Automatic CT Whole-Lung Segmentation in Radiomics Discrimination: Methodology and Application in Pneumonia Diagnosis and Distinguishment

Radiomics based on lesion segmentation has been widely accepted for disease diagnosis;however, it is difficult to precisely determine the boundary for pneumonia due to its diffuse characteristics. In this study, we aimed to propose an automatic radiomics method using whole-lung segmentation in pneumonia discrimination and assist clinical practitioners in fast and accurate diagnosis. In the discovery set, data from 151 participants diagnosed with type A or B influenza virus pneumonia, 63 diagnosed with coronavirus disease 2019 (COVID-19) and 50 healthy participants were collected. The three groups of data were compared in pairs. A total of 117 radiomics features were extracted from whole-lung images segmented by a four-layer U-net. We then utilized a logistic regression model to train the model and used the area under the receiver operating characteristic curve (AUC) to assess its performance. The L1 regularization term was used in feature selection, and 10-fold cross-validation was used to tune the hyperparameters. Fourteen radiomics features were selected to classify influenza pneumonia and health, and the AUC was 0.957 (95% confidential interval (CI): 0.939, 0.976) in the training set and 0.914 (95% CI: 0.866, 0.963) in the testing set. Eighteen features were selected for COVID-19 and health, and the AUC was 0.949 (95% CI: 0.926, 0.973)in the training set and 0.911 (95% CI: 0.859, 0.963) in the testing set. Twenty-eight features were selected for influenza virus pneumonia and COVID-19, and the AUC was 0.895 (95% CI: 0.870, 0.920) in the training set and 0.839 (95% CI: 0.791, 0.887) in the testing set. The results show that the automatic radiomics model based on whole lung segmentation is effective in distinguishing influenza virus pneumonia, COVID-19 and health, and may assist in the diagnosis of influenza virus pneumonia and COVID-19.

Radiomics Is Effective for Distinguishing Coronavirus Disease 2019 Pneumonia From Influenza Virus Pneumonia.

Objectives: To develop and validate a radiomics model for distinguishing coronavirus disease 2019 (COVID-19) pneumonia from influenza virus pneumonia. Materials and Methods: A radiomics model was developed on the basis of 56 patients with COVID-19 pneumonia and 90 patients with influenza virus pneumonia in this retrospective study. Radiomics features were extracted from CT images. The radiomics features were reduced by the Max-Relevance and Min-Redundancy algorithm and the least absolute shrinkage and selection operator method. The radiomics model was built using the multivariate backward stepwise logistic regression. A nomogram of the radiomics model was established, and the decision curve showed the clinical usefulness of the radiomics nomogram. Results: The radiomics features, consisting of nine selected features, were significantly different between COVID-19 pneumonia and influenza virus pneumonia in both training and validation data sets. The receiver operator characteristic curve of the radiomics model showed good discrimination in the training sample [area under the receiver operating characteristic curve (AUC), 0.909; 95% confidence interval (CI), 0.859-0.958] and in the validation sample (AUC, 0.911; 95% CI, 0.753-1.000). The nomogram was established and had good calibration. Decision curve analysis showed that the radiomics nomogram was clinically useful. Conclusions: The radiomics model has good performance for distinguishing COVID-19 pneumonia from influenza virus pneumonia and may aid in the diagnosis of COVID-19 pneumonia.

Dynamic changes on chest CT of COVID-19 patients with solitary pulmonary lesion in initial CT.

PURPOSE: To investigate the dynamic evolution of image features of COVID-19 patients appearing as a solitary lesion at initial chest CT scan. MATERIALS AND METHODS: Twenty-two COVID-19 patients with solitary pulmonary lesion from three hospitals in China were enrolled from January 18, 2020 to March 18, 2020. The clinical feature and laboratory findings at first visit, as well as characteristics and dynamic evolution of chest CT images were analyzed. Among them, the CT score evaluation was the sum of the lung involvement in five lobes (0-5 points for each lobe, with a total score ranging from 0 to 25). RESULTS: 22 COVID-19 patients (11 males and 11 females, with an average age of 40.7 ± 10.3) developed a solitary pulmonary lesion within 4 days after the onset of symptoms, the peak time of CT score was about 11 days (with a median CT score of 6), and was discharged about 19 days. The peak of CT score was positively correlated with the peak time and the discharge time (p < 0.001, r = 0.793; p < 0.001, r = 0.715). Scan-1 (first visit): 22 cases (100%) showed GGO and one lobe was involved, CT score was 1.0/1.0 (median/IQR). Scan-2 (peak): 15 cases (68%) showed crazy-paving pattern, 19 cases (86%) showed consolidation, and 2.5 lobes were involved, CT score was 6.0/12.0. Scan-3 (before discharge): ten cases (45%) showed linear opacities, none had crazy-paving pattern, and 2.5 lobes were involved, CT score was 6.0/11.0. Scan-4 (after discharge): three cases (19%) showed linear opacities and one lobe was involved, CT score was 2.0/5.0. CONCLUSION: The chest CT features are related to the course of COVID-19 disease, and dynamic chest CT scan are helpful to monitor disease progress and patients' condition. In recovered patients with COVID-19, the positive CT manifestations were found within 4 days, lung involvement peaking at approximately 11 days, and discharged at about 19 days. The patients with more severe the lung injury was, the later the peak time appeared and the longer the recovery time was. Although the lesion was resolved over time, isolation and reexamination were required after discharge.

CT Manifestations of Coronavirus Disease (COVID-19) Pneumonia and Influenza Virus Pneumonia: A Comparative Study.

OBJECTIVE. The purpose of this study was to investigate differences in CT manifestations of coronavirus disease (COVID-19) pneumonia and those of influenza virus pneumonia. MATERIALS AND METHODS. We conducted a retrospective study of 52 patients with COVID-19 pneumonia and 45 patients with influenza virus pneumonia. All patients had positive results for the respective viruses from nucleic acid testing and had complete clinical data and CT images. CT findings of pulmonary inflammation, CT score, and length of largest lesion were evaluated in all patients. Mean density, volume, and mass of lesions were further calculated using artificial intelligence software. CT findings and clinical data were evaluated. RESULTS. Between the group of patients with COVID-19 pneumonia and the group of patients with influenza virus pneumonia, the largest lesion close to the pleura (i.e., no pulmonary parenchyma between the lesion and the pleura), mucoid impaction, presence of pleural effusion, and axial distribution showed statistical difference (p < 0.05). The properties of the largest lesion, presence of ground-glass opacity, presence of consolidation, mosaic attenuation, bronchial wall thickening, centrilobular nodules, interlobular septal thickening, crazy paving pattern, air bronchogram, unilateral or bilateral distribution, and longitudinal distribution did not show significant differences (p > 0.05). In addition, no significant difference was seen in CT score, length of the largest lesion, mean density, volume, or mass of the lesions between the two groups (p > 0.05). CONCLUSION. Most lesions in patients with COVID-19 pneumonia were located in the peripheral zone and close to the pleura, whereas influenza virus pneumonia was more prone to show mucoid impaction and pleural effusion. However, differentiating between COVID-19 pneumonia and influenza virus pneumonia in clinical practice remains difficult.