Lung nodule detection algorithms on CT low dose scans.
apply AI algorithms in identifying lung cancer lesions in thoracic low-dose computed tomography images. Compared to conventional CT, low-dose CT enables multi-acquisition of images at a lower radiation dose. Infervision has developed an advanced algorithm for lung cancer detection. Computer-aided diagnosis will help radiologists better detect abnormalities and easily obtain quantitative metrics such as nodule size and the possibility of malignancy. We believe that deep learning strategies will enhance the efficiency and accuracy of radiologists and clinicians in the diagnosis and prognosis of lung cancer.
We plan to carry out a multi-reader multi-case (MRMC) study using existing data in the NLST study to validate our algorithm. The study will also investigate whether our algorithm improves the efficiency of diagnosis by measuring the time required for radiologist review with or without our algorithm. Finally, we will analyze the risk reduction resulted from using the Infervision algorithm. The detailed steps include:
1. Identify low dose CT image from subjects with or without lung cancer.
2. Randomize the dataset into training and validation sets. Use the training set to train our algorithm and validate the results using the validation set.
3. Perform triplicate reviews by radiologists with or without Infervision technology.
4. Perform statistical analysis to investigate the impact of Infervision on
diagnosis.
5. Report findings.
Dr. Yufeng Deng, Infervision North America
ChiaHua Chang, Infervision North America
-
Detection and staging of chronic obstructive pulmonary disease using a computed tomography-based weakly supervised deep learning approach.
Sun J, Liao X, Yan Y, Zhang X, Sun J, Tan W, Liu B, Wu J, Guo Q, Gao S, Li Z, Wang K, Li Q
Eur Radiol. 2022 Feb 24 PUBMED