- Despite limited data, deep learning in Chest X-Ray (CXR) of patients with COVID-19 could increase the diagnostic abilities of physicians at point of care. It could also highlight subtle abnormalities missed by physicians and triage patients for CT.
- With limited test data, deep learning can accurately detect COVID-19 and differentiate it from community acquired pneumonia and other non-pneumonic lung diseases. However, there exists an overlap in chest CT findings of viral pneumonias with other chest diseases. This puts an emphasis on the importance of multidisciplinary approach to reach a final diagnosis.
- Despite limited data, deep learning models in chest CT and CXRs could assist radiologists by providing workload relief and earlier advance interpretation in patients with COVID-19.
- To confirm generalizability of reported diagnostic performance in clinical practice, clinical validation studies with external test data needs to be done.
This data was provided by :
- A research letter article from the USA of 10 frontal chest X-ray from 5 patients.
- An original research article from China. The study included 4356 chest CT from 3322. Exclusion criteria included use of contrast and slice thickness of more than 3mm.
- Hurt, B., Kligerman, S., & Hsiao, A. (2020). Deep Learning Localization of Pneumonia. Journal Of Thoracic Imaging, 1. doi: 10.1097/rti.0000000000000512. Retrieved from : https://journals.lww.com/thoracicimaging/Citation/publishahead/Deep_Learning_Localization_of_Pneumonia__2019.99441.aspx
- Li, L., Qin, L., Xu, Z., Yin, Y., Wang, X., & Kong, B. et al. (2020). Artificial Intelligence Distinguishes COVID-19 from Community Acquired Pneumonia on Chest CT. Radiology, 200905. doi: 10.1148/radiol.2020200905. Retrieved from : https://pubs.rsna.org/doi/10.1148/radiol.2020200905