Multicenter Study on COVID-19 Lung Computed Tomography Segmentation with varying Glass Ground Opacities using Unseen Deep Learning Artificial Intelligence Paradigms: COVLIAS 1.0 Validation.

Variations in COVID-19 lesions such as glass ground opacities (GGO), consolidations, and crazy paving can compromise the ability of solo-deep learning (SDL) or hybrid-deep learning (HDL) artificial intelligence (AI) models in predicting automated COVID-19 lung segmentation in Computed Tomography (CT) from unseen data leading to poor clinical manifestations. As the first study of its kind, "COVLIAS 1.0-Unseen" proves two hypotheses, (i) contrast adjustment is vital for AI, and (ii) HDL is superior to SDL. In a multicenter study, 10,000 CT slices were collected from 72 Italian (ITA) patients with low-GGO, and 80 Croatian (CRO) patients with high-GGO. Hounsfield Units (HU) were automatically adjusted to train the AI models and predict from test data, leading to four combinations-two Unseen sets: (i) train-CRO:test-ITA, (ii) train-ITA:test-CRO, and two Seen sets: (iii) train-CRO:test-CRO, (iv) train-ITA:test-ITA. COVILAS used three SDL models: PSPNet, SegNet, UNet and six HDL models: VGG-PSPNet, VGG-SegNet, VGG-UNet, ResNet-PSPNet, ResNet-SegNet, and ResNet-UNet. Two trained, blinded senior radiologists conducted ground truth annotations. Five types of performance metrics were used to validate COVLIAS 1.0-Unseen which was further benchmarked against MedSeg, an open-source web-based system. After HU adjustment for DS and JI, HDL (Unseen AI) > SDL (Unseen AI) by 4% and 5%, respectively. For CC, HDL (Unseen AI) > SDL (Unseen AI) by 6%. The COVLIAS-MedSeg difference was < 5%, meeting regulatory guidelines.Unseen AI was successfully demonstrated using automated HU adjustment. HDL was found to be superior to SDL.

Screening for Myocardial Injury after Mild SARS-CoV-2 Infection with Advanced Transthoracic Echocardiography Modalities.

Although the clinical manifestations of SARS-CoV-2 viral infection affect mainly the respiratory system, cardiac complications are common and are associated with increased morbidity and mortality. While echocardiographic alterations indicating myocardial involvement are widely reported in patients hospitalized for acute COVID-19 infection, much fewer data available in non-hospitalized, mildly symptomatic COVID-19 patients. In our work, we aimed to investigate subclinical cardiac alterations characterized by parameters provided by advanced echocardiographic techniques following mild SARS-CoV-2 viral infection. A total of 86 patients (30 males, age: 39.5 ± 13.0 yrs) were assessed 59 ± 33 days after mild SARS-CoV-2 viral infection (requiring no hospital or <5 days in-hospital treatment) by advanced echocardiographic examination including 2-dimensional (2D) speckle tracking echocardiography and non-invasive myocardial work analysis, and were compared to an age-and sex-matched control group. Altogether, variables from eleven echocardiographic categories representing morphological or functional echocardiographic parameters showed statistical difference between the post-COVID patient group and the control group. The magnitude of change was subtle or mild in the case of these parameters, ranging from 1-11.7% of relative change. Among the parameters, global longitudinal strain [-20.3 (-21.1--19.0) vs. -19.1 (-20.4--17.6) %; p = 0.0007], global myocardial work index [1975 (1789-2105) vs. 1829 (1656-2057) Hgmm%; p = 0.007] and right ventricular free wall strain values (-26.6 ± 3.80 vs. -23.8 ± 4.0%; p = 0.0003) showed the most significant differences between the two groups. Subclinical cardiac alterations are present following even mild SARS-CoV-2 viral infection. These more subtle alterations are difficult to detect by routine echocardiography. Extended protocols, involving speckle-tracking echocardiography, non-invasive measurement of cardiac hemodynamics, and possibly myocardial work are necessary for detection and adequate follow-up.

Eight pruning deep learning models for low storage and high-speed COVID-19 computed tomography lung segmentation and heatmap-based lesion localization: A multicenter study using COVLIAS 2.0.

BACKGROUND: COVLIAS 1.0: an automated lung segmentation was designed for COVID-19 diagnosis. It has issues related to storage space and speed. This study shows that COVLIAS 2.0 uses pruned AI (PAI) networks for improving both storage and speed, wiliest high performance on lung segmentation and lesion localization. METHOD: ology: The proposed study uses multicenter ∼9,000 CT slices from two different nations, namely, CroMed from Croatia (80 patients, experimental data), and NovMed from Italy (72 patients, validation data). We hypothesize that by using pruning and evolutionary optimization algorithms, the size of the AI models can be reduced significantly, ensuring optimal performance. Eight different pruning techniques (i) differential evolution (DE), (ii) genetic algorithm (GA), (iii) particle swarm optimization algorithm (PSO), and (iv) whale optimization algorithm (WO) in two deep learning frameworks (i) Fully connected network (FCN) and (ii) SegNet were designed. COVLIAS 2.0 was validated using "Unseen NovMed" and benchmarked against MedSeg. Statistical tests for stability and reliability were also conducted. RESULTS: Pruning algorithms (i) FCN-DE, (ii) FCN-GA, (iii) FCN-PSO, and (iv) FCN-WO showed improvement in storage by 92.4%, 95.3%, 98.7%, and 99.8% respectively when compared against solo FCN, and (v) SegNet-DE, (vi) SegNet-GA, (vii) SegNet-PSO, and (viii) SegNet-WO showed improvement by 97.1%, 97.9%, 98.8%, and 99.2% respectively when compared against solo SegNet. AUC > 0.94 (p < 0.0001) on CroMed and > 0.86 (p < 0.0001) on NovMed data set for all eight EA model. PAI <0.25 s per image. DenseNet-121-based Grad-CAM heatmaps showed validation on glass ground opacity lesions. CONCLUSIONS: Eight PAI networks that were successfully validated are five times faster, storage efficient, and could be used in clinical settings.

The sensitivity and specificity of chest CT in the diagnosis of COVID-19.

PURPOSE: The identification of patients infected by SARS-CoV-2 is highly important to control the disease; however, the clinical presentation is often unspecific and a large portion of the patients develop mild or no symptoms at all. For this reason, there is an emphasis on evaluating diagnostic tools for screening. Chest CT scans are emerging as a useful tool in the diagnostic process of viral pneumonia cases associated with COVID-19. This review examines the sensitivity, specificity, and feasibility of chest CT in detecting COVID-19 compared with real-time polymerase chain reaction (RT-PCR). METHODS: Sensitivity and specificity of chest CT in detecting COVID-19 in its various phases was compared using RT-PCR as a gold standard. A "reverse calculation approach" was applied and treated chest CT as a hypothetical gold standard and compared RT-PCR to it point out the flaw of the standard approach. RESULTS: High sensitivity (67-100%) and relatively low specificity (25-80%) was reported for the CT scans. However, the sensitivity of RT-PCR was reported to be modest (53-88%), hence cannot serve as an appropriate ground truth. The "reverse calculation approach" showed that CT could have a higher specificity (83-100%) if we consider the modest sensitivity of the RT-PCR. CONCLUSIONS: The sensitivity and specificity of the chest CT in diagnosing COVID-19 and the radiation exposure have to be judged together. Arguments are presented that chest CT scans have added value in diagnosing COVID-19 especially in patients, who exhibit typical clinical symptoms and have negative RT-PCR results in highly infected regions. KEY POINTS: • CT scans have higher specificity if we take into account the low sensitivity of the RT-PCR. • Avoid chest CT as a sole diagnostic approach for COVID-19 infection. • Patients who had negative RT-PCR result with typical clinical symptoms in highly infected regions or with close contact of COVID-19-infected patients; the use of chest CT is warranted.