Abstract
The efficiency of lung cancer screening for reducing mortality is hindered by the high rate of false positives. Artificial intelligence applied to radiomics could help to early discard benign cases from the analysis of CT scans. The available amount of data and the fact that benign cases are a minority, constitutes a main challenge for the successful use of state of the art methods (like deep learning), which can be biased, over-fitted and lack of clinical reproducibility. We present an hybrid approach combining the potential of radiomic features to characterize nodules in CT scans and the generalization of the feed forward networks. In order to obtain maximal reproducibility with minimal training data, we propose an embedding of nodules based on the statistical significance of radiomic features for malignancy detection. This representation space of lesions is the input to a feed forward network, which architecture and hyperparameters are optimized using own-defined metrics of the diagnostic power of the whole system. Results of the best model on an independent set of patients achieve 100% of sensitivity and 83% of specificity (AUC = 0.94) for malignancy detection.
Highlights
Lung cancer is both the most frequently diagnosed cancer and cause of cancer death [1].The National Lung Screening Trial (NLST) (Study performed in United States) [2] and DutchBelgian Randomized Lung Cancer Screening Trial (NELSON) [3] have shown that lung cancer screening (LCS) with computed tomography of low dose (CTLD) reduces mortality by 20–25%
This suggests that the application of radiomics [4], could represent a critical shift in the reduction of the false positive rate and an improvement of early diagnosis of lung cancer
To address the above challenges, we propose to embed 2D slices into a low dimensional radiomic space defined by the classic radiomic features that significantly correlate to malignancy
Summary
Lung cancer is both the most frequently diagnosed cancer and cause of cancer death [1].The National Lung Screening Trial (NLST) (Study performed in United States) [2] and DutchBelgian Randomized Lung Cancer Screening Trial (NELSON) [3] have shown that lung cancer screening (LCS) with computed tomography of low dose (CTLD) reduces mortality by 20–25%. The largest screening program in Europe, the NELSON study, introduced volumetry of the nodules in consecutive CT, which meant a significant reduction of the average of false positive rate to 13%. This suggests that the application of radiomics [4] (a recent discipline that extracts a large number of image features correlating to treatment outcome), could represent a critical shift in the reduction of the false positive rate and an improvement of early diagnosis of lung cancer
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have