Automated quantification of COVID-19 severity and progression using chest CT images

Abstract

Objective To develop and test computer software to detect, quantify, and monitor progression of pneumonia associated with COVID-19 using chest CT scans. Methods One hundred twenty chest CT scans from subjects with lung infiltrates were used for training deep learning algorithms to segment lung regions and vessels. Seventy-two serial scans from 24 COVID-19 subjects were used to develop and test algorithms to detect and quantify the presence and progression of infiltrates associated with COVID-19. The algorithm included (1) automated lung boundary and vessel segmentation, (2) registration of the lung boundary between serial scans, (3) computerized identification of the pneumonitis regions, and (4) assessment of disease progression. Agreement between radiologist manually delineated regions and computer-detected regions was assessed using the Dice coefficient. Serial scans were registered and used to generate a heatmap visualizing the change between scans. Two radiologists, using a five-point Likert scale, subjectively rated heatmap accuracy in representing progression. Results There was strong agreement between computer detection and the manual delineation of pneumonic regions with a Dice coefficient of 81% (CI 76-86%). In detecting large pneumonia regions (> 200 mm 3), the algorithm had a sensitivity of 95% (CI 94-97%) and specificity of 84% (CI 81-86%). Radiologists rated 95% (CI 72 to 99) of heatmaps at least "acceptable" for representing disease progression. Conclusion The preliminary results suggested the feasibility of using computer software to detect and quantify pneumonic regions associated with COVID-19 and to generate heatmaps that can be used to visualize and assess progression. Key Points • Both computer vision and deep learning technology were used to develop computer software to quantify the presence and progression of pneumonia associated with COVID-19 depicted on CT images. • The computer software was tested using both quantitative experiments and subjective assessment. • The computer software has the potential to assist in the detection of the pneumonic regions, monitor disease progression, and assess treatment efficacy related to COVID-19.

Deep COVID DeteCT: an international experience on COVID-19 lung detection and prognosis using chest CT

“…Unlike prior published AI works that combine lung segmentation and predictions 17,18,[21][22][23] , we report use of a simple 3D model that uses whole CT chest that might facilitate clinical translation. Furthermore, while prior studies use human visual inspection 23 , software-based segmentation for scoring disease severity 22,24,25 , we leverage learned features to conduct both supervised and unsupervised learning. Prior chest CT studies have shown characteristic COVID+ patterns, such as peripheral ground glass opacities that are often bilateral, peripheral with contiguous and multi-lobar extensions depending on disease severity [26][27][28] .…”

Practical clinical and radiological models to diagnose COVID-19 based on a multicentric teleradiological emergency chest CT cohort

“…Using a quantitative assessment of the main COVID-19 radiological features (through volumes of GGO and fibrotic alterations) did not increase sensitivity and specificity (0.68 and 0.59 with GGO, respectively, and 0.86 and 0.44 with fibrotic alterations, respectively) 31 . The AUCs of these deep learning models mostly ranged between 0.70 and 0.95, while sensitivity and specificity, when given, ranged between 0.84 and 1, and 0.25 and 0.96, respectively 15 – 20 .…”

“…In parallel, predictive models have been issued to facilitate and even automate the diagnosis of COVID-19 on chest CT with good performances and in an objective manner. Indeed, regarding deep-learning models, diagnostic performances (estimated with area under the receiver operating characteristics curves [AUC]) ranged from 0.70 to 0.95 in retrospective testing cohorts 15 – 20 . When detailed, sensitivity and specificity were 0.84–1 and 0.25–0.96, respectively.…”

Densely connected convolutional networks-based COVID-19 screening model

“…The main limitations of this model are biases in the results due to the small sample size, lack of transparency, and interpretability. Pu et al [27] developed an automatic approach to perceive and quantify the pneumonia regions of CT scans associated with COVID-19. UNet framework was used to segment lung boundary, which is further used to assess the progression of disease.…”