A new artificial intelligence-powered deep learning model has helped radiologists in China to distinguish COVID-19 from community-acquired pneumonia and other lung diseases in chest CT imaging. In the retrospective and multi-center study, the COVID-19 Detection Neural Network—or COVNet for short—detected COVID-19 in the independent test set with 90% sensitivity and 96% specificity. “These results demonstrate that a machine learning approach using convolutional networks model has the ability to distinguish COVID-19 from community-acquired pneumonia,” concluded Lin Li, from the Department of Radiology at Wuhan Huangpi People's Hospital in China, and colleagues. With radiologists cautioning about overlap between coronavirus imaging findings and other lung issues, Li and colleagues believe AI can provide a useful assist to radiologists concerned about specificity. “There is overlap in the chest CT imaging findings of all viral pneumonias with other chest diseases that encourages a multidisciplinary approach to the final diagnosis used for patient treatment,” they added.
Researchers have trained algorithms to predict tumor sensitivity to three systemic cancer therapies using CT scans from patients with advanced non-small cell lung cancer (NSCLC). Dercle and colleagues utilized data from multiple phase II/phase III clinical trials that evaluated systemic treatment in patients with NSCLC including the immunotherapeutic agent nivolumab (Opdivo), the chemotherapeutic agent docetaxel (Taxotere), or the targeted therapeutic gefitinib (Iressa). To develop the model, the researchers used the CT images taken at baseline and on first-treatment assessment (three weeks for patients treated with gefitinib; eight weeks for patients treated with either nivolumab or docetaxel). Tumors were classified as treatment-sensitive or treatment-insensitive based on the reference standard of each trial (median progression-free survival in the nivolumab and docetaxel cohorts; analysis of surgical specimen following gefitinib treatment). The performance of each signature was evaluated by calculating the area under the curve (AUC), a measure of the model's accuracy, where a score of 1 corresponds to perfect prediction. The nivolumab, docetaxel, and gefitinib prediction models achieved an AUC of 0.77, 0.67, and 0.82 in the validation cohorts, respectively. As these types of tools continue to be developed and refined, radiomic signatures will continue to enhance clinical decision-making and the ability of physicians to provider superior care to their patients.
Looking for news updates personalized to your needs? Claim your BioTrack invite and get the latest market and research activity customized by news type, source, keywords and more.
Headlines curated by BioTrack beta and edited by Seth Schachter, Associate at DeciBio Consulting
If you would like to discuss the field of big data and AI in healthcare in more detail or provide feedback on our newsletters, please don't hesitate to reach out to me at Schachter@decibio.com or connect with me on LinkedIn