Description:
<jats:sec><jats:title>Purpose:</jats:title><jats:p>Respiratory gating is commonly used to reduce motion artifacts in positron emission tomography (PET). Clinically established methods for respiratory gating in PET require contact to the patient or a direct optical line between the sensor and the patient's torso and time consuming preparation. In this work, a contactless method for capturing a respiratory signal during PET is presented based on continuous‐wave radar.</jats:p></jats:sec><jats:sec><jats:title>Methods:</jats:title><jats:p>The proposed method relies on the principle of emitting an electromagnetic wave and detecting the phase shift of the reflected wave, modulated due to the respiratory movement of the patient's torso. A 24 GHz carrier frequency was chosen allowing wave propagation through plastic and clothing with high reflections at the skin surface. A detector module and signal processing algorithms were developed to extract a quantitative respiratory signal. The sensor was validated using a high precision linear table. During volunteer measurements and [<jats:sup>18</jats:sup>F] FDG PET scans, the radar sensor was positioned inside the scanner bore of a PET/computed tomography scanner. As reference, pressure belt (one volunteer), depth camera‐based (two volunteers, two patients), and PET data‐driven (six patients) signals were acquired simultaneously and the signal correlation was quantified.</jats:p></jats:sec><jats:sec><jats:title>Results:</jats:title><jats:p>The developed system demonstrated a high measurement accuracy for movement detection within the submillimeter range. With the proposed method, small displacements of 25 <jats:italic>μ</jats:italic>m could be detected, not considerably influenced by clothing or blankets. From the patient studies, the extracted respiratory radar signals revealed high correlation (Pearson correlation coefficient) to those derived from the external pressure belt and depth camera signals (<jats:italic>r</jats:italic> = 0.69–0.99) and moderate correlation to those of the internal data‐driven signals (<jats:italic>r</jats:italic> = 0.53–0.70). In some cases, a cardiac signal could be visualized, due to the representation of the mechanical heart motion on the skin.</jats:p></jats:sec><jats:sec><jats:title>Conclusions:</jats:title><jats:p>Accurate respiratory signals were obtained successfully by the proposed method with high spatial and temporal resolution. By working without contact and passing through clothing and blankets, this approach minimizes preparation time and increases the convenience of the patient during the scan.</jats:p></jats:sec>