Exploring the feasibility of millimeter-wave sensors for non-invasive respiratory motion visualization in diagnostic imaging and therapy.

Background: Management of respiratory motion during radiation therapy is essential for accurate dose delivery and minimizing the risk to organs. In diagnostic imaging, respiratory monitoring is required for confirmation of breath-hold and four-dimensional computed tomography (CT) reconstruction. Although respiratory monitoring systems are widely used in radiation therapy, they are not often used for diagnostic imaging, where they could improve image quality.

Purpose: The purpose of this study was to use a millimeter-wave sensor (MWS) to noninvasively visualize respiratory motion, confirm breath-holding, and explore the potential for clinical implementation of an MWS in diagnostic x-ray imaging, CT, and radiation therapy.

Methods: A 24 GHz microMWS was used in this study. The MWS directionality was determined using a radio-wave dark-box system. An antenna directionality test evaluated the effective azimuthal and elevational beamwidths. Respiratory waveforms were detected by optimizing the fast Fourier transform threshold and the cutoff frequencies of the bandpass filter. To confirm the reproducibility of the MWS, the detected waveforms were compared with those of a respiratory motion phantom (QUASAR), the amplitude of motion of which could be controlled. The time from valley to peak of the waveforms obtained by normalized MWS and the QUASAR were compared. The MWS was used to acquire respiratory waveforms of 20 healthy volunteers (including an infant and a child) in geometries adopted during chest CT (supine position; anteroposterior view; source-to-surface distance, 400 mm) and chest x-ray imaging (standing position; posteroanterior view; source-to-surface distance, 1800 mm).

Results: The effective azimuthal and elevational beamwidths of the MWS were approximately ± 20° and ± 40°, respectively. By optimizing the acquisition parameters (high-sensitivity setting; with noise cancelling; frequency range, 10-20 min), the waveforms detected using the MWS approximately matched those of the respiratory motion phantom at all amplitudes. The MWS was also used to confirm breath-holding in 18 volunteers in both supine (anteroposterior view) and standing (posteroanterior view) positions. In addition, for an infant and a child who were unable to follow the instruction to stop breathing, a visual count of their inhalations matched the number of respiratory cycles measured using the MWS.

Conclusion: The 24 GHz MWS successfully monitored respiratory motion and breath-holding during radiographic and CT imaging. With effective directionality and stability, this system holds promise for clinical management of respiratory motion during diagnostic imaging and radiation therapy.