Development of a quantitative segmentation model to assess the effect of comorbidity on patients with COVID-19

Abstract

Background The coronavirus disease 2019 (COVID-19) has brought a global disaster. Quantitative lesions may provide the radiological evidence of the severity of pneumonia and further to assess the effect of comorbidity on patients with COVID-19. Methods 294 patients with COVID-19 were enrolled from February, 24, 2020 to June, 1, 2020 from six centers. Multi-task Unet network was used to segment the whole lung and lesions from chest CT images. This deep learning method was pre-trained in 650 CT images (550 in primary dataset and 100 in test dataset) with COVID-19 or community-acquired pneumonia and Dice coefficients in test dataset were calculated. 50 CT scans of 50 patients (15 with comorbidity and 35 without comorbidity) were random selected to mark lesions manually. The results will be compared with the automatic segmentation model. Eight quantitative parameters were calculated based on the segmentation results to evaluate the effect of comorbidity on patients with COVID-19. Results Quantitative segmentation model was proved to be effective and accurate with all Dice coefficients more than 0.85 and all accuracies more than 0.95. Of the 294 patients, 52 (17.7%) patients were reported having at least one comorbidity; 14 (4.8%) having more than one comorbidity. Patients with any comorbidity were older (P < 0.001), had longer incubation period (P < 0.001), were more likely to have abnormal laboratory findings (P < 0.05), and be in severity status (P < 0.001). More lesions (including larger volume of lesion, consolidation, and ground-glass opacity) were shown in patients with any comorbidity than patients without comorbidity (all P < 0.001). More lesions were found on CT images in patients with more comorbidities. The median volumes of lesion, consolidation, and ground-glass opacity in diabetes mellitus group were largest among the groups with single comorbidity that had the incidence rate of top three. Conclusions Multi-task Unet network can make quantitative CT analysis of lesions to assess the effect of comorbidity on patients with COVID-19, further to provide the radiological evidence of the severity of pneumonia. More lesions (including GGO and consolidation) were found in CT images of cases with comorbidity. The more comorbidities patients have, the more lesions CT images show.

Classification Performance for COVID Patient Prognosis from Automatic AI Segmentation—A Single-Center Study

“…An initial prospective made by Huang et al [12] on chest CT scans of patients affected by COVID-19 has shown that the examined subjects have a bilateral GGO and CS. The same medical results were also confirmed by other authors [13,14], who posed the basis for the next quantification studies allowing a better characterization of the COVID-19 features. The classification between early-stage patients and progressive phases has been thoroughly investigated [15][16][17][18], and all of them lead to the same conclusion: the main COVID-19 features can be difficult to detect in early stages of the disease, and their correct identification is strongly dependent on the radiologist's expertise.…”

The Applications of Artificial Intelligence in Chest Imaging of COVID-19 Patients: A Literature Review

“…Differently from the previous study, Zhang et al included hypertension-the most common, COPD and cerebrovascular diseases in addition to DM, already described as major risk factors [126]: authors found a significant correlation with age, length of incubation period, abnormal laboratory findings and severity status. Moreover, a higher number of comorbidities resulted in a higher number of CT lesions, especially in presence of DM as main risk factors for lung volume involvement [127].…”

Artificial intelligence for reducing the radiation burden of medical imaging for the diagnosis of coronavirus disease

“…AI has demonstrated its potential in assisting the diagnosis of COVID-19, and its corresponding application in computer-aided diagnosis has successfully improved diagnostic efficiency in many applications. The current literature mainly focuses on revealing AI’s potentiality in assisted diagnosis [ 44 – 48 ], and some in segmentation [ 49 , 50 ]. More specifically, to address the difficulty of emerging false positives in CT findings, Lin et al trained a 2D convolutional neural networking (CNN) method to differentiate COVID-19 from community-acquired pneumonia and non-pneumonia [ 51 ].…”