CT cardiac stationary period detection and imaging triggering method and device based on millimeter wave radar

Abstract

The invention provides a CT cardiac stationary period detection and imaging triggering method and device based on millimeter wave radar, and the method comprises the steps: carrying out the self-adaptive offset through a chest wall micro-motion signal collected in a non-contact manner and a reference channel signal, and separating breathing and heartbeat components; obtaining a high-precision instantaneous cardiac phase and the uncertainty thereof; generating a trigger availability score in combination with the multi-dimensional features, and adaptively selecting a trigger strategy; and system delay is calibrated on line, an RR interval is predicted, an advanced triggering moment is calculated based on a target phase proportion, and accurate imaging triggering is realized in combination with a safety guard condition. According to the method, non-contact and self-adaptive accurate cardiac stationary phase detection and imaging triggering can be realized in a strong interference environment of a CT machine room, multiple challenges such as metal scattering, cardiac respiration coupling, individual difference and system delay are overcome, CT environment interference and individual difference can be effectively overcome, and the cardiac CT image quality and the scanning success rate are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of medical image data processing technology, and in particular to a method and device for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar. Background Technology

[0002] With the rapid development of medical diagnostic technology, computed tomography (CT) has become an important medical imaging tool, widely used in clinical diagnosis, disease monitoring, treatment decision-making, and intraoperative guidance. CT technology uses X-rays to penetrate human tissue and acquire cross-sectional images of tissues of different densities, providing crucial information for early disease diagnosis and precise treatment. It is irreplaceable not only in the imaging diagnosis of routine diseases but also plays a key role in radiotherapy (RT) for tumors—doctors rely on high-quality CT images for target delineation, organ at risk identification, and radiation dose distribution design; any image distortion can lead to treatment errors. Furthermore, CT has core value in emergency trauma assessment, rapid stroke screening, coronary computed tomography angiography (CTA), solid tumor staging, preoperative planning, biopsy guidance, and efficacy follow-up; image quality is decisive for clinical judgment and operational safety. However, the accuracy of CT images is highly dependent on the subject remaining stationary during the scan; if the subject moves during the scan, it will cause significant motion artifacts. Motion artifacts typically manifest as blurred images, structural distortions, or overlaps, severely impacting the accuracy of medical diagnosis and treatment planning. This is particularly pronounced in clinical practice. For example, during scans of children, elderly patients, patients with impaired consciousness or agitation, and during prolonged scans, even involuntary micro-movements by the subject can lead to decreased image quality. Especially in applications requiring extremely high image temporal stability, such as cardiac imaging, chest radiotherapy simulation positioning, and cardiac protection design for left breast cancer radiotherapy, subtle movements caused by heartbeat and respiration can result in unclear target identification or dosage errors, directly affecting clinical decisions and treatment efficacy. Traditional methods such as mechanical restraint, sedation, or ECG synchronization suffer from poor comfort, complex procedures, and limited applicability, failing to meet the clinical needs for non-invasive, efficient, and universally applicable solutions.

[0003] Currently, non-contact physiological monitoring offers a new direction for reducing artifacts without altering the CT imaging chain. Millimeter-wave radar (FMCW) can detect sub-millimeter-level micro-movements in the chest cavity, theoretically reconstructing mechanical vibration waveforms reflecting the cardiac process and indicating the relatively stationary phase within a cardiac cycle for scan gating / triggering. However, the unique environment of CT presents additional challenges to this approach: scattering and multipath propagation: the gantry / bed metal structure and rotating components generate strong echoes, easily obscuring the cardiac micro-Doppler components; cardiorespiratory coupling and body motion: increasing phase curve jitter and unstable quiescent phase localization; cross-device timing: total delay exists in the radar-to-CT control chain, and there are differences in manufacturer interfaces; individual differences: BMI / chest wall thickness, arrhythmia, etc., cause the failure of fixed thresholds and fixed window widths. In summary, due to factors such as gantry scattering, cardiorespiratory coupling, timing delays, and individual differences, simply reproducing cardiac rhythms with radar is insufficient to guarantee stable quiescent phase identification and scan triggering. Summary of the Invention

[0004] In view of this, embodiments of this application provide a method and device for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar, in order to eliminate or improve one or more defects existing in the prior art.

[0005] One aspect of this application provides a method for CT cardiac quiescent phase detection and imaging triggering based on millimeter-wave radar, comprising: Based on millimeter-wave radar array, the reference channel signal of the metal interference source in the space where the CT scanning table is located is acquired non-contactly, rather than acquiring the chest wall micro-motion observation signal of the target user located on the CT scanning table. Based on the reference channel signal, the chest wall micromotion observation signal is subjected to reference channel adaptive cancellation processing to obtain the corresponding chest wall micromotion purification signal. Obtain the estimated respiratory rate and initial heart rate signal corresponding to the chest wall micro-motion purification signal, respectively; Construct the analytical signal corresponding to the initial heartbeat signal, and obtain the instantaneous cardiac phase and the uncertainty of the instantaneous cardiac phase of the analytical signal; Multi-dimensional feature vectors are extracted from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate, and a trigger availability score for the CT scanner is determined based on the multi-dimensional feature vectors. Select the corresponding triggering strategy based on the trigger availability score; The system delay is calibrated online, and the RR interval of the next cardiac cycle is predicted based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, an early triggering time is determined. When the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, a scan triggering command is sent to the CT machine at the early triggering time to acquire the CT image data of the target user.

[0006] In some embodiments of this application, the non-contact acquisition of reference channel signals from metal interference sources within the space where the CT scan bed is located, based on a millimeter-wave radar array, is not the non-contact acquisition of chest wall micro-motion observation signals from the target user located on the CT scan bed, but includes: Based on a millimeter-wave radar array, a reference channel signal of a metal interference source in the space where the CT scanning bed is located is acquired non-contactly. The metal interference source includes: metal parts on the CT gantry or CT scanning bed. The main lobe of the millimeter-wave radar array is directed towards the chest wall of the target user located on the CT scanning table. Based on the millimeter-wave radar array, the raw chest wall micro-motion signal of the target user located on the CT scanning table is acquired non-contactly. Null adaptive beamforming processing is performed on the raw chest wall micro-motion signal to spatially suppress scattering interference from the metal structure by forming a null in the direction of the interference source, thereby obtaining the chest wall micro-motion observation signal of the target user.

[0007] In some embodiments of this application, the step of performing reference channel adaptive cancellation processing on the chest wall micromotion observation signal based on the reference channel signal to obtain the corresponding chest wall micromotion purification signal includes: The reference channel signal is input into an adaptive filter so that the adaptive filter estimates the interference contribution of the reference channel signal to the chest wall micromotion observation signal based on the current filter coefficients, and obtains the chest wall micromotion purification signal by calculating the residuals; wherein, the filter coefficients are updated using the normalized least mean square algorithm.

[0008] In some embodiments of this application, obtaining the estimated respiratory rate and initial heart rate signal corresponding to the chest wall micro-motion purification signal respectively includes: The respiratory rate is estimated from the chest wall micro-movement purification signal; Construct a time-varying bandstop filter whose stopband center frequency dynamically corresponds to the estimated breathing frequency and its harmonics. The time-varying band-stop filter is used to filter the chest wall micro-motion purification signal to remove the respiratory component and its harmonics, thereby obtaining the initial heartbeat signal.

[0009] In some embodiments of this application, constructing the analytical signal corresponding to the initial heartbeat signal and obtaining the instantaneous cardiac phase and uncertainty of the instantaneous cardiac phase of the analytical signal includes: Perform a Hilbert transform on the initial heartbeat signal to construct the analytic signal corresponding to the initial heartbeat signal; Calculate the phase angle of the analyzed signal and perform phase unwrapping to obtain the unwrapped instantaneous phase; The unwrapped instantaneous phase is smoothed and optimally estimated using a Kalman filter to obtain the smoothed instantaneous cardiac phase and its uncertainty.

[0010] In some embodiments of this application, prior to the non-contact acquisition of chest wall micromotion observation signals of the target user located on the CT scan table, the method further includes: The ambient noise in the space where the currently unmanned CT scanning bed is located is collected based on a millimeter-wave radar array; Correspondingly, the extraction of multi-dimensional feature vectors from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate, and the determination of the trigger availability score for the CT scanner based on the multi-dimensional feature vectors, includes: The heart rate value, heart rate variability value, and PVC / AF segment ratio of the rhythm irregularity index are extracted from the instantaneous cardiac phase. The chest wall attenuation amplitude is extracted from the initial heartbeat signal. The signal-to-noise ratio is calculated based on the initial heartbeat signal and the environmental noise. The respiratory rate value is extracted from the respiratory rate estimate to form a multi-dimensional feature vector containing a preset initial system delay value, the heart rate value, heart rate variability value, PVC / AF segment ratio of the rhythm irregularity index, chest wall attenuation amplitude, signal-to-noise ratio, and respiratory rate value. The multi-dimensional feature vector is input into a pre-defined discrimination model so that the discrimination model outputs a value between 0 and 1 as a trigger availability score for the CT scanner.

[0011] In some embodiments of this application, selecting the corresponding triggering strategy based on the trigger availability score includes: If the trigger availability score is greater than or equal to a preset first threshold, then the first triggering strategy with mid-to-late diastole as the target cardiac phase is selected. If the trigger availability score is less than the first threshold and the trigger availability score is greater than or equal to the second threshold, then a second triggering strategy with end-systole as the target cardiac phase is selected; wherein, the first threshold is greater than the second threshold. If the trigger availability score is less than the second threshold, or if a continuous irregular rhythm is detected, a third triggering strategy is selected, which involves opening multiple trigger windows on multiple cardiac phases and limiting triggering to the end of expiration.

[0012] In some embodiments of this application, the online calibration system delays and predicts the RR interval of the next cardiac cycle based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, it determines an early triggering time. When the uncertainty and the RR interval of the next cardiac cycle meet preset safety protection conditions, a scan trigger command is issued to the CT scanner at the early triggering time to acquire CT image data of the target user, including: A test trigger command is sent to the CT scanner and the response timestamp of the CT scanner is recorded. The time difference between the time when the test trigger command is sent and the response timestamp is calculated. Based on the time difference, the current system latency is estimated using a rolling average algorithm. Based on historical RR interval data, a hybrid prediction model combining weighted moving average and median filtering is used to predict the RR interval of the next cardiac cycle. The RR interval of the next cardiac cycle is obtained by weighted summation of the median of the most recent historical RR interval data and multiple other historical RR interval data, and the weight coefficient of the most recent historical RR interval is between 0.4 and 0.7. Starting from the previous reference time, add the product between the target phase ratio of the cardiac resting period corresponding to the triggering strategy and the RR interval of the next cardiac cycle, and then subtract the current system delay to obtain the advance triggering time; If the absolute value of the prediction error between the RR interval of the next cardiac cycle and the RR interval measured after the end of the next cardiac cycle is less than or equal to a preset error threshold, and the current uncertainty is less than or equal to a preset uncertainty threshold, then it is determined that the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, and a scan trigger command is sent to the CT machine at the early trigger time to collect the CT image data of the target user; If the absolute value of the prediction error is greater than the error threshold, or if the current uncertainty is greater than the uncertainty threshold, then the triggering of the current cycle is paused and the re-evaluation is postponed, or the triggering strategy is downgraded.

[0013] Another aspect of this application provides a CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar, used to execute the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in the first aspect.

[0014] A third aspect of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the described millimeter-wave radar-based CT cardiac quiescent period detection and imaging triggering method.

[0015] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the described millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method.

[0016] The fifth aspect of this application provides a computer program product comprising a computer program that, when executed by a processor, implements the described millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method.

[0017] The CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application uses a millimeter-wave radar array to non-contactly acquire reference channel signals of metal interference sources in the space where the CT scanning table is located, rather than acquiring chest wall micro-motion observation signals of the target user on the CT scanning table. Based on the reference channel signals, the chest wall micro-motion observation signals are subjected to reference channel adaptive cancellation processing to obtain corresponding chest wall micro-motion purification signals. The estimated respiratory rate and initial heartbeat signals corresponding to the chest wall micro-motion purification signals are obtained. An analytical signal corresponding to the initial heartbeat signal is constructed, and the instantaneous cardiac phase and uncertainty of the instantaneous cardiac phase are obtained. Multidimensional signals are extracted from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate. The system uses a multi-dimensional feature vector to determine the trigger availability score for the CT scanner. A corresponding triggering strategy is selected based on the trigger availability score. The system delay is calibrated online, and the RR interval of the next cardiac cycle is predicted based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent phase corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, an early triggering time is determined. When the uncertainty and the RR interval of the next cardiac cycle meet preset safety protection conditions, a scan trigger command is sent to the CT scanner at the early triggering time to acquire CT images of the target user. Non-contact, electrode-free chest wall micro-motion acquisition is achieved through millimeter-wave radar, effectively overcoming the limitations of ECG gating and patient discomfort. An anti-interference mechanism with adaptive cancellation of the reference channel enables stable extraction of sub-millimeter-level heartbeat signals in the strongly scattering CT scanner environment. Instantaneous phase estimation and uncertainty quantification provide a high-precision time reference and reliability index for the cardiac cycle. The trigger availability score and adaptive strategy selection based on multi-dimensional features intelligently address individual differences and physiological changes. By using online system delay calibration and RR interval prediction, proactive and accurate triggering can be achieved, while safety safeguards effectively prevent false triggering. This entire solution significantly improves cardiac CT image quality, reduces motion artifacts, increases scan success rate and clinical efficiency, and provides reliable technical support for cardiac imaging diagnosis.

[0018] Additional advantages, objectives, and features of this application will be set forth in part in the description which follows, and will in part become apparent to those skilled in the art upon review of the following description, or may be learned by practice of the application. The objectives and other advantages of this application can be realized and obtained by means of the structures specifically pointed out in the specification and drawings.

[0019] Those skilled in the art will understand that the purposes and advantages that can be achieved with this application are not limited to those specifically described above, and that the above and other purposes that this application can achieve will be more clearly understood from the following detailed description. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, do not constitute a limitation thereof. The components in the drawings are not drawn to scale but are merely for illustrating the principles of this application. For ease of illustration and description of certain parts of this application, corresponding portions in the drawings may be enlarged, i.e., may appear larger relative to other components in an exemplary device actually manufactured according to this application. In the drawings: Figure 1 This is a schematic diagram of the first process of a CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram of the second process of the CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar in one embodiment of this application.

[0022] Figure 3 This is a schematic diagram of the structure of a CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar in one embodiment of this application.

[0023] Figure 4 This is a schematic diagram of the process of a CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar in an application example of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and their descriptions are used to explain this application, but are not intended to limit it.

[0025] It should also be noted that, in order to avoid obscuring this application with unnecessary details, only the structures and / or processing steps closely related to the scheme according to this application are shown in the accompanying drawings, while other details that are not closely related to this application are omitted.

[0026] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0027] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0028] In the following description, embodiments of the present application will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0029] It should be noted that the main methods currently used in clinical practice to reduce motion artifacts during CT scans include the following: (a) ECG gating / triggering Acquiring cardiac phase by attaching electrodes and triggering scanning at end-diastole can reduce cardiac artifacts to some extent. However, it is sensitive to electrode attachment quality, lead management, and patient compliance; it has poor applicability in emergency departments, infants, young children, and people with sensitive skin or limited body surface conditions; it is prone to phase drift or triggering failure in cases of arrhythmia, and the procedure is complex and uncomfortable.

[0030] (ii) Drug intervention and sedation / heart rate reduction Using beta-blockers or sedatives to prolong diastole and reduce heart rate carries risks and individual variability, limiting its applicability to the elderly and those with underlying diseases; adverse drug reactions further restrict its clinical application.

[0031] (III) Image Post-processing Motion Correction Relying on post-reconstruction registration / AI correction makes it impossible to intervene in the acquisition sequence in real time; when the range of motion is large or quality control is insufficient, the correction accuracy is limited and may mask the true lesion boundary.

[0032] (iv) Other non-contact motion detection Optical, laser, and video technologies focus more on overall body position and respiration monitoring, but are greatly affected by lighting, obstruction, and safety regulations, making it difficult to reliably capture the subtle mechanical movements inside the heart.

[0033] (v) Millimeter-wave radar for monitoring vital signs Existing research and applications have shown that millimeter-wave radar can reconstruct thoracic mechanical vibrations and characterize cardiac rhythms under non-contact conditions, but these works are mostly limited to general vital sign monitoring or rhythm recognition; applications in CT environments are still lacking. 1. Anti-interference acquisition link for rack / bed strong scattering and multipath interference; 2. Joint estimation of instantaneous phase and uncertainty is based on the stable quiescent period. 3. Adaptive strategies based on patient characteristics (window width / phase / multi-window voting) rather than fixed thresholds; 4. Online calibration and forward-looking prediction triggering of system delays across device timings; Therefore, existing technologies have not yet provided a complete solution for non-contact cardiac quiescent phase identification and scan control that can be deployed on a large scale for real-world CT scenarios.

[0034] Based on this, in order to achieve non-contact, adaptive, and accurate cardiac quiescent phase detection and imaging triggering in the strong interference environment of a CT room, and to overcome multiple challenges such as metal scattering, cardiorespiratory coupling, individual differences, and system delays, this application provides a CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar, a CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar for executing the millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method, a physical device, a computer-readable storage medium, and a computer program product. It provides a complete set of solutions for anti-interference acquisition, joint estimation of phase and uncertainty, adaptive patient feature-driven strategy, and time-series compensation and predictive triggering for CT scenarios.

[0035] The following examples will provide a detailed description.

[0036] Based on this, embodiments of this application provide a millimeter-wave radar-based method for detecting and triggering the quiescent phase of cardiac activity in CT scans, which can be implemented by a millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering device. See [link to relevant documentation]. Figure 1 The method for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar specifically includes the following: Step 100: Based on the millimeter-wave radar array, non-contact acquisition of the reference channel signal of the metal interference source in the space where the CT scanning table is located, and non-contact acquisition of the chest wall micro-motion observation signal of the target user located on the CT scanning table.

[0037] In one or more embodiments of this application, the millimeter-wave radar array may refer to a multi-input multi-output radar system employing frequency modulated continuous wave (FMCW) technology, with no fewer than 8 virtual array elements, and supports polarization selection to reduce coupling interference.

[0038] Understandably, non-contact acquisition refers to radar sensing physiological activity by emitting electromagnetic waves and receiving their echoes without physical contact with the patient. The reference channel signal is acquired through a dedicated antenna channel pointed towards metallic interference sources such as the CT gantry / bed, primarily used to obtain clean interference samples. The chest wall micromotion observation signal is a raw, mixed signal acquired by pointing the main array towards the patient's anterior chest wall, containing components of heartbeat, respiration, and environmental interference; it can be simply referred to as the observation signal.

[0039] Step 200: Based on the reference channel signal, perform reference channel adaptive cancellation processing on the chest wall micro-motion observation signal to obtain the corresponding chest wall micro-motion purification signal.

[0040] In step 200, the chest wall micro-motion purification signal refers to the output signal after dual anti-interference processing, which retains the main vital signs information of the target user (such as the patient).

[0041] Step 300: Obtain the estimated respiratory rate and initial heart rate signal corresponding to the chest wall micro-motion purification signal.

[0042] In step 300, the respiratory rate estimate can be extracted by estimating the dominant respiratory frequency and its harmonic components through spectral analysis. The initial heartbeat signal refers to the mechanical vibration signal of the heartbeat obtained after dynamically filtering out the respiratory component using a time-varying band-stop filter; the initial heartbeat signal can also be called the chest wall displacement signal.

[0043] Step 400: Construct the analytical signal corresponding to the initial heartbeat signal, and obtain the instantaneous cardiac phase and the uncertainty of the instantaneous cardiac phase of the analytical signal.

[0044] In step 400, the analytic signal z(t) can be obtained in complex form as z(t) = x(t) + j·H{x(t)} by performing a Hilbert transform on the initial heartbeat signal x(t), where j represents the imaginary unit and H{} represents the Hilbert transform.

[0045] The instantaneous cardiac phase refers to the smoothed phase curve φ(t) obtained by calculating arg z(t), followed by phase unwrapping and Kalman filtering, ranging from 0° to 360°. Here, arg z(t) represents the phase angle of the complex analytic signal z(t). Uncertainty It is a confidence index for phase estimation calculated from the covariance of Kalman filtering.

[0046] Step 500: Extract multi-dimensional feature vectors from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate, and determine the trigger availability score for the CT scanner based on the multi-dimensional feature vectors.

[0047] Understandably, a multi-dimensional feature vector is a set of features including heart rate, heart rate variability, signal-to-noise ratio, chest wall attenuation amplitude, respiratory rate, and rhythm irregularity indicators. Triggering usability scoring refers to using a logistic regression model... Or the confidence score in the [0,1] interval calculated by the gradient boosting tree model, where, A multidimensional feature vector is a list containing all key physiological and signal quality indicators, such as: z = [heart rate, heart rate variability, signal-to-noise ratio, phase uncertainty, respiratory rate, chest wall attenuation amplitude, ...]. This represents the first weight vector; This represents the bias term, which can be a constant. It adjusts the output baseline of the entire model, ensuring that the model has a reasonable output even when all feature values ​​are general. It is also learned during the training phase. This represents the Sigmoid function (activation function). This indicates that an availability score has been triggered.

[0048] Step 600: Select the corresponding triggering strategy based on the trigger availability score.

[0049] It is understandable that the triggering strategy refers to different modes of prediction triggering, such as the mid-to-late diastolic window, the end-systolic window, or multiple windows, which are pre-set by the human based on the score.

[0050] Step 700: The system delay is calibrated online and the RR interval of the next cardiac cycle is predicted based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, an early triggering time is determined. When the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, a scan triggering command is sent to the CT machine at the early triggering time to acquire the CT image data of the target user.

[0051] As is understandable, the RR interval refers to the time interval between two consecutive R waves (representing the peak value of ventricular depolarization) on an electrocardiogram (ECG), used to assess the rhythm and frequency of cardiac electrical activity. It is a key indicator for measuring ventricular rhythm and, together with the PP interval (atrial rhythm), reflects the function of the cardiac conduction system.

[0052] It should be noted that system delay calibration refers to the end-to-end delay from the radar to the CT sensor, achieved through testing. This, also known as the total delay, is continuously updated using a rolling average algorithm. RR interval prediction can be achieved using... A hybrid prediction model, in which, It is the second weighting coefficient; α∈[0.4,0.7]; It is the RR interval of the next cardiac cycle (i.e., the kth RR interval), which is the predicted value; It is the most recent historical RR interval data (i.e., the (k-1)th RR interval). This is a statistical value used to represent the median of four historical RR intervals, from the (k-2)th to the (k-5)th. It reflects the general level and stable trend of recent heart rate and can effectively filter out occasional abnormal fluctuations (such as a single, accidental premature beat).

[0053] Among them, the safety protection conditions include protection mechanisms when the prediction error exceeds ±5%RR or the phase uncertainty exceeds the threshold.

[0054] In step 700, the trigger time is advanced. It can be pressed The calculated precise trigger time. Among them, Indicates the target phase ratio during the cardiac resting period; It represents the previous reference time as the starting point of time, that is, the starting time of the previous heartbeat cycle, and also the time when the mechanical R wave occurs as defined by the radar signal. It is the reference point for all time calculations and can be abbreviated as reference R time.

[0055] The scan trigger command is an exposure control signal sent to the CT equipment via a hardware trigger interface or network protocol. CT image data is tomographic scan data acquired during cardiac rest and is used to reconstruct diagnostic images free of motion artifacts.

[0056] Furthermore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application embodiment is executed by a CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar. This device can serve as the signal processing and control unit in a CT cardiac quiescent phase detection and imaging triggering system based on millimeter-wave radar (which can be simply referred to as the system). The system also includes a millimeter-wave MIMO radar, a reference channel / sensor module, a signal processing and control unit, a CT interface module, and a human-machine interface. The signal processing and control unit is connected to the millimeter-wave MIMO radar, the reference channel / sensor module, the CT interface module, and the human-machine interface. The communication connection between the human-machine interface and the millimeter-wave MIMO radar is as follows: The millimeter-wave MIMO radar and the reference channel / sensor module are also connected. When executing step 100, the signal processing and control unit can send an acquisition command to the millimeter-wave MIMO radar, enabling the reference channel / sensor module, which is connected to the millimeter-wave MIMO radar, to non-contactly acquire the reference channel signal of the metal interference source in the space where the CT scan bed is located, instead of contactively acquiring the chest wall micro-motion observation signal of the target user on the CT scan bed. Then, the signal processing and control unit obtains the reference channel signal and the chest wall micro-motion observation signal from the reference channel / sensor module. Furthermore, when executing step 700, the signal processing and control unit can send a scan trigger command to the CT interface module of the CT machine at the pre-triggered moment, causing the CT machine to acquire the CT image data of the target user. The CT image data can then be sent to the signal processing and control unit, which then transmits it to the human-machine interface for display. Simultaneously, during the execution of steps 100 to 600, the signal processing and control unit can also send corresponding data to the human-machine interface for display, such as signal result indicators and trigger status.

[0057] As described above, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application achieves non-contact, electrode-free chest wall micro-motion acquisition through millimeter-wave radar, effectively overcoming the applicability limitations and patient discomfort of ECG gating. The anti-interference mechanism of adaptive cancellation via the reference channel enables stable extraction of sub-millimeter-level heartbeat signals in the strong scattering environment of a CT room. Instantaneous phase estimation and uncertainty quantification provide a high-precision time reference and reliability index for the cardiac cycle. Trigger availability scoring based on multi-dimensional features and adaptive strategy selection intelligently address individual differences and physiological changes. Online system delay calibration and RR interval prediction enable forward-looking and accurate triggering, effectively preventing false triggering by combining safety protection conditions. The entire solution significantly improves cardiac CT image quality, reduces motion artifacts, increases scan success rate and clinical work efficiency, providing reliable technical support for cardiac imaging diagnosis.

[0058] To actively suppress strong scattering interference from fixed metal structures such as CT gantry and scanning bed, and to prevent the target signal from being overwhelmed, a CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar is provided in this application embodiment. (See also...) Figure 2 Step 100 of the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 110: Based on a millimeter-wave radar array, non-contact acquisition of reference channel signals of metal interference sources in the space where the CT scanning bed is located, wherein the metal interference sources include: metal components on the CT gantry or CT scanning bed.

[0059] In one example, the millimeter-wave radar antenna can be located at the window opening of the CT gantry, with the main lobe pointing towards the anterior chest wall, and orthogonal polarization is used to reduce metal scattering coupling.

[0060] Step 120: Control the main lobe of the millimeter-wave radar array to point towards the chest wall of the target user located on the CT scanning table. Based on the millimeter-wave radar array, non-contact acquisition of the original chest wall micro-motion signal of the target user located on the CT scanning table is performed. Null adaptive beamforming processing is performed on the original chest wall micro-motion signal to spatially suppress scattering interference from the metal structure by forming a null in the direction of the interference source, thereby obtaining the chest wall micro-motion observation signal of the target user.

[0061] In step 120, the null adaptive beamforming specifically involves: setting the array weight w, so that the main lobe points towards the anterior chest wall. Side lobe relative to frame / bed direction Forming a zero trap, optimization: Online fine-tuning with small steps is achieved using LMS / RLS to adapt to changes in patient micro-displacement and gantry angle.

[0062] in, This indicates that the array weights w are optimized to find the optimal solution; Represent the optimization objective (cost function); Indicates constraints; This represents the Hermitian transpose of the array weights w. This represents the desired signal steering vector, used to describe the direction from the target. The signal (i.e., the patient's anterior chest wall) generates a phase difference on each antenna element, which defines the direction of the "main lobe". This indicates a main lobe constraint, requiring the constraint to originate from the target direction. The signal, after being processed by the array weight w, must have a gain of 1 (i.e., lossless transmission). This ensures clear reception of the patient's chest wall signal.

[0063] This represents the steering vector of the interference signal, used to describe the direction from the interference source. The signal (i.e., the CT gantry or CT scanning bed) generates a phase difference on each antenna element. This represents a zero trap constraint, requiring the constraint to originate from the direction of disturbance. The signal, after being processed by the array weight w, must have a gain of 0 (i.e., be completely canceled). This creates a null trap in the direction of the interference source, achieving spatial anti-interference.

[0064] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application embodiment forms a "silent zone" in the direction of the interference source by suppressing spatial domain interference, which greatly improves the signal-to-noise ratio and provides a high-quality input signal for subsequent signal processing.

[0065] To further eliminate quasi-static environmental interference in the signal domain and provide a second line of defense against interference, a CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar is provided in this application embodiment. (See also...) Figure 2 Step 200 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 210: Input the reference channel signal into an adaptive filter so that the adaptive filter estimates the interference contribution of the reference channel signal to the chest wall micromotion observation signal based on the current filter coefficients, and obtains the chest wall micromotion purification signal by calculating the residuals; wherein, the filter coefficients are updated using the normalized least mean square algorithm.

[0066] Specifically, the adaptive cancellation of the reference channel can employ the NLMS algorithm to update the filter coefficients in real time. Through adaptive filters Eliminate quasi-static interference from the observed signal.

[0067] in, Indicates the reference channel signal; Indicates the observed signal; This indicates a chest wall micro-movement purification signal, which can be simply referred to as a purification signal. The algorithm is updated online using the Normalized Least Mean Square (NLMS) algorithm to suppress quasi-static strong echoes.

[0068] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application complements the null beamforming, and achieves dual purification by adaptive filtering to track and cancel residual interference in real time, ensuring the purity of the heartbeat signal.

[0069] To effectively separate overlapping respiratory and heartbeat signals and avoid interference from respiratory harmonics in heartbeat detection, a method for detecting and triggering the quiescent phase of cardiac activity in CT based on millimeter-wave radar, as provided in this application embodiment, is described below. Figure 2 Step 300 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 310: Estimate the respiratory rate from the chest wall micro-movement purification signal.

[0070] Specifically, the chest wall micro-movement purification signal can be windowed first using a sliding window (such as a Hamming window) with a length of 4-8 seconds and an overlap rate of 50% to balance frequency resolution and real-time performance. Then, a Fast Fourier Transform (FFT) is performed on each signal segment to obtain a power spectral density estimate. Next, within a typical respiratory rate range (0.1-0.5 Hz, corresponding to 6-30 breaths / minute), the dominant peak in the power spectrum is identified. If the peak amplitude is significantly higher than the background noise level (signal-to-noise ratio > 3 dB), remains stable within three or more consecutive analysis windows (frequency change < 0.05 Hz), and its second harmonic (2f_R) is visible in the spectrum, then this dominant peak is determined as the respiratory frequency. A Kalman filter or α-β tracker can then be used to smooth the detected respiratory frequency, suppressing transient fluctuations and outputting a stable respiratory rate estimate. .

[0071] Step 320: Construct a time-varying bandstop filter whose stopband center frequency dynamically corresponds to the estimated breathing frequency and its harmonics.

[0072] In step 320, a stopband with a width of 0.15 Hz can be set with the current estimated respiratory rate as the center frequency; simultaneously, additional stopbands with widths of 0.2 Hz and 0.25 Hz, centered at twice and three times the estimated respiratory rate respectively (considering potential harmonic frequency spread), can be configured. An infinite impulse response (IIR) elliptic filter structure can be used, and the filter coefficients can be updated in real time according to the estimated respiratory rate to obtain a time-varying bandstop filter B(f, t). The strategy for real-time updating the filter coefficients can be: updating the coefficients immediately when |ΔfR|>0.02 Hz, where ΔfR is the change in the estimated respiratory rate per unit time; under normal circumstances, a forced update is performed every 2 seconds to prevent coefficient stagnation; a smooth transition mechanism is used to avoid transient responses caused by abrupt coefficient changes. This allows the time-varying bandstop filter B(f, t) to suppress the mechanical micro-motion components within the cardiac subband (approximately 5–40 Hz). and harmonics.

[0073] Step 330: Use the time-varying band-stop filter to filter the chest wall micro-motion purification signal to remove the respiratory component and its harmonics, and obtain the initial heartbeat signal.

[0074] Specifically, the chest wall micro-movement purification signal can be passed through the aforementioned time-varying band-stop filter to filter out the respiratory fundamental wave and its harmonic components in real time. Due to the phase distortion introduced by the IIR filter, the system adopts zero-phase filtering technology, that is, forward-backward filtering of the signal to eliminate the influence of nonlinear phase, and the overlap-retention method is used at the data block boundary to ensure continuity.

[0075] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application uses a time-varying band-stop filter to dynamically track respiratory changes, accurately filter out the respiratory fundamental wave and its harmonics, and retain the complete mechanical vibration characteristics of the heartbeat.

[0076] To reliably extract high-precision instantaneous cardiac phase from noisy environments and quantify the reliability of the estimation, a method for detecting and triggering the quiescent phase of cardiac activity in CT based on millimeter-wave radar is provided in this application embodiment. (See also...) Figure 2 Step 400 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 410: Perform a Hilbert transform on the initial heartbeat signal to construct the analytic signal corresponding to the initial heartbeat signal.

[0077] Specifically, the Hilbert transform H{x(t)} can be achieved using a 90-degree phase shift filter. The specific implementation adopts an FIR filter structure with a length of 101 sampling points to ensure an accurate 90-degree phase shift within the heartbeat frequency band, resulting in the constructed analytic signal z(t). This analytic signal can eliminate negative frequency components, providing a basis for subsequent phase calculations.

[0078] Step 420: Calculate the phase angle of the analyzed signal and perform phase unwrapping to obtain the unwrapped instantaneous phase.

[0079] Specifically, the phase angle of the analytical signal can be calculated using the four-quadrant arctangent function atan2 to ensure that the phase value is within the range of [-π, π] radians. When the absolute value of the phase difference between adjacent sampling points is greater than π, a phase jump is determined to have occurred, and an unwrapping operation is performed on argz(t). The instantaneous phase after unwrapping exhibits a monotonically increasing characteristic, ranging from 0 to +∞.

[0080] Step 430: Use a Kalman filter to smooth and optimally estimate the unwrapped instantaneous phase to obtain the smoothed instantaneous cardiac phase and the uncertainty of the instantaneous cardiac phase.

[0081] Specifically, the state vector of the Kalman filter can be X = [φ, ω]ᵀ, containing the instantaneous cardiac phase φ and instantaneous angular velocity ω; the state transition model of the Kalman filter can be φ k = φ {k-1} + ω {k-1} ·Δt+ w φ ;ω k = ω {k-1} + w ω ; Observation model z of Kalman filter k = φ k + v k v k To observe noise, the instantaneous cardiac phase can also be denoted as φ(t).

[0082] Where Δt is the sampling interval, w φ and w ω For process noise; φ {k-1} φ represents the optimal estimate of the instantaneous cardiac phase at the previous discrete time point k-1; k ω represents the optimal estimate of the instantaneous cardiac phase at the current discrete time point k; {k-1} ω represents the optimal estimate of the instantaneous angular velocity at the previous discrete time point k-1; k This represents the optimal estimate of the instantaneous angular velocity at the current discrete time point k.

[0083] Based on this, the specific process of smoothing and optimally estimating the unwrapped instantaneous phase using a Kalman filter is as follows: Before filtering begins, the Kalman filter is initialized. Based on the cycle length of the first few heartbeats, the average heart rate is estimated, and the initial angular velocity is calculated accordingly. Simultaneously, the initial phase is set to a reasonable starting point (e.g., 0 degrees). The Kalman filter also initializes two key uncertainty indices: one for phase and one for angular velocity, indicating significant uncertainty in the initial perception of these states. Once initialization is complete, the Kalman filter enters a continuous "prediction-update" loop. The first step is prediction, which, based on the optimally estimated phase and angular velocity from the previous moment, predicts the most likely phase the heart is in at the current moment. Specifically, assuming the heart moves at a constant angular velocity over a very short time interval, the new predicted phase equals the old phase plus the product of angular velocity and time. Simultaneously, the Kalman filter also predicts that the angular velocity remains essentially constant. Due to model imperfections, the Kalman filter appropriately increases the uncertainty of these predictions. The second step is updating. The Kalman filter compares the latest actual observation (i.e., the unwrapped instantaneous phase) with the prediction from the previous step, calculating an optimal weighting coefficient called the Kalman gain. This weight determines the degree to which the current observation should be trusted over the previous prediction. If the observed signal quality is high (high signal-to-noise ratio), the weight will tend to trust the new observation and use it to significantly correct the prediction; if the observed signal is noisy, the weight will tend to trust the model-based prediction and be more cautious about the observed value. Finally, the Kalman filter fuses the prediction and observation values ​​according to this weight to obtain the optimal smoothed phase estimate for the current moment, and simultaneously calculates the optimal estimate of the angular velocity. At the same time, it also precisely updates the uncertainty of the phase estimate at this moment. That is, calculating based on the filter covariance and instantaneous SNR. It is used as a criterion for window width adaptation and backoff.

[0084] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application provides a smooth and continuous phase trajectory and its uncertainty through the combination of Hilbert transform and Kalman filtering, providing a reliable time reference and confidence assessment for triggering decisions.

[0085] To comprehensively analyze multi-dimensional information and quantitatively evaluate the success probability and reliability of triggering a scan under the current state, this application provides a CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar, see [link to relevant documentation]. Figure 2 The method for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar, before step 120, specifically includes the following: Step 111: Collect the ambient noise of the space where the currently unmanned CT scanning bed is located based on the millimeter-wave radar array.

[0086] Correspondingly, step 500 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 510: Extract the corresponding heart rate value, heart rate variability value, and rhythm irregularity index PVC / AF segment ratio from the instantaneous cardiac phase; extract the corresponding chest wall attenuation amplitude from the initial heartbeat signal; calculate the signal-to-noise ratio based on the initial heartbeat signal and the environmental noise; and extract the respiratory rate value from the respiratory rate estimate to construct a multi-dimensional feature vector containing a preset initial system delay value, the heart rate value, heart rate variability value, rhythm irregularity index PVC / AF segment ratio, chest wall attenuation amplitude, signal-to-noise ratio, and respiratory rate value.

[0087] Specifically, by analyzing the instantaneous cardiac phase curve, detecting two consecutive 360° phase reset points (equivalent to the R wave moment), and calculating the time interval to obtain the RR interval, the system converts this to heart rate per minute using the formula HR = 60 / RR interval. The system maintains a heart rate queue of length 8, outputting the median as the current heart rate value, i.e., the HR value. It can collect RR interval sequences from the most recent 2 minutes, using time-domain analysis to calculate SDNN (standard deviation of all normal RR intervals) and RMSSD (root mean square of the difference between adjacent RR intervals), and then using the weighted sum of these two indicators as the heart rate variability (HRV) value (SDNN / RMSSD). The system can monitor the RR interval sequence in real time, marking suspected premature beats when the difference between adjacent RR intervals exceeds 20%. By analyzing the spectral entropy of the heart rate signal through a sliding window, it identifies the disordered patterns unique to atrial fibrillation. Finally, it calculates the proportion of abnormal rhythm segments within the most recent 30 seconds as the PVC / AF segment ratio, a rhythm irregularity indicator. Envelope detection can be performed on the initial heartbeat signal, and the peak-to-peak amplitude of the signal can be calculated within each cardiac cycle. The average amplitude of the most recent 8 cycles is taken as the chest wall attenuation amplitude. The signal-to-noise ratio (SNR) reflects the intensity of the heart's mechanical contraction. The power of the initial heartbeat signal after passing through a bandpass filter (0.8-5Hz) can be used as the signal power, and the power of ambient noise within the same frequency band can be used as the noise power. The SNR value can be calculated using the formula SNR = 10·log10(signal power / noise power). The respiratory rate estimate can be directly used as the respiratory rate (BR) value, in breaths per minute. The total delay time from radar to CT trigger, pre-calibrated by the system, can be used as the system delay. Initial values ​​are then used. All features can then be Z-score standardized before integration to eliminate the influence of dimensions. The final 7-dimensional feature vector is arranged in a fixed order to obtain a multi-dimensional feature vector z = [heart rate value, heart rate variability value, rhythm irregularity index, chest wall attenuation amplitude, signal-to-noise ratio, respiratory rate value, initial system delay].

[0088] Step 520: Input the multi-dimensional feature vector into a pre-defined discriminant model so that the discriminant model outputs a value between 0 and 1 as a trigger availability score for the CT scanner.

[0089] Specifically, two optional discriminant models can be built-in: a logistic regression model and a lightweight gradient boosting tree model. Model parameters are trained using thousands of clinical data points and pre-loaded into memory upon system startup. When a new feature vector arrives, the system first performs an integrity check to ensure all feature values ​​are within a reasonable range. For the logistic regression model, the inner product of the feature vector and the pre-trained weight vector is calculated, and after adding a bias term, it is mapped to the 0-1 interval using the sigmoid function. For the gradient boosting tree model, the feature vector can be input into an ensemble model consisting of hundreds of weak decision trees, and a voting mechanism is used to derive the final confidence score, which serves as the trigger for usability scoring.

[0090] In addition, the following further processing can be performed: the raw trigger availability score output by the model is filtered using a moving average to eliminate instantaneous fluctuations and ensure the smoothness of changes in the trigger availability score. When a sharp drop in signal quality is detected, the system activates an emergency weighting mechanism, automatically limiting the trigger availability score to below 0.4. The final output trigger availability score is updated in real time at a resolution of 0.01 and visualized through a human-machine interface. The actual effect of each trigger is recorded (assessed through image quality). After accumulating a certain number of new samples, an incremental learning process is automatically initiated to fine-tune the model parameters, gradually adapting it to the specific equipment characteristics and patient demographics of the hospital.

[0091] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application transforms complex physiological signals and system states into a single, operable trigger availability score through systematic feature engineering and machine learning inference. This achieves a leap from multi-parameter monitoring to intelligent decision-making and provides a reliable basis for adaptive triggering in different clinical scenarios.

[0092] To automatically select the optimal triggering strategy based on different confidence levels and balance image quality and success rate, this application provides a method for CT cardiac quiescent phase detection and imaging triggering based on millimeter-wave radar, see [link to relevant documentation]. Figure 2Step 600 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 610: If the trigger availability score is greater than or equal to a preset first threshold, then select the mid-to-late diastolic phase as the target cardiac phase as the first triggering strategy.

[0093] Specifically, when the system calculates a trigger availability score of 0.7 or higher for one or more cardiac cycles, it determines that the current signal quality is excellent and the physiological state is stable, thus entering a high-confidence working mode. At this point, the first triggering strategy is selected: the mid-to-late diastolic phase, where the heart has the longest relative rest time, is chosen as the target phase. This corresponds to the 70%-80% stage of the cardiac cycle, during which the heart is in full diastole after contraction, ventricular motion amplitude is minimal, and blood flow is relatively still, which is most conducive to obtaining high-quality CT images.

[0094] The parameter configuration for the first triggering strategy can include: (1) The trigger window width is dynamically adjusted according to the signal-to-noise ratio, ranging from 60 to 80 milliseconds; (2) The target phase ratio p is set to 0.75 (i.e., the phase point 75% after the R wave); (3) A single, precise trigger window is used to pursue the optimal image temporal resolution; (4) Breathing coordination requirements are relaxed and can be triggered throughout the entire respiratory cycle.

[0095] In other words, the first trigger strategy is suitable for ideal examination conditions with a stable heart rate (60-80 bpm), regular breathing, and good signal quality (SNR>20dB), and is commonly seen in young and middle-aged patients with high cooperation.

[0096] Step 620: If the trigger availability score is less than the first threshold and the trigger availability score is greater than or equal to the second threshold, then select the second triggering strategy with end-systole as the target cardiac phase; wherein, the first threshold is greater than the second threshold.

[0097] Specifically, when the trigger availability score fluctuates between 0.4 and 0.7, or when signal quality shows a slight decrease but remains acceptable, the system switches to a robust operating mode. At this point, a second triggering strategy is selected: using the end of systole as the target phase, corresponding to the 35%-45% phase of the cardiac cycle. Although the resting window in this phase is relatively short, it is relatively stable when the heart rate is fast or arrhythmic, making it a reliable alternative.

[0098] The parameter configuration for the second triggering strategy can include: (1) The trigger window width is appropriately increased to 80-100 milliseconds; (2) The target phase ratio p is set to 0.40 (i.e., the phase point 40% after the R wave). (3) Enable the prediction error compensation mechanism to improve trigger robustness; (4) There are basic requirements for respiratory coordination to avoid obvious end-inspiratory triggering.

[0099] In other words, the second triggering strategy is suitable for moderate heart rate (80-100 bpm), mild arrhythmia, and moderate signal quality (SNR=10-20dB), and is commonly seen in elderly patients or those with mild poor cooperation.

[0100] And, step 630: if the trigger availability score is less than the second threshold, or if a continuous irregular rhythm is detected, then a third triggering strategy is selected, which involves opening multiple trigger windows on multiple cardiac phases and limiting triggering to the end of expiration.

[0101] Specifically, when the trigger availability score remains below 0.4, or more than three consecutive abnormal RR intervals are detected (difference > 25%), or the uncertainty remains above 0.3 radians, the system enters a safety-priority mode. In this case, the third triggering strategy is selected: multiple trigger windows are opened across multiple cardiac phases and limited to triggering at the end of expiration, including: (1) Multi-phase trigger window: The system sets 2-4 overlapping trigger windows within a single cardiac cycle, covering the end of systole (30%-40%), early diastole (50%-60%), and mid-diastole (70%-80%). Through the multi-location waiting strategy, it ensures that usable images can still be captured under complex conditions.

[0102] (2) Respiratory cycle limitation: The triggering opportunity is strictly limited to the end of expiration (the last 1 / 3 of the respiratory cycle), when the range of chest movement is minimal, which can minimize respiratory motion artifacts.

[0103] (3) Image selection mechanism: During the same cardiac cycle, the system automatically selects the best image for diagnosis based on the image quality control score (such as sharpness, noise level, and artifact degree).

[0104] The parameter configuration for the third triggering strategy can include: (1) Set the width of each child window to 100-120 milliseconds; (2) Use conservative prediction parameters to reduce the sensitivity to heart rhythm tracking; (3) Implement strict security measures; triggering will be immediately suspended if any uncertainty exceeds the limit. (4) The respiratory monitoring sensitivity is set to the highest level, and even slight body movement will trigger a reassessment.

[0105] In other words, the third trigger strategy is suitable for difficult situations such as high heart rate (>100 bpm), significant arrhythmia (e.g., atrial fibrillation, frequent premature beats), poor signal quality (SNR<10dB), or poor patient cooperation.

[0106] In simple terms, triggering strategies can include: If the availability score x ≥ 0.7 is triggered, select the first trigger strategy: mid-to-late expansion window (70-80% of R-R), with the window width scaled according to SNR; If 0.4 ≤ x < 0.7, then the second triggering strategy is selected: shrink the end window (35–45% of R–R) to take into account the high heart rate population; If x < 0.4 or there is a continuous irregular rhythm, then the third triggering strategy is selected: multi-window prediction triggering + end-expiratory limitation (triggered only at the end of expiration).

[0107] In addition, the system can monitor status changes in real time and automatically switch strategies upwards (e.g., from the third trigger strategy to the second trigger strategy) when conditions improve, achieving dynamic adaptive optimization. For example, if the confidence level of the discriminant model is <0.6, it can safely fall back to the manufacturer's default cardiac phase (e.g., 75% RR) or "no trigger + post-reconstruction". The execution of "no trigger + post-reconstruction" can include: stopping all real-time trigger command issuance, switching to continuous spiral scanning mode to collect raw data, and after the scan, performing image-based retrospective ECG-gated reconstruction on the workstation based on the acquired ECG signals or the quality assessment of the reconstructed images, selecting the image sequence with the fewest motion artifacts from multiple cardiac phases. This approach is suitable for severe arrhythmias or other extreme situations. All strategy switching is clearly prompted on the interface for operator confirmation or intervention. This safe fallback mechanism ensures that the system can maintain basic functions even under extreme abnormal conditions, preventing complete scan failure due to algorithm failure, and significantly improving the clinical usability and robustness of the entire solution.

[0108] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application can achieve intelligent degradation from pursuing the best image to ensuring basic success, maintain a high triggering success rate under various complex conditions, and significantly improve the robustness of the system.

[0109] To overcome the effects of system latency and physiological variability, achieve millisecond-level precision in forward-looking triggering, and ensure triggering safety, this application provides a method for CT cardiac quiescent phase detection and imaging triggering based on millimeter-wave radar, see [link to relevant documentation]. Figure 2 Step 700 in the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar specifically includes the following: Step 710: Send a test trigger command to the CT machine and record the response timestamp of the CT machine. Calculate the time difference between the time of sending the test trigger command and the response timestamp, and estimate the current system delay based on the time difference using a rolling average algorithm.

[0110] Specifically, the delay calibration process can be automatically executed before the scan begins and during the scan intervals. The process involves sending high-frequency test trigger pulses (typically 5-10 times) to the CT scanner via a dedicated test thread, simultaneously recording the precise moment the test trigger command is sent, i.e., the command issuance timestamp t. send When a CT scanner receives a trigger command and begins exposure, it returns the CT scanner's response timestamp, which is also the actual exposure timestamp t, through its status interface. exposure The system calculates the single-transaction delay τ. single =t exposure -t send The system delay is then smoothed using a rolling average algorithm with a window size of 8 to obtain the current optimal system delay. This can also be called the system latency estimate. This process runs continuously and can adapt to latency fluctuations caused by changes in the CT system state. In other words, by sending test triggers, recording CT response timestamps, and combining this with interface protocol latency estimation... The initial value is calculated and averaged to accurately map the "target time in the physiological phase" to the CT-executable trigger time.

[0111] Step 720: Based on historical RR interval data, a hybrid prediction model combining weighted moving average and median filtering is used to predict the RR interval of the next cardiac cycle; wherein, the RR interval of the next cardiac cycle is obtained by weighted summation of the median of the most recent historical RR interval data and multiple other historical RR interval data, and the weight coefficient of the most recent historical RR interval is between 0.4 and 0.7.

[0112] Specifically, a historical RR interval buffer with a capacity of 16 can be maintained, employing an innovative hybrid prediction model. This model comprehensively considers both the latest trends in heart rate changes and long-term stability: it uses the most recently measured historical RR interval data... As a short-term trend term, the median of the four historical RR intervals from k-2 to k-5 is taken as the long-term stable benchmark. The RR interval for the next cardiac cycle is then predicted. The weighting coefficient α is dynamically adjusted based on heart rate variability. When the heart rhythm is stable, α=0.7 focuses on tracking the latest changes; when the heart rhythm is irregular, α=0.4 focuses on historical stability to achieve adaptive prediction.

[0113] Step 730: Using the previous reference time as the starting point, add the product between the target phase ratio of the cardiac resting period corresponding to the triggering strategy and the RR interval of the next cardiac cycle, and then subtract the current system delay to obtain the advance triggering time.

[0114] Specifically, based on the selected triggering strategy, the system obtains the corresponding target phase ratio p during the cardiac resting phase (e.g., p=0.75 for the diastolic strategy and p=0.40 for the systolic strategy). This is referenced to time R. Using the time origin, calculate the advance trigger time. This calculation ensures that the CT exposure command, after a delay, is executed precisely at the target cardiac phase.

[0115] When step 730 calculates the theoretical advance triggering time Then, the system immediately calculates the actual window width used for this trigger, centered on the target phase (such as mid-to-late diastole or end-systole), based on the current signal-to-noise ratio and phase uncertainty. , where g(·) is a monotonically increasing function (SNR high, (A smaller window means a narrower window), clip is for upper and lower limit clipping. Specifically, you can read the baseline window width corresponding to the current triggering strategy. (such as diastolic strategies) =80ms, contraction phase strategy =60ms), will monitor the SNR and Substitute into function g(SNR, To ensure the best signal quality and the most certain phase, the calculated window width is narrower, aiming for the highest image quality. The clip function limits the final window width to a physically feasible range of 40ms to 120ms.

[0116] In other words, the final output of step 730 can be a... A trigger time window centered on a point with a width of W. - W / 2, + W / 2].

[0117] Step 740: If the absolute value of the prediction error between the RR interval of the next cardiac cycle and the RR interval measured after the end of the next cardiac cycle is less than or equal to a preset error threshold, and the current uncertainty is less than or equal to a preset uncertainty threshold, then it is determined that the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, and a scan trigger command is sent to the CT machine at the early trigger time to acquire the CT image data of the target user.

[0118] Specifically, 5-10ms before the expected trigger time, the system performs a final security verification: (1) Verification of prediction accuracy: The RR interval measured after the end of the next cardiac cycle (i.e., the next actual RR interval) k Calculate the prediction error immediately when available. - RR k |, and must not exceed the preset threshold ε=±5%RR; (2) Reliability verification of estimation: check the uncertainty σ φ Is it below the threshold σ? threshold =0.3 radians.

[0119] The system determines that the security protection condition has passed only if all of the above conditions are met simultaneously, based on precise calculations. The system continuously sends scan trigger commands to the CT scanner to initiate the image acquisition process. During trigger execution, the system does not calculate only one trigger window, but rather calculates 2-4 trigger windows located at different cardiac phases (e.g., end-systole, early diastole, mid-diastole). The system attempts to trigger the CT scan within all these windows, thereby acquiring multiple sets of image data within the same heartbeat cycle. After the scan is completed, the system calls an image quality control algorithm to score these sets of images (e.g., based on sharpness, noise, and artifact levels), and automatically votes to select the best image for diagnostic use. In other words, for detected irregular rhythms, multi-window predictive triggering (2-4 overlapping small windows) is used, and after imaging, the best image is retained based on quality control scoring.

[0120] And, step 750: if the absolute value of the prediction error is greater than the error threshold, or if the current uncertainty is greater than the uncertainty threshold, then pause the triggering of the current period and postpone the re-evaluation, or execute the downgrade of the triggering strategy.

[0121] Specifically, the handling mechanism when any security protection indicator exceeds the standard is as follows: (1) Prediction error exceeds the standard (| - RR k |>5%RR): This indicates an unexpected and drastic change in heart rate. The system immediately pauses the current cycle triggering and waits for the next cardiac cycle to reassess.

[0122] (2) Phase uncertainty exceeds the limit (σ φ >0.3 radians): This indicates that the current signal quality is poor, the phase estimation is unreliable, and the system executes a downgraded triggering strategy—such as downgrading from high-precision diastolic single-window triggering to robust multi-window triggering, or switching to a safe backoff mode in extreme cases.

[0123] Where, when σ φWhen the phase uncertainty is large (high), the function g(SNR, σ) φ The output decreases, causing the window width W to automatically widen. This is a passive, built-in degradation strategy that relaxes the requirements in step 730 to ensure that the quiescent period can still be captured when the signal is unreliable. If σ φ If the value exceeds the security protection threshold in step 740, the process in step 750 will be automatically triggered to pause the triggering or forcibly switch to a more conservative multi-window strategy.

[0124] When both prediction error and uncertainty exceed the limits, the system enters a high-alert state, suspends triggering, and prompts the operator for intervention. After handling the anomaly, the system continuously monitors the status indicators. When the indicators return to normal for three consecutive cycles, it automatically reverts to the original triggering strategy, achieving self-healing.

[0125] Therefore, the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar provided in this application can form a complete closed-loop control system through delay calibration, hybrid prediction model and safety protection mechanism, so as to achieve accurate triggering under the premise of ensuring safety and effectively avoid false triggering and trigger failure.

[0126] From a software perspective, this application also provides a millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering device for executing all or part of the millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method, see [link to relevant documentation]. Figure 3 The CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar specifically includes the following components: The dual-channel signal acquisition module 10 is used to acquire reference channel signals of metal interference sources in the space where the CT scanning bed is located in a non-contact manner based on a millimeter-wave radar array, and to acquire chest wall micro-motion observation signals of the target user located on the CT scanning bed in a non-contact manner.

[0127] The adaptive interference cancellation module 20 is used to perform reference channel adaptive cancellation processing on the chest wall micro-motion observation signal according to the reference channel signal, so as to obtain the corresponding chest wall micro-motion purification signal.

[0128] The vital signs separation module 30 is used to acquire the estimated respiratory rate and initial heartbeat signal corresponding to the chest wall micro-movement purification signal, respectively.

[0129] The instantaneous phase estimation module 40 is used to construct the analytical signal corresponding to the initial heartbeat signal, and to obtain the instantaneous heartbeat phase and the uncertainty of the instantaneous heartbeat phase of the analytical signal.

[0130] The intelligent scoring and decision module 50 is used to extract multi-dimensional feature vectors from the instantaneous cardiac phase, the initial heartbeat signal and the estimated respiratory rate, and to determine the trigger availability score for the CT scanner based on the multi-dimensional feature vectors.

[0131] The adaptive strategy mapping module 60 is used to select the corresponding triggering strategy based on the trigger availability score.

[0132] The look-ahead trigger execution module 70 is used to calibrate the system delay online and predict the RR interval of the next cardiac cycle based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, it determines the advance triggering time. When the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, the module sends a scan trigger command to the CT machine at the advance triggering time to acquire the CT image data of the target user.

[0133] The embodiments of the CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar provided in this application can be used to execute the processing flow of the embodiments of the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar in the above embodiments. Its functions will not be repeated here, but can be referred to the detailed description of the embodiments of the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar.

[0134] The portion of the millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering device that performs the millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering can be completed in either a server or client device. The choice can be made based on the processing power of the client device and the limitations of the user's usage scenario. This application does not impose any limitations in this regard. If all operations are completed in the client device, the client device may further include a processor for the specific processing of the millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering.

[0135] The aforementioned client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side; in other implementation scenarios, it may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a distributed server structure.

[0136] The server and the client device can communicate using any suitable network protocol, including those not yet developed as of the date of this application. Such network protocols may include, for example, TCP / IP, UDP / IP, HTTP, HTTPS, etc. Furthermore, such network protocols may also include RPC (Remote Procedure Call Protocol) and REST (Representational State Transfer Protocol) protocols used on top of the aforementioned protocols.

[0137] As described above, the CT cardiac quiescent phase detection and imaging triggering device based on millimeter-wave radar provided in this application achieves non-contact, electrode-free chest wall micro-motion acquisition via millimeter-wave radar, effectively overcoming the limitations of ECG gating and patient discomfort. Through an anti-interference mechanism with adaptive cancellation of the reference channel, it can stably extract sub-millimeter-level heartbeat signals in the strong scattering environment of a CT room. Utilizing instantaneous phase estimation and uncertainty quantification, it can provide a high-precision time reference and reliability index for the cardiac cycle. Based on multi-dimensional feature-based trigger availability scoring and adaptive strategy selection, it can intelligently respond to individual differences and physiological changes. Through online system delay calibration and RR interval prediction, it can achieve forward-looking and accurate triggering, effectively preventing false triggering combined with safety protection conditions. The entire solution can significantly improve the quality of cardiac CT images, reduce motion artifacts, increase scan success rate and clinical work efficiency, and provide reliable technical support for cardiac imaging diagnosis.

[0138] To further illustrate the above embodiments, this application also provides a specific application example of a CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar, which is executed by a CT cardiac quiescent period detection and imaging triggering system based on millimeter-wave radar. The signal processing and control unit in this system can be the CT cardiac quiescent period detection and imaging triggering device based on millimeter-wave radar mentioned in the foregoing embodiments.

[0139] This application example focuses on cardiac quiescent phase identification and scan control in CT imaging, and proposes systematic solutions to the following CT scenario-specific problems (1) to (4): (1) Strong scattering and multipath of the rack / bed: The metal structure and rotating parts generate strong echoes, which drown out the cardiac micro-Doppler signal; (2) Cardiac-respiratory coupling and body movement: The respiratory fundamental frequency and its harmonics cause the cardiac phase curve to fluctuate, and the resting phase positioning is unstable; (3) Individual differences and irregular rhythm: BMI / chest wall thickness, heart rate range and occasional premature beats / atrial fibrillation, etc., cause the fixed threshold / window width to fail; (4) Differences in timing and interface across devices: There is a total delay in the radar to CT control link, and the timing of the interfaces of different manufacturers is different.

[0140] The application example of this application constructs a millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering system that integrates millimeter-wave radar micro-motion sensing and intelligent signal processing. This system can stably extract cardiac motion features in complex imaging environments and achieve high-precision imaging phase control, thereby effectively suppressing cardiac motion artifacts and improving CT image quality and diagnostic reliability.

[0141] The CT cardiac quiescent phase detection and imaging triggering system based on millimeter-wave radar consists of a millimeter-wave MIMO radar, a reference channel / sensor module, a signal processing and control unit, a CT interface module, and a human-machine interface. The methods include radar anti-scattering acquisition, joint estimation of instantaneous phase and uncertainty, patient feature-driven strategy selection, online system delay calibration, and prospective predictive triggering, as detailed below: 1) Radar front-end: FMCW millimeter-wave radar, MIMO array (total virtual array elements of transmit / receive antennas ≥ 8), supports polarization selection; 2) Reference channel: Used to collect static or quasi-static scattering from the rack / bed, and can be a passive corner reflector or a fixed scatterer channel; 3) Signal processing and control unit: performs beamforming, reference cancellation, phase estimation, feature calculation, strategy mapping, timing compensation, and triggering; 4) CT Interface Module: Interacts with the CT scanner's digital trigger / Ethernet control port and records response timestamps; 5) Human-machine interface: Displays signal result indicators and trigger status.

[0142] Based on the above, see Figure 4 The adaptive identification and scan triggering method for cardiac quiescent phase based on millimeter-wave radar provided in this application example specifically includes the following steps: S1 Arrangement and Layout: The millimeter-wave radar antenna is located at the window of the rack, with the main lobe pointing towards the front wall of the chest. Orthogonal polarization is used to reduce metal scattering coupling.

[0143] S2 pre-scan self-calibration: Acquire static noise and reference channel signals to estimate rack / bed echo.

[0144] S3 Anti-scattering Acquisition: Performs "zero-dip adaptive beamforming and reference channel adaptive cancellation" to suppress strong scattering and multipath, obtaining target echoes dominated by the chest wall.

[0145] S4 Decoupling and Phase Tracking: First, estimate the respiratory dominant frequency and perform time-varying notch filtering. Then, construct an analytical signal from the chest wall displacement signal and perform Kalman filtering to obtain the instantaneous cardiac phase and its uncertainty.

[0146] S5 Feature Extraction: Real-time calculation of HR, HRV, BR, SNR, chest wall attenuation amplitude, rhythm irregularity index, etc., to form a multi-dimensional feature vector z; through x=σ(wᵀz+b) mapping, the trigger availability score is obtained, which is used to automatically select the "resting period window / trigger" strategy and window width.

[0147] S6 Timing Compensation and Look-Ahead Triggering: Online calibration of system delay and prediction of the next RR interval, at the early triggering time. The trigger is issued appropriately in advance; the response timestamp is recorded through the CT interface for continuous calibration.

[0148] S7 Human-Machine Interface: The interface displays the trigger status in real time, aligned with the CT response, highlighting the personnel movement status (stable / minor movement / significant body movement) and risk warnings; it summarizes core indicators to determine trigger availability and the current trigger strategy; it provides one-click intervention (pause / resume, manual trigger, resampling reference, strategy switching) when body movement or quality degradation is detected, and sets guard conditions to prevent false triggering.

[0149] In summary, this application example utilizes FMCW millimeter-wave radar combined with MIMO arrays and polarization selection to acquire chest wall micro-movements without contact with the subject. By setting a reference channel and jointly processing it with the target channel, static or quasi-static scattering from the gantry / bed is suppressed, improving the signal-to-noise ratio and directivity for chest wall displacement, thus simultaneously supporting cardiac gating and body motion monitoring. The cardiac phase and its uncertainty are robustly estimated from the millimeter-wave displacement signal. Combined with online calibration of system link delay and RR interval prediction, a trigger command is issued before the target phase arrives, achieving precise gating with "controllable lead time" to reduce cardiac-related artifacts and improve trigger success rate. HR / HRV, respiratory rate, SNR, phase uncertainty, and body motion index are used to construct a feature vector, which is mapped to obtain a trigger availability score. When the score or threshold exceeds the limit, the system automatically enters a guard state and executes a fallback strategy (such as pausing triggering, switching to the default cardiac phase, or only acquiring without triggering) to avoid imaging failure caused by false triggering. It supports both hardware digital trigger interface and Ethernet control / status readback interface, and records CT side response timestamps; by estimating and continuously correcting end-to-end delay, it achieves a unified time base and traceable recording for trigger issuance and CT response, ensuring accurate and reliable gating timing.

[0150] Based on this, the application example of this application uses non-contact millimeter-wave radar to acquire minute displacements of the chest wall, extract stable cardiac phases, and send gating / trigger signals to CT in real time. This significantly reduces motion artifacts caused by cardiac and non-periodic body movements, improving the imaging clarity and interpretability of detailed structures such as the heart and coronary arteries. By suppressing fixed scattering through a reference channel, using MIMO and polarization selection to reduce multipath interference, and combining temporal compensation and prospective prediction to improve trigger accuracy and robustness, a high trigger success rate and image stability can be maintained even under conditions of high heart rate, irregular rhythm, or poor respiratory coordination. At the clinical and workflow level, the system displays the trigger status and personnel movement (stable / minor / significant body movement) in real time through an intuitive human-machine interface, and provides one-click intervention and rollback strategies, facilitating quick decision-making by operators and reducing failures and interruptions. The device interfaces with CT scanners via a standard digital trigger / Ethernet interface and records response timestamps. It is low-cost to modify, easy to integrate and maintain; it reduces rescanning and waiting time, which helps to reduce radiation dose and contrast agent usage, save consumables and labor costs, increase examination throughput and patient experience, and is suitable for a variety of populations including adults, the elderly and pediatricians.

[0151] This application also provides an electronic device, which may include a processor, a memory, a receiver, and a transmitter. The processor is used to execute the CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar mentioned in the above embodiments. The processor and memory can be connected via a bus or other means, taking a bus connection as an example. The receiver can be connected to the processor and memory via wired or wireless means.

[0152] The processor can be a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of the above types of chips.

[0153] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar in the embodiments of this application. The processor executes various functional applications and data processing by running the non-transitory software programs, instructions, and modules stored in the memory, thereby implementing the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar in the above method embodiments.

[0154] The memory may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor, etc. Furthermore, the memory may include high-speed random access memory and non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the processor, which can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0155] The one or more modules are stored in the memory, and when executed by the processor, the CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar in the embodiment is executed.

[0156] In some embodiments of this application, the user equipment may include a processor, a memory, and a transceiver unit. The transceiver unit may include a receiver and a transmitter. The processor, memory, receiver, and transmitter may be connected via a bus system. The memory is used to store computer instructions, and the processor is used to execute the computer instructions stored in the memory to control the transceiver unit to send and receive signals.

[0157] As one implementation method, the functions of the receiver and transmitter in this application can be implemented by transceiver circuits or dedicated transceiver chips, and the processor can be implemented by dedicated processing chips, processing circuits or general-purpose chips.

[0158] As another implementation approach, the server provided in this application embodiment can be implemented using a general-purpose computer. That is, the program code implementing the processor, receiver, and transmitter functions is stored in memory, and the general-purpose processor implements the processor, receiver, and transmitter functions by executing the code in memory.

[0159] This application also provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the steps of the aforementioned millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.

[0160] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the aforementioned millimeter-wave radar-based CT cardiac quiescent phase detection and imaging triggering method.

[0161] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. The programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave.

[0162] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0163] In this application, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0164] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to the embodiments of this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for detecting and triggering cardiac quiescent phase in CT imaging based on millimeter-wave radar, characterized in that, include: Based on millimeter-wave radar array, the reference channel signal of the metal interference source in the space where the CT scanning table is located is acquired non-contactly, rather than acquiring the chest wall micro-motion observation signal of the target user located on the CT scanning table. Based on the reference channel signal, the chest wall micromotion observation signal is subjected to reference channel adaptive cancellation processing to obtain the corresponding chest wall micromotion purification signal. Obtain the estimated respiratory rate and initial heart rate signal corresponding to the chest wall micro-motion purification signal, respectively; Construct the analytical signal corresponding to the initial heartbeat signal, and obtain the instantaneous cardiac phase and the uncertainty of the instantaneous cardiac phase of the analytical signal; Multi-dimensional feature vectors are extracted from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate, and a trigger availability score for the CT scanner is determined based on the multi-dimensional feature vectors. Select the corresponding triggering strategy based on the trigger availability score; The system delay is calibrated online, and the RR interval of the next cardiac cycle is predicted based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, an early triggering time is determined. When the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, a scan triggering command is sent to the CT machine at the early triggering time to acquire the CT image data of the target user.

2. The method for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar according to claim 1, characterized in that, The non-contact acquisition of reference channel signals from metallic interference sources within the space where the CT scan table is located, based on a millimeter-wave radar array, and the non-contact acquisition of chest wall micro-motion observation signals from the target user located on the CT scan table, include: Based on a millimeter-wave radar array, a reference channel signal of a metal interference source in the space where the CT scanning bed is located is acquired non-contactly. The metal interference source includes: metal parts on the CT gantry or CT scanning bed. The main lobe of the millimeter-wave radar array is directed towards the chest wall of the target user located on the CT scanning table. Based on the millimeter-wave radar array, the raw chest wall micro-motion signal of the target user located on the CT scanning table is acquired non-contactly. Null adaptive beamforming processing is performed on the raw chest wall micro-motion signal to spatially suppress scattering interference from the metal structure by forming a null in the direction of the interference source, thereby obtaining the chest wall micro-motion observation signal of the target user.

3. The method for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar according to claim 1, characterized in that, The step of performing reference channel adaptive cancellation processing on the chest wall micromotion observation signal based on the reference channel signal to obtain the corresponding chest wall micromotion purification signal includes: The reference channel signal is input into an adaptive filter so that the adaptive filter estimates the interference contribution of the reference channel signal to the chest wall micromotion observation signal based on the current filter coefficients, and obtains the chest wall micromotion purification signal by calculating the residuals; wherein, the filter coefficients are updated using the normalized least mean square algorithm.

4. The CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar according to claim 1, characterized in that, The acquisition of the estimated respiratory rate and initial heart rate signals corresponding to the chest wall micro-motion purification signals includes: The respiratory rate is estimated from the chest wall micro-movement purification signal; Construct a time-varying bandstop filter whose stopband center frequency dynamically corresponds to the estimated breathing frequency and its harmonics. The time-varying band-stop filter is used to filter the chest wall micro-motion purification signal to remove the respiratory component and its harmonics, thereby obtaining the initial heartbeat signal.

5. The CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar according to claim 1, characterized in that, The process of constructing the analytical signal corresponding to the initial heartbeat signal and obtaining the instantaneous cardiac phase and uncertainty of the instantaneous cardiac phase includes: Perform a Hilbert transform on the initial heartbeat signal to construct the analytic signal corresponding to the initial heartbeat signal; Calculate the phase angle of the analyzed signal and perform phase unwrapping to obtain the unwrapped instantaneous phase; The unwrapped instantaneous phase is smoothed and optimally estimated using a Kalman filter to obtain the smoothed instantaneous cardiac phase and its uncertainty.

6. The CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar according to claim 1, characterized in that, Before the non-contact acquisition of chest wall micromotion observation signals of the target user located on the CT scanning table, the following is also included: The ambient noise in the space where the currently unmanned CT scanning bed is located is collected based on a millimeter-wave radar array; Correspondingly, the extraction of multi-dimensional feature vectors from the instantaneous cardiac phase, the initial heartbeat signal, and the estimated respiratory rate, and the determination of the trigger availability score for the CT scanner based on the multi-dimensional feature vectors, includes: The heart rate value, heart rate variability value, and PVC / AF segment ratio of the rhythm irregularity index are extracted from the instantaneous cardiac phase. The chest wall attenuation amplitude is extracted from the initial heartbeat signal. The signal-to-noise ratio is calculated based on the initial heartbeat signal and the environmental noise. The respiratory rate value is extracted from the respiratory rate estimate to form a multi-dimensional feature vector containing a preset initial system delay value, the heart rate value, heart rate variability value, PVC / AF segment ratio of the rhythm irregularity index, chest wall attenuation amplitude, signal-to-noise ratio, and respiratory rate value. The multi-dimensional feature vector is input into a pre-defined discrimination model so that the discrimination model outputs a value between 0 and 1 as a trigger availability score for the CT scanner.

7. The method for detecting and triggering cardiac quiescent phase in CT based on millimeter-wave radar according to claim 1, characterized in that, The step of selecting the corresponding triggering strategy based on the trigger availability score includes: If the trigger availability score is greater than or equal to a preset first threshold, then the first triggering strategy with mid-to-late diastole as the target cardiac phase is selected. If the trigger availability score is less than the first threshold and the trigger availability score is greater than or equal to the second threshold, then a second triggering strategy with end-systole as the target cardiac phase is selected; wherein, the first threshold is greater than the second threshold. If the trigger availability score is less than the second threshold, or if a continuous irregular rhythm is detected, a third triggering strategy is selected, which involves opening multiple trigger windows on multiple cardiac phases and limiting triggering to the end of expiration.

8. The CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar according to claim 7, characterized in that, The online calibration system delays and predicts the RR interval of the next cardiac cycle based on historical RR interval data. Based on the target phase ratio of the cardiac quiescent period corresponding to the triggering strategy, the RR interval of the next cardiac cycle, and the system delay, it determines an early triggering time. When the uncertainty and the RR interval of the next cardiac cycle meet preset safety protection conditions, a scan trigger command is issued to the CT scanner at the early triggering time to acquire the target user's CT image data, including: A test trigger command is sent to the CT scanner and the response timestamp of the CT scanner is recorded. The time difference between the time when the test trigger command is sent and the response timestamp is calculated. Based on the time difference, the current system latency is estimated using a rolling average algorithm. Based on historical RR interval data, a hybrid prediction model combining weighted moving average and median filtering is used to predict the RR interval of the next cardiac cycle. The RR interval of the next cardiac cycle is obtained by weighted summation of the median of the most recent historical RR interval data and multiple other historical RR interval data, and the weight coefficient of the most recent historical RR interval is between 0.4 and 0.

7. Starting from the previous reference time, add the product between the target phase ratio of the cardiac resting period corresponding to the triggering strategy and the RR interval of the next cardiac cycle, and then subtract the current system delay to obtain the advance triggering time; If the absolute value of the prediction error between the RR interval of the next cardiac cycle and the RR interval measured after the end of the next cardiac cycle is less than or equal to a preset error threshold, and the current uncertainty is less than or equal to a preset uncertainty threshold, then it is determined that the uncertainty and the RR interval of the next cardiac cycle meet the preset safety protection conditions, and a scan trigger command is sent to the CT machine at the early trigger time to collect the CT image data of the target user; If the absolute value of the prediction error is greater than the error threshold, or if the current uncertainty is greater than the uncertainty threshold, then the triggering of the current cycle is paused and the re-evaluation is postponed, or the triggering strategy is downgraded.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the CT cardiac quiescent period detection and imaging triggering method based on millimeter-wave radar as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the CT cardiac quiescent phase detection and imaging triggering method based on millimeter-wave radar as described in any one of claims 1 to 8.

Images