Chest CT plays an important role in the assessment of patients with positive COVID-19 infection and their follow-up. These examinations are vital in the early detection and assessment of the COVID-19 disease course [8].
CT study of the chest plays a vital role in the evaluation of the severity of COVID-19, monitoring progression of the disease and consequent evaluation of the therapeutic efficiency [9].
At first, confirmed cases of COVID-19 by PCR underwent CT chest study; then, the obtained CT chest cuts were introduced to a software computer program using deep learning and programming to enable the software to make auto-detection of lesions when future CT cuts are processed.
In this study, the median number of lesions detected in the study population was 2 lesions ranged from 1 to 12 lesions. The most common affected site of the lesions was the right lower lung lobe (29.2%), left lower lung lobe (28.6%), followed by the left upper lung lobe (18.2%), the right upper lung lobe (14.3%) and at the last the middle lung lobe (9.7%).
In agreement with the current study results, Bao et al. [10] performed a systematic review and meta-analysis of the results obtained from published studies to provide a summary of the detection of COVID-19 by CT chest study. It was found that the most commonly affected lobe was the right lower lobe (87.21%). Also, Salahshour et al. [11] reported in their study that the most commonly affected lobe is the right lower lobe (43.3%).
According to the current study results, it was found that there was a significant strong agreement (P value < 0.001) between the radiologist and the semiquantitative CT assessment in the detection of GGO among patients with COVID-19 pneumonia. The radiologist failed to detect only 9 (5.8%) lesions in 7 patients only. The sites of false negative GGO findings detected by the radiologist in the present study were most prevalent in the left upper lobe (14.3%) followed by the left lower lobe (4.5%) and left lower lobe (4.5%).
In agreement with our results, a study conducted by Pan et al. [12] aimed to explore a novel deep learning-based quantification and compare its efficacy with the conventional semiquantitative CT scoring for the serial chest CT scans of patients with COVID-19. They found that there is a good correlation between conventional CT scoring and novel deep learning-based quantification (P < 0.001).
Shan et al. [13] in a study evaluated the overlap ratio (Dice similarity coefficient) between an automatically segmented infection region (S) and the corresponding reference region (R) provided by radiologist(s). It was found that the average Dice similarity coefficient is 91.6% ± 10.0%.
In a previous study conducted by Wu et al. [14], it was found that the proposed diagnosis model trained on multi-view images of chest CT images, based on the deep learning method, showed great potential to improve the diagnosis and reduce the heavy workload on radiologists in the initial screening of COVID-19 pneumonia.
As regards the patients who were followed up by semiquantitative CT and radiologist in this study population, it was found that the median number of lesions was 1 ranging from 1 to 8 lesions. The most common affected site of the lesions was the lower lobes (right lower lobe 44.4% with 33.3% for the left lower lobe) followed by both upper lobes (11.1% for each lobe).
According to the current study findings, it was reported that there was a significant strong agreement (P value = 0.001) between the radiologist and the semiquantitative CT assessment in the detection of GGO during follow-up among patients with COVID-19 pneumonia. The radiologist failed to detect only 1 (11.1%) lesion in one patient only.
In agreement with these results, Gieraerts et al. [15] conducted a study to compare the prognostic value of detection of lung involvement by the visual versus artificial intelligence (AI)–assisted analysis at chest CT in patients with COVID-19 pneumonia and found that there is an average agreement between them.
However, a study conducted by Li et al. [16] reported different results. Their study results showed the superiority of AI-assisted quantification over conventional CT severity scores for the prediction of disease progression. Also, Kimura et al. [17] also concluded that the AI-calculated CT severity score and total opacity percentage showed higher diagnostic accuracy.
Our study had some limitations, first of them is the limited number of patients in the study and moreover the limited number of patients who performed follow-up CT scan. That’s why we recommend to perform future studies on a larger scale of population.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Disclaimer:
This article is autogenerated using RSS feeds and has not been created or edited by OA JF.
Click here for Source link (https://www.springeropen.com/)
