Low-dose Image Recognition with Quantum Computational Electron Microscopy

TL;DR

This paper demonstrates that quantum computational imaging can improve low-dose electron microscopy of sensitive specimens by using quantum algorithms and qudits to enhance image identification.

Contribution

It introduces a quantum scheme utilizing qudits for full quantum information transfer and a novel quantum algorithm for identifying images among many candidates.

Findings

Quantum information transfer is feasible with qudits near the electron beam.

The quantum algorithm can identify the correct image among more candidates than the Hilbert space dimension.

Quantum approach offers advantages in low-dose, beam-sensitive electron microscopy.

Abstract

We show that quantum computational imaging is advantageous in the setting of low-dose electron microscopy of beam-sensitive specimens. Two qudits placed near the electron beam enable full transfer of quantum information between the electron microscope and a quantum computer in the proposed scheme, providing the specimen is a phase object. We present a quantum algorithm that identifies the correct image among n candidate images, where n is larger than the effective dimension of the Hilbert space of the imaging electron.