AI Close to Radiologists on CT Lung Cancer Screenings
An open competition study among the top three Data Science Bowl 2017 deep-learning AI algorithms shows that when evaluating low-dose CT lung cancer screening images, AI can perform comparably to radiologists, according to research published in Radiology: Artificial Intelligence.
“A multinational team of researchers led by Colin Jacobs, PhD, of Radboud University Medical Center in Nijmegen, the Netherlands compared the performance of three high-performing deep-learning models to that of 11 radiologists on 300 cases. Although the mean radiologist performance was higher than all three algorithms, the difference was only statistically significant for one of them.
“‘These results offer several opportunities to optimize the reading of screening CT scans in lung cancer screening,’ the authors wrote,” reported Ridley.
Ridley noted that the AI did not provide the location of the suspicious pulmonary nodes and that, for clinical use “the algorithms’ predictions will need to be calibrated, and the optimal cut-off points for decision making should be investigated in future studies, according to the authors.”
Colin Jacobs, Arnaud A. A. Setio, Ernst T. Scholten, Paul K. Gerke, Haimasree Bhattacharya, Firdaus A. M. Hoesein, Monique Brink, Erik Ranschaert, Pim A. de Jong, Mario Silva, Bram Geurts, Kaman Chung, Steven Schalekamp, Joke Meersschaert, Anand Devaraj, Paul F. Pinsky, Stephen C. Lam, Bram van Ginneken, and Keyvan Farahani. Deep Learning for Lung Cancer Detection in Screening CT Scans: Results of a Large-Scale Public Competition and an Observer Study with 11 Radiologists. Radiology: Artificial Intelligence. Published online October 27,2021, at https://pubs.rsna.org/doi/10.1148/ryai.2021210027. Accessed November 15, 2021.