Denoising of Fourier domain quantum optical coherence tomography spectrums based on deep-learning methods

Abstract

One of the promising optical coherence tomography schemes using quantum light is the Fourier domain quantum optical coherence tomography. It combines the advantage of quantum metrology and the optical coherence tomography, providing a great detection with a high axial resolution. However, the application of the Fourier domain quantum optical coherence tomography is hard to implement since various types of noise would affect the quality and finally eliminate the advantages. In this paper, we quantitively analyze the affection of the noise induced by quantum fluctuation on Fourier domain quantum optical coherence tomography and propose to suppress the effect by deep-learning method. Our simulation shows that it could severely lower the accuracy of the detection, and can be removed by our deep-learning model. We believe that our results will promote the application of similar quantum optical coherence tomography strategies to real scenarios.