Ophthalmic nerve scanning monitoring method and system based on eye movement gating

Abstract

The invention provides an optic nerve scanning monitoring method and system based on eye movement gating, and relates to the technical field of optic nerve scanning monitoring. The method comprises the steps of judging whether an acquisition condition of an optic nerve magnetic resonance image is met or not based on an eye movement event obtained through image acquisition of an eye tracker and eye movement recognition, and if yes, generating a feedback signal; triggering MR sequence acquisition according to the feedback signal; marking the eye movement event in the k space data to obtain eye movement data marking information; and correcting the k-space data according to the eye movement data mark information to obtain corrected data, and performing image reconstruction on the corrected data by using an MR reconstruction machine. Through the real-time monitoring and correction technology based on eye movement gating, the image quality of optic nerve imaging is remarkably improved, the influence of motion artifacts is reduced, and therefore more accurate and reliable image support is provided for clinical diagnosis.

Description

Technical Field

[0001] The present invention relates to the technical field of optic nerve scan monitoring, and particularly to a method and system for optic nerve scan monitoring based on eye movement gating. Background Art

[0002] The motion artifact suppression technology based on K-space data correction is a currently widely used clinical solution. Existing motion correction technologies based on K-space trajectory optimization (such as Propeller, Blade, etc.) mainly adopt segmented radial or spiral K-space filling strategies, and achieve motion compensation by repeatedly acquiring the central K-space region and using data redundancy. Although such technologies can reduce stripe artifacts within the conventional motion range, their algorithms rely on the assumption of temporal correlation of adjacent data blocks. When the eyeball undergoes non-periodic tremors (such as horizontal tremor frequency > 3 Hz) or the proportion of the rotational component exceeds 30%, the phase difference between data blocks will exceed the algorithm correction range, resulting in annular artifacts or local signal distortion in the reconstructed image.

[0003] Existing motion correction technologies usually need to extend the scanning time to 1.5 - 2 times that of the conventional FSE sequence to improve data redundancy. (Taking the acquisition of the oblique sagittal T2WI of the optic nerve by a 3T MRI system as an example, the conventional FSE takes about 3 minutes and 20 seconds, while the Propeller technology needs to be extended to 5 minutes and 50 seconds). This time extension leads to two key problems: First, the decrease in patient compliance makes it easier to form a vicious cycle of increased motion in the later scanning stage; second, for cases of nystagmus that require dynamic evaluation (such as the acute stage of multiple sclerosis), the reduction in time resolution will mask the time-varying characteristics of the motion pattern.

[0004] The real-time gating technology based on navigator echoes can also be theoretically used for optic nerve imaging. Its principle is to insert a fast gradient echo module in the sequence interval to monitor the eyeball position. However, this method faces two major technical bottlenecks: On the one hand, the acquisition of navigator echoes requires a time window of 10 - 15 ms, resulting in the forced shortening of the echo train length (ETL) of the traditional TSE sequence. To ensure the signal-to-noise ratio, the number of signal averages needs to be increased, and finally the single-scan time is extended to 8 - 10 minutes; on the other hand, the anatomical adjacent relationship between the extraocular muscles and the optic nerve causes the navigator echo signal to be easily interfered by electromyographic activity, and the signal false trigger rate is relatively high, seriously affecting the scanning efficiency. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a method for optic nerve scan monitoring based on eye movement gating. Through real-time monitoring and correction technologies based on eye movement gating, the image quality of optic nerve imaging is significantly improved, and the influence of motion artifacts is reduced, thereby providing more accurate and reliable image support for clinical diagnosis.

[0006] To achieve the above object, the present invention provides the following solutions: A method for optic nerve scan monitoring based on eye movement gating, comprising: Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determining whether the acquisition conditions for magnetic resonance imaging of the optic nerve are met. If so, generating a feedback signal; Triggering the MR sequence acquisition according to the feedback signal; Annotating the eye movement events in the k-space data to obtain eye movement data marking information; Correcting the k-space data according to the eye movement data marking information to obtain corrected data, and performing image reconstruction on the corrected data using an MR reconstructor; Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determining whether the acquisition conditions for magnetic resonance imaging of the optic nerve are met. If so, generating a feedback signal, including: Taking as a judgment window, and statistically calculating the proportion of the eye movement events and time within the judgment window; is the threshold parameter of the judgment window; Judging whether the number of saccade and blink eye movement events within the judgment window is less than a preset threshold, or whether the time proportion is less than a preset ratio. If so, it is considered that the acquisition conditions are met, and a feedback signal for consenting to acquisition is sent to the main control device of the MR system; If not, it is considered that the acquisition conditions are not met, and an instruction is sent to the main control device of the MR system to re-acquire the eye movement events.

[0007] Preferably, the preset threshold and the preset ratio are dynamically adjusted according to historical data or environmental factors.

[0008] Preferably, the transmission of the feedback signal relies on FPGA timing optimization to ensure that the scanning trigger synchronization error < 1ms; if the acquisition conditions are not met for three consecutive statistical periods, the online calibration process is automatically triggered and a user prompt is pushed, and the scanning device is frozen synchronously to avoid invalid data acquisition until the acquisition conditions are met.

[0009] Preferably, the trigger modes for triggering the MR sequence acquisition include: immediate trigger and delayed trigger; the immediate trigger starts a preset scanning sequence after receiving the feedback signal, with a response delay < 2ms; the delayed trigger is based on a deep learning eye movement prediction model to predict the fixation stable period within the next 50 - 100ms, and dynamically optimizes the scanning start time to match the best imaging window.

[0010] Preferably, during the MR sequence acquisition, it further includes: Adjust the scanning parameters according to the real-time eye movement stability index; the scanning parameters include repetition time and echo time; Preferentially call the motion-sensitive sequence, perform high b-value scanning during the stable period, and switch to the artifact-tolerant sequence during the non-stable period.

[0011] Preferably, annotate the eye movement events in the k-space data to obtain eye movement data marking information, including: Based on the PTP / NTP protocol, enable the eye tracker and the MR acquisition device to share a high-precision clock source to ensure global alignment of timestamps; Insert an eye movement event tag into the header file of the k-space data, and record the event type, start time, and duration in binary coding format to achieve millisecond-level event-scan timing association, thereby obtaining the eye movement data marking information; Dynamically associate the eye movement coordinates and the corresponding eye movement states with the acquisition moment of the k-space data through a hash index to generate a multi-dimensional mapping table.

[0012] Preferably, correct the k-space data according to the eye movement data marking information to obtain corrected data, and use an MR reconstructor to perform image reconstruction on the corrected data, including: If the acquisition time window of the k-space data overlaps with the eye movement event, mark the corresponding overlapping data block as contaminated data; For each data block marked as contaminated data, calculate the contamination confidence based on the eye movement intensity. When the contamination confidence is greater than the preset contamination threshold, discard the data block marked as contaminated data; Use adjacent uncontaminated data blocks to generate a mask matrix to reduce the diffusion effect of artifacts; For the discarded contaminated data blocks, use adjacent uncontaminated data to fill the missing regions in the k-space data; Predict the displacement of the tissue according to the eye movement vector, correct the phase-encoding direction artifacts to ensure the accuracy and clarity of the image, and obtain the corrected data; Use an MR reconstructor to perform image reconstruction on the corrected data to generate the final MR image.

[0013] Preferably, the preset contamination threshold is 0.8.

[0014] A system for optic nerve scan monitoring based on eye movement gating, for the above method, the system for optic nerve scan monitoring based on eye movement gating includes: a head coil, an eye tracker mirror, an MR acquisition device, and an eye tracker, an eye tracker main control computer, an MR system main control device, and an MR reconstruction device connected in sequence; The head coil is arranged at a designated position on the magnetic resonance bedplate; the eye tracker mirror is placed on the upper part of the head coil; the eye tracker is placed at the rear of the head coil; the head coil is used to emit radio frequency pulses to the head tissue during magnetic resonance signal acquisition to stimulate the generation of magnetic resonance signals and receive the magnetic resonance signals generated by the response of the head tissue; the eye tracker mirror is used to reflect the eye image into the field of view of the camera of the eye tracker; the MR acquisition device is used to complete the acquisition of magnetic resonance signals; the eye tracker is used to capture eye movements using the dark pupil technology in combination with a near-infrared light source and a camera to obtain eye detection information; the main control computer of the eye tracker is used to process the eye detection information recognized by the eye tracker to obtain eye movement events, generate feedback signals, and send the feedback signals to the main control device of the MR system; the main control device of the MR system is used to receive the feedback signals sent by the eye movement signal processing device, perform feedback operations, and control the acquisition of MR sequences; the MR reconstruction device is used to reconstruct the original MR signals collected in the k-space data into MR images.

[0015] The present invention discloses the following technical effects: The present invention uses an eye tracker to monitor the eye movement state of a patient in real time, comprehensively judges the collected eye movement signals through an algorithm within a fixed time window. If the acquisition requirements are met, the pulse sequence acquisition is triggered; if the judgment result does not meet the acquisition requirements, the pulse sequence acquisition is stopped. And the collected data is labeled using eye movement information for motion artifact correction during the reconstruction process. The present invention can accurately capture any minute eye movements of the patient during the scanning process, including saccades, blinks, and fixations in real time during the optic nerve scanning process in cooperation with the use of the eye tracker, and can feedback to the scanning system in real time, shield the sequence trigger within a certain period of time, and make marks in the collected data for artifact correction of the reconstruction, ultimately improving the quality of optic nerve imaging and the scanning work efficiency. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is the flowchart of the method provided by the embodiment of the present invention; Figure 2 It is the schematic diagram of the technical route provided by the embodiment of the present invention; Figure 3 It is the schematic diagram of the system structure provided by the embodiment of the present invention; Figure 4Schematic diagram of the core hardware of the eye tracker provided by the embodiment of the present invention. Detailed implementation manners

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] The purpose of the present invention is to provide a method and system for monitoring optic nerve scanning based on eye movement gating. Through real-time monitoring and correction technologies based on eye movement gating, the image quality of optic nerve imaging is significantly improved, and the influence of motion artifacts is reduced, thereby providing more accurate and reliable image support for clinical diagnosis.

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0021] Figure 1 Flowchart of the method provided by the embodiment of the present invention. As Figure 1 shown, the present invention provides a method for monitoring optic nerve scanning based on eye movement gating, including: Step 100: Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determine whether the acquisition conditions for magnetic resonance imaging of the optic nerve are met. If so, generate a feedback signal. Step 200: Trigger the MR sequence acquisition according to the feedback signal. Step 300: Mark the eye movement events in the k-space data to obtain eye movement data marking information. Step 400: Correct the k-space data according to the eye movement data marking information to obtain corrected data, and use an MR reconstructor to reconstruct an image of the corrected data. Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determine whether the acquisition conditions for magnetic resonance imaging of the optic nerve are met. If so, generate a feedback signal, including: Taking as a judgment window, count the eye movement events and time ratio within the judgment window. is the threshold parameter of the judgment window. Judge whether the number of saccade and blink eye movement events within the judgment window is less than a preset threshold, or whether the time ratio is less than a preset ratio. If so, it is considered that the acquisition conditions are met, and a feedback signal of consent to acquire is sent to the main control device of the MR system. If so, it is considered that the acquisition condition is not met, and an instruction is sent to the MR system master device to re-acquire the eye movement event.

[0022] Specifically, as Figure 2 shown, the technical route of this embodiment is as follows: Step 1: Use an eye tracker to collect eye images. The eye tracker directly collects eye images using a high-speed camera.

[0023] Specifically, based on the dark pupil effect, the eye tracker uses an active near-infrared light source projection to significantly enhance the contrast between the pupil and the iris. A high-speed camera equipped with a global shutter CMOS sensor synchronously captures dynamic eye images with microsecond-level exposure accuracy, effectively eliminating motion blur and ensuring accurate parsing of details at high frame rates.

[0024] Step 2: The main control computer of the eye tracker preprocesses the eye images collected in Step 1, and performs pupil detection and Purkinje spot localization to obtain eye movement coordinates.

[0025] Specifically, Gaussian filtering is used in preprocessing to eliminate high-frequency noise in the image, and the distinguishability between the pupil and the corneal reflection point is improved through adaptive histogram equalization.

[0026] Pupil detection includes three steps: threshold segmentation, edge extraction, and ellipse fitting to obtain the pupil center coordinates : (1) Threshold segmentation: Use the Otsu adaptive threshold method to binarize the image and segment the low-brightness pupil area; (2) Edge extraction: Use the Canny edge detection algorithm to extract the pupil contour and eliminate the interference of eyelashes and eyelids; (3) Ellipse fitting: First, use the random Hough transform to efficiently sample from the edge points of the eye image, and quickly estimate key parameters such as the center position, major axis, minor axis, and rotation angle of the ellipse; subsequently, introduce a deep learning model to dynamically optimize the preliminary fitting result, effectively overcoming the interference of complex scenarios (such as eyelash occlusion and uneven illumination) on pupil localization.

[0027] Purkinje spot tracking, separate the high-brightness reflection point through morphological opening operation, and calculate the center coordinates of the Purkinje spot based on pixel brightness weighting 。

[0028] Step 3: Convert the pupil position signal collected by the eye tracker in Step 2 into the screen coordinate system in real time and stably.

[0029] Specifically, the gaze mapping model uses a second-order polynomial to convert the pupil center coordinates in the eye tracker coordinate system and the offset of the Purkinje spot ( Map it to the screen coordinate system (X, Y), and optimize the model coefficients using the least squares method based on the calibration data of the nine calibration points that the user fixated on during pupil calibration. , Achieve high-precision conversion with a residual error < 0.5° visual angle. The formula is as follows:

[0030]

[0031] Step 4: Eye movement event recognition. By calculating the angular velocity of eye movement in real time and angular acceleration , achieve precise classification by combining spatio-temporal features.

[0032] Specifically, calculate the angular velocity and angular acceleration based on the screen coordinate differences of consecutive frames:

[0033]

[0034] The classification conditions for eye movement events are as follows: Fixation: within 5 consecutive frames (≥ 20 ms) < 30° / s, and the coordinate standard deviation σ < 0.5°; Saccade: ≥ 30° / s and ≥ 8000° / s 2 , with a duration of 5 - 100 ms; Blink: the pixel area of the pupil region drops suddenly by ≥ 90% and lasts for 3 - 5 frames (20 - 50 ms); Smooth pursuit: speed [5° / s, 30° / s], and is in the same direction as the target movement direction (cosine similarity > 0.9).

[0035] Step 5: MR acquisition condition judgment. Based on the eye movement events identified in Step 4, judge whether the acquisition conditions for optic nerve magnetic resonance imaging are met, and send a feedback signal to the MR system main control device; As a judgment window, count the eye movement events and their time proportions within the window.

[0036] According to actual usage experience, select a judgment threshold. When the number of saccade and blink events within the judgment window is less than a certain value, or the time proportion is less than a certain percentage, it is considered that the acquisition conditions are met, and an instruction to agree to acquire is sent to the MR system main control device to enter the next step; When this threshold is not met, the acquisition condition is not met at this time, and a command is sent to the MR system main control device to return to step 1. The threshold for this condition judgment can be dynamically adjusted according to historical data (such as user fatigue level) or environmental factors (such as ambient light intensity); This embodiment also provides a feedback mechanism, which ensures the reliability of acquisition through hardware-level triggering and exception handling dual modules: (1) Hardware-level triggering: Generate TTL pulses or API instructions and transmit them to the MR system main control device in real time. Rely on FPGA timing optimization to ensure that the scan trigger synchronization error is less than 1 ms, avoiding phase mismatch between the imaging sequence and the eye movement state; (2) Exception handling: If the collection conditions are not met for three consecutive statistical cycles, the online calibration process (such as nine-point calibration recheck) will be automatically triggered and a user prompt (such as "Please keep your head fixed") will be pushed, and the scanning device will be frozen simultaneously to avoid invalid data collection until the system returns to a stable state.

[0037] This mechanism significantly improves the success rate and data validity of optic nerve scanning in complex environments through real-time closed-loop feedback and dynamic fault-tolerant strategies.

[0038] Step 6: MR sequence acquisition trigger. After receiving the feedback signal of consent acquisition in step 5, the MR system main control device automatically triggers the sequence acquisition and realizes accurate scanning through dynamic parameter adaptation; Specifically, the trigger modes of this embodiment include: (1) Immediate triggering: Immediately start the preset scanning sequence (such as T1-weighted imaging) after receiving the eye movement gating signal, with a response delay of <2 ms; (2) Delayed triggering: Based on the deep learning eye movement prediction model, the gaze stability period within the next 50-100 ms is predicted, and the scanning start time is dynamically optimized to match the optimal imaging window; illustratively, the eye movement prediction model in this embodiment includes but is not limited to deep learning networks such as LSTM and TCN, and this embodiment does not impose additional restrictions or elaboration on the network structure.

[0039] Optionally, the parameter dynamic adaptation steps of this embodiment are as follows: (1) Adjust scanning parameters according to the real-time eye movement stability index (such as coordinate standard deviation σ), including shortening the repetition time TR (adaptive to 200-1000 ms) and echo time TE (adaptive to 10-50 ms) to reduce the risk of motion artifacts; (2)Multi-modal sequence intelligent switching: preferentially call sequences sensitive to motion (such as diffusion-weighted imaging DWI), perform high b-value scanning (b = 3000 s / mm²) during the stable period, and switch to artifact-tolerant sequences (such as T2-FLAIR) during the unstable period to improve the processing efficiency of data. This logic realizes the collaborative optimization of scanning efficiency and imaging quality through predictive triggering and parameter closed-loop regulation.

[0040] Step 7: Eye movement data annotation and synchronization. Annotate the eye movement events obtained in Step 4 in the k-space data, and achieve accurate data fusion through multi-level time alignment and spatio-temporal mapping; The specific process is as follows: (1)Hardware-level time synchronization: The eye tracker and the MR acquisition device share a high-precision clock source (error < 1 ms) based on the PTP / NTP protocol to ensure global alignment of timestamps; Exemplarily, in this embodiment, the hardware-level time synchronization realizes the millisecond-level clock alignment between the eye tracker and the MRI device through the cooperation of PTP (Precision Time Protocol, IEEE 1588 standard) and NTP (Network Time Protocol, RFC 1305 standard). In specific implementation, the MRI master control device serves as the PTP grandmaster, sends a sequence of timing synchronization signals including synchronization messages and follow-up messages to the eye tracker through gigabit Ethernet, and uses the timestamp marks generated by the physical layer (PHY) hardware module to measure the network transmission delay in real time, dynamically correct the deviation between the local clock of the eye tracker and the MRI master clock, and ensure that the timestamp data of the eye movement events is globally aligned with the scan start signal triggered by the gradient controller in the MRI scan timing, with a time synchronization error of less than 1 millisecond. When the PTP synchronization link fails, the system automatically switches to the backup time protocol (NTP), obtains the reference time from the locally deployed high-precision time server (directly connected to the atomic clock or GPS timing module), and maintains the global time synchronization error within 5 milliseconds by dynamically screening the optimal intervals of multiple time sources and excluding network jitter or abnormal server interference. During the timestamp embedding process, the time difference (Δt) between the eye movement event label and the k-space acquisition time is verified in real time. If Δt < 1 ms, the label is valid; otherwise, a hardware reset and scan freeze are triggered.

[0041] (2)Event marker embedding: Insert eye movement event labels (such as saccade_start, blink_end) into the k-space data header file, and record the event type, start time, and duration in binary coding format to achieve millisecond-level event-scan timing association; (3)Space-time mapping construction: Through hash indexing, the eye movement coordinates and their corresponding eye movement states (fixation / saccade / blink) are dynamically associated with the k-space acquisition time tk to generate a multi-dimensional mapping table { } ↔ {fixation, saccade, blink}, which supports the subsequent motion artifact correction module to selectively remove the interfered k-space data according to the eye movement state, improving the signal-to-noise ratio of optic nerve imaging.

[0042] Step 8: k-space data correction and image reconstruction. Based on the eye movement data marking information in Step 7 and the original k-space data collected in Step 6, the k-space data is corrected, and image reconstruction is performed using an MR reconstructor, thereby reducing the impact of motion on image quality.

[0043] The specific process is as follows: (1)Intelligent removal of artifact data: First, identify the contaminated data. If the k-space acquisition time window overlaps with an eye movement event (such as a saccade), mark this data block as "contaminated data", and calculate the contamination confidence level C∈[0,1] based on the eye movement intensity (speed / acceleration). When C > 0.8, discard the contaminated data block, and generate a mask matrix through adjacent uncontaminated data to reduce the spread of artifacts; (2)Motion compensation enhancement: First, perform k-space data interpolation: Use adjacent uncontaminated data to fill the missing areas of the k-space (i.e., the "contaminated data" discarded in the previous step) through GRAPPA or SPIRiT; then perform motion field estimation: Predict tissue displacement based on the eye movement vector to correct the artifacts in the phase encoding direction.

[0044] (3)Use an MR reconstructor to perform image reconstruction on the k-space data corrected in the previous two steps.

[0045] As another alternative implementation, as Figure 3 shown, the system of this embodiment includes: a head coil, an eye tracker mirror, an MR, an eye tracker, an eye tracker main control computer, an MR system main control device, and an MR reconstruction device; Among them, the head coil is a radio frequency transmitting / receiving head coil. During the magnetic resonance signal acquisition process, it is responsible for transmitting radio frequency pulses to the head tissue to excite the generation of magnetic resonance signals, and accurately receiving the magnetic resonance signals generated by the response of the head tissue, providing basic data for subsequent steps such as image reconstruction; The eye tracker mirror is used to reflect the eye image into the field of view of the eye tracker camera, so that the high-speed camera of the eye tracker can accurately capture the eye image; The MR acquisition device is the general term for the hardware that completes the magnetic resonance signal acquisition; as Figure 4 ​As shown, the eye tracker uses the dark pupil technology, combines a near-infrared light source and a camera to achieve high-precision eye movement capture; the main control computer of the eye tracker is used to process the eye detection information recognized by the eye tracker and generate a feedback signal, and send the feedback signal to the main control device of the MR system; the main control device of the MR system is used to receive the feedback signal sent by the eye movement signal processing device, perform a feedback operation, and control the acquisition of the MR sequence; the MR reconstruction device is used to reconstruct the original MR signal collected in the k-space into an MR image.

[0046] Optionally, the head coil of this embodiment is set at a specified position on the magnetic resonance bedplate and can only be used after being recognized by the magnetic resonance system; the eye tracker mirror is placed on the upper part of the head coil, and the angle of the mirror is adjusted until the eye image is located at the center of the field of view of the eye tracker camera; the eye tracker is placed at the rear of the head coil and is lifted to a suitable height by a special rack, and the rack is located at the rear position inside the aperture; the main control device of the MR system, the MR reconstruction device, and the main control computer of the eye tracker are all located in the equipment room.

[0047] Exemplarily, the system of this embodiment is used for monitoring and control during MR optic nerve scanning. The sampling frequency of the eye tracker above 1000Hz can capture the movement of the eyeball more sensitively, enabling the present invention to efficiently shield the triggering of the sequence at inappropriate scanning moments. The eye tracker can provide richer eye movement information and more finely distinguish different eye movement states. Using this information, the present invention can make more accurate markings in the collected data for artifact correction during reconstruction. In addition, the present invention can suppress motion artifacts in a natural state, is applicable to patients with low cooperation such as children, and reduces the use of sedatives.

[0048] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0049] Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, there will be changes in the specific implementation manner and application scope according to the idea of the present invention. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for monitoring optic nerve scanning based on eye movement gating, characterized in that, Including: Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determine whether the acquisition conditions for optic nerve magnetic resonance imaging are met. If so, generate a feedback signal; Trigger the MR sequence acquisition according to the feedback signal; Annotate the eye movement events in the k-space data to obtain eye movement data marking information; Correct the k-space data according to the eye movement data marking information to obtain corrected data, and use an MR reconstructor to perform image reconstruction on the corrected data; Based on the eye movement events obtained through eye tracker image acquisition and eye movement recognition, determine whether the acquisition conditions for optic nerve magnetic resonance imaging are met. If so, generate a feedback signal, including: Taking as a judgment window, counting the eye movement events and time proportion within the judgment window; is the judgment window threshold parameter; Judge whether the number of saccade and blink eye movement events within the judgment window is less than a preset threshold, or whether the time ratio is less than a preset ratio. If so, it is considered that the acquisition conditions are met, and a feedback signal of consent to acquisition is sent to the MR system main control device; If not, it is considered that the acquisition conditions are not met, and an instruction is sent to the MR system main control device to re-acquire the eye movement events.

2. The method for optic nerve scan monitoring based on eye movement gating according to claim 1, characterized in that The preset threshold and the preset ratio are dynamically adjusted according to historical data or environmental factors.

3. The method for monitoring optic nerve scanning based on eye movement gating according to claim 1, wherein The transmission of the feedback signal relies on FPGA timing optimization to ensure that the scanning trigger synchronization error < 1ms; if the acquisition conditions are not met for 3 consecutive statistical periods, the online calibration process is automatically triggered and a user prompt is pushed, and the scanning device is frozen synchronously to avoid invalid data acquisition until the acquisition conditions are met.

4. The method for optic nerve scan monitoring based on eye movement gating according to claim 1, characterized in that The trigger modes for triggering the MR sequence acquisition include: immediate trigger and delayed trigger; the immediate trigger starts a preset scanning sequence after receiving the feedback signal, and the response delay < 2ms; the delayed trigger is based on a deep learning eye movement prediction model to predict the fixation stable period within the next 50 - 100ms, and dynamically optimize the scanning start time to match the best imaging window.

5. The method for optic nerve scan monitoring based on eye movement gating according to claim 1, wherein, During the MR sequence acquisition process, it also includes: Adjust the scanning parameters according to the real-time eye movement stability index; the scanning parameters include repetition time and echo time; Prioritize the call of motion-sensitive sequences, perform high b-value scanning during the stable period, and switch to artifact-tolerant sequences during the unstable period.

6. The method for optic nerve scan monitoring based on eye movement gating according to claim 1, characterized in that, Annotate the eye movement events in the k-space data to obtain eye movement data marking information, including: Based on the PTP / NTP protocol, enable the eye tracker and the MR acquisition device to share a high-precision clock source to ensure global alignment of timestamps; Insert an eye movement event label into the header file of the k-space data, and record the event type, start time, and duration in binary coding format to achieve millisecond-level event-scanning timing association to obtain the eye movement data marking information; Dynamically associate the eye movement coordinates and the corresponding eye movement states with the acquisition moment of the k-space data through hash indexing to generate a multi-dimensional mapping table.

7. The method for monitoring optic nerve scanning based on eye movement gating according to claim 1, characterized in that, Correct the k-space data according to the eye movement data marking information to obtain corrected data, and use an MR reconstructor to perform image reconstruction on the corrected data, including: If the acquisition time window of the k-space data overlaps with the eye movement events, mark the corresponding overlapping data blocks as contaminated data; For each data block marked as contaminated data, calculate the contamination confidence based on the eye movement intensity. When the contamination confidence is greater than the preset contamination threshold, discard the data block marked as contaminated data; Utilize adjacent uncontaminated data blocks to generate a mask matrix to reduce the spreading effect of artifacts; For the discarded contaminated data blocks, use adjacent uncontaminated data to fill the missing regions in the k-space data; Predict the displacement of the tissue according to the eye movement vector, correct the phase encoding direction artifacts to ensure the accuracy and clarity of the image, and obtain the corrected data; Use an MR reconstructor to reconstruct the corrected data into a final MR image.

8. The method for optic nerve scan monitoring based on eye movement gating according to claim 7, wherein The preset contamination threshold is 0.

8.

9. A system for optic nerve scan monitoring based on eye movement gating, characterized in that, A method for implementing the eye movement gating-based optic nerve scan monitoring as described in any one of claims 1 to 8, the eye movement gating-based optic nerve scan monitoring system comprising: a head coil, an eye tracker mirror, an MR acquisition device, and an eye tracker, an eye tracker main control computer, an MR system main control device, and an MR reconstruction device connected in sequence; The head coil is arranged at a specified position on the magnetic resonance bedplate; the eye tracker mirror is placed on the upper part of the head coil; the eye tracker is placed at the rear of the head coil; the head coil is used to emit radio frequency pulses to the head tissue during the magnetic resonance signal acquisition process to excite the generation of magnetic resonance signals and receive the magnetic resonance signals generated by the response of the head tissue; the eye tracker mirror is used to reflect the eye image into the field of view of the camera of the eye tracker; the MR acquisition device is used to complete the acquisition of magnetic resonance signals; the eye tracker is used to capture the eye movement using the dark pupil technology in combination with a near-infrared light source and a camera to obtain eye detection information; the eye tracker main control computer is used to process the eye detection information recognized by the eye tracker to obtain eye movement events, generate a feedback signal, and send the feedback signal to the MR system main control device; the MR system main control device is used to receive the feedback signal sent by the eye movement signal processing device, perform a feedback operation, and control the acquisition of the MR sequence; the MR reconstruction device is used to reconstruct the original MR signals acquired in the k-space data into an MR image.

Images