A Physics-informed Deep Neural Network for Harmonization of CT Images

Computed Tomography (CT) quantification is affected by the variability in image acquisition and rendition. This paper aimed to reduce this variability by harmonizing the images utilizing physics-based deep neural networks (DNNs). An adversarial generative network was trained on virtual CT images acquired under various imaging conditions using a virtual imaging platform with 40 computational patient models. These models featured anthropomorphic lungs with different levels of pulmonary diseases, including nodules and emphysema. Imaging was conducted using a validated CT simulator at two dose levels and varying reconstruction kernels. The trained model was tested on an independent virtual test dataset and two clinical datasets. On the virtual test set, the harmonizer improved the structural similarity index from 79.316.4% to 95.81.7%, normalized mean squared error from 16.79.7% to 9.21.7%, and peak signal-to-noise ratio from 27.73.7 dB to 32.21.6 dB. Moreover, the harmonized images yielded more precise quantification of emphysema-based imaging biomarkers for lung attenuation, LAA -950 from 5.68.7% to 0.230.16%, Perc 15 from 43.445.4 HU to 20.07.5 HU, and Lung Mass from 0.30.3 g to 0.10.2 g. In clinical data, the harmonizer reduced biomarker variability by an average of 70%. For lung nodules, harmonized images improved the detectability index by 6.5-fold and DNN-based precision by 6%. The proposed harmonizer significantly enhances image quality and quantification accuracy in CT imaging. The study demonstrated the potential utility of image harmonization for consistent CT image quality and reliable quantification, which is crucial for clinical applications and patient management.

Duke Scholars

Author Ehsan Abadi Radiology