A deep learning method was developed to distinguish COVID-19 from community acquired pneumonia (CAP) and other non-pneumonic lung diseases using chest CT scans. The model, called COVNet, achieved high sensitivity and specificity in detecting COVID-19 (90% sensitivity, 96% specificity) and CAP (87% sensitivity, 92% specificity), with area under the receiver operating characteristic curve (AUC) values of 0.96 and 0.95, respectively. The study used a dataset of 4,356 chest CT exams from 3,322 patients, collected from six hospitals between August 2016 and February 2020. The model was trained on 90% of the data and tested on the remaining 10%. The results showed that COVNet could accurately detect and differentiate between COVID-19 and CAP, as well as other lung diseases. The model's performance was evaluated using statistical methods, including ANOVA and chi-square tests. The study also included examples of misclassification, highlighting the challenges in distinguishing between similar diseases. The model's ability to identify key regions of interest using Grad-CAM was demonstrated, providing insights into the decision-making process. The study concluded that a deep learning model can accurately detect and differentiate between COVID-19 and CAP using chest CT scans, offering a promising tool for early diagnosis and treatment. However, the study had limitations, including the lack of laboratory confirmation for some cases and the need for further research to improve the model's accuracy in distinguishing between different types of viral pneumonias.A deep learning method was developed to distinguish COVID-19 from community acquired pneumonia (CAP) and other non-pneumonic lung diseases using chest CT scans. The model, called COVNet, achieved high sensitivity and specificity in detecting COVID-19 (90% sensitivity, 96% specificity) and CAP (87% sensitivity, 92% specificity), with area under the receiver operating characteristic curve (AUC) values of 0.96 and 0.95, respectively. The study used a dataset of 4,356 chest CT exams from 3,322 patients, collected from six hospitals between August 2016 and February 2020. The model was trained on 90% of the data and tested on the remaining 10%. The results showed that COVNet could accurately detect and differentiate between COVID-19 and CAP, as well as other lung diseases. The model's performance was evaluated using statistical methods, including ANOVA and chi-square tests. The study also included examples of misclassification, highlighting the challenges in distinguishing between similar diseases. The model's ability to identify key regions of interest using Grad-CAM was demonstrated, providing insights into the decision-making process. The study concluded that a deep learning model can accurately detect and differentiate between COVID-19 and CAP using chest CT scans, offering a promising tool for early diagnosis and treatment. However, the study had limitations, including the lack of laboratory confirmation for some cases and the need for further research to improve the model's accuracy in distinguishing between different types of viral pneumonias.