligence (AI) can potentially be the option. During the COVID-19 Pandemic, AI provided cutting-edge applications in disease, medicine, treatment, and target recognition. This paper reviewed the literature on the intersection of AI and medicine to analyze and compare different AI model applications in the COVID-19 Pandemic, evaluate their effectiveness, show their advantages and differences, and introduce the main models and their characteristics.

Methods

We searched PubMed, arXiv, medRxiv, and Google Scholar through February 2020 to identify studies on AI applications in the medical areas for the COVID-19 Pandemic.

Results

We summarize the main AI applications in six areas: (I) epidemiology, (II) diagnosis, (III) progression, (IV) treatment, (V) psychological health impact, and (VI) data security. The ongoing development in AI has significantly improved prediction, contact tracing, screening, diagnosis, treatment, medication, and vaccine development for the COVID-19 Pandemic and reducing human intervention in medical practice.

Discussion

This paper provides strong advice for using AI-based auxiliary tools for related applications of human diseases. We also discuss the clinicians’ role in the further development of AI. They and AI researchers can integrate AI technology with current clinical processes and information systems into applications. In the future, AI personnel and medical workers will further cooperate closely.

Pandemic strategies with computational and structural biology against COVID-19: A retrospective

The emergence of the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), which is the etiologic agent of the coronavirus disease 2019 (COVID-19) pandemic, has dominated all aspects of life since of 2020. Research studies on the virus and exploration of therapeutic and preventive strategies has been moving at rapid rates to control the pandemic. In the field of bioinformatics or computational and structural biology, recent research strategies have used multiple disciplines to compile large datasets to uncover statistical correlations and significance, visualize and model proteins, perform molecular dynamics simulations, and employ the help of artificial intelligence and machine learning to harness computational processing power to further the research on COVID-19, including drug screening, drug design, vaccine development, prognosis prediction, and outbreak prediction. These recent developments should help us better understand the viral disease and develop the much-needed therapies and strategies for the management of COVID-19.

A meta-analysis on the risk factors adjusted association between cardiovascular disease and COVID-19 severity

BACKGROUND

Cardiovascular disease (CVD), one of the most common comorbidities of coronavirus disease 2019 (COVID-19), has been suspected to be associated with adverse outcomes in COVID-19 patients, but their correlation remains controversial.

METHOD

This is a quantitative meta-analysis on the basis of adjusted effect estimates. PubMed, Web of Science, MedRxiv, Scopus, Elsevier ScienceDirect, Cochrane Library and EMBASE were searched comprehensively to obtain a complete data source up to January 7, 2021. Pooled effects (hazard ratio (HR), odds ratio (OR)) and the 95% confidence intervals (CIs) were estimated to evaluate the risk of the adverse outcomes in COVID-19 patients with CVD. Heterogeneity was assessed by Cochran's Q-statistic, I2test, and meta-regression. In addition, we also provided the prediction interval, which was helpful for assessing whether the variation across studies was clinically significant. The robustness of the results was evaluated by sensitivity analysis. Publication bias was assessed by Begg's test, Egger's test, and trim-and-fill method.

RESULT

Our results revealed that COVID-19 patients with pre-existing CVD tended more to adverse outcomes on the basis of 203 eligible studies with 24,032,712 cases (pooled ORs = 1.41, 95% CIs: 1.32-1.51, prediction interval: 0.84-2.39; pooled HRs = 1.34, 95% CIs: 1.23-1.46, prediction interval: 0.82-2.21). Further subgroup analyses stratified by age, the proportion of males, study design, disease types, sample size, region and disease outcomes also showed that pre-existing CVD was significantly associated with adverse outcomes among COVID-19 patients.

CONCLUSION

Our findings demonstrated that pre-existing CVD was an independent risk factor associated with adverse outcomes among COVID-19 patients.

Clinical and radiological characteristics of COVID‑19 patients without comorbidities : A single-center study

Objective

To evaluate the clinical characteristics and detailed imaging features in coronavirus disease 2019 (COVID-19) patients without comorbidities.

Material and methods

This retrospective study included laboratory-confirmed and symptomatic COVID-19 patients without comorbid diseases who were admitted to our second level hospital between March 2020 and September 2020. We assessed the clinical, biochemical and imaging diagnostic parameters on admission. The patients were classified as non-severe and progress to severe group and then the initial parameters were compared.

Results

We enrolled 135 adult COVID-19 patients, 12 progressed to severe disease during hospitalization. Compared to the non-severe group, patients who progressed to severe were older (p < 0.001) and were more likely to manifest coughing (p = 0.011) and have higher lactate dehydrogenase (LDH) levels (p = 0.011). On chest computed tomography (CT) images, multilobar (p = 0.016), peripherally (p = 0.001) distributed mixed ground glass opacities and consolidation (p < 0.001), crazy paving (p = 0.007) and higher total CT severity score (p < 0.001) were significantly associated with severe disease.

Conclusion

Knowledge of the clinical and radiological parameters associated with disease severity might be useful to guide clinical decision-making for COVID-19 patients without comorbidities.

Quantitative evaluation of COVID-19 pneumonia severity by CT pneumonia analysis algorithm using deep learning technology and blood test results

PURPOSE

To evaluate whether early chest computed tomography (CT) lesions quantified by an artificial intelligence (AI)-based commercial software and blood test values at the initial presentation can differentiate the severity of COVID-19 pneumonia.

MATERIALS AND METHODS

This retrospective study included 100 SARS-CoV-2-positive patients with mild (n = 23), moderate (n = 37) or severe (n = 40) pneumonia classified according to the Japanese guidelines. Univariate Kruskal-Wallis and multivariate ordinal logistic analyses were used to examine whether CT parameters (opacity score, volume of opacity, % opacity, volume of high opacity, % high opacity and mean HU total on CT) as well as blood test parameters [procalcitonin, estimated glomerular filtration rate (eGFR), C-reactive protein, % lymphocyte, ferritin, aspartate aminotransferase, lactate dehydrogenase, alanine aminotransferase, creatine kinase, hemoglobin A1c, prothrombin time, activated partial prothrombin time (APTT), white blood cell count and creatinine] differed by disease severity.

RESULTS

All CT parameters and all blood test parameters except procalcitonin and APPT were significantly different among mild, moderate and severe groups. By multivariate analysis, mean HU total and eGFR were two independent factors associated with severity (p < 0.0001). Cutoff values for mean HU total and eGFR were, respectively, - 801 HU and 77 ml/min/1.73 m² between mild and moderate pneumonia and - 704 HU and 53 ml/min/1.73 m² between moderate and severe pneumonia.

CONCLUSION

The mean HU total of the whole lung, determined by the AI algorithm, and eGFR reflect the severity of COVID-19 pneumonia.

The diagnostic accuracy of Artificial Intelligence-Assisted CT imaging in COVID-19 disease: A systematic review and meta-analysis

Artificial intelligence (AI) systems have become critical in support of decision-making. This systematic review summarizes all the data currently available on the AI-assisted CT-Scan prediction accuracy for COVID-19. The ISI Web of Science, Cochrane Library, PubMed, Scopus, CINAHL, Science Direct, PROSPERO, and EMBASE were systematically searched. We used the revised Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool to assess all included studies' quality and potential bias. A hierarchical receiver-operating characteristic summary (HSROC) curve and a summary receiver operating characteristic (SROC) curve have been implemented. The area under the curve (AUC) was computed to determine the diagnostic accuracy. Finally, 36 studies (a total of 39,246 image data) were selected for inclusion into the final meta-analysis. The pooled sensitivity for AI was 0.90 (95% CI, 0.90–0.91), specificity was 0.91 (95% CI, 0.90–0.92) and the AUC was 0.96 (95% CI, 0.91–0.98). For deep learning (DL) method, the pooled sensitivity was 0.90 (95% CI, 0.90–0.91), specificity was 0.88 (95% CI, 0.87–0.88) and the AUC was 0.96 (95% CI, 0.93–0.97). In case of machine learning (ML), the pooled sensitivity was 0.90 (95% CI, 0.90–0.91), specificity was 0.95 (95% CI, 0.94–0.95) and the AUC was 0.97 (95% CI, 0.96–0.99). AI in COVID-19 patients is useful in identifying symptoms of lung involvement. More prospective real-time trials are required to confirm AI's role for high and quick COVID-19 diagnosis due to the possible selection bias and retrospective existence of currently available studies.