Fetal biometric angle ultrasound image measurement method and system based on a watch coordinate system

Abstract

This invention, entitled "A Method and System for Fetal Biological Angle Ultrasound Imaging Measurement Based on a Clock Coordinate System," belongs to the field of ultrasound imaging medical technology. The technical problem it aims to solve is the cumbersome operation, lack of intuitive results, and significant influence from changes in body position in existing fetal biological angle measurements. The key technical points are: acquiring fetal ultrasound cross-sectional images; determining the anatomical reference point and the target point to be measured; constructing a virtual clock coordinate system using the anatomical reference point as the origin and the direction pointing from the origin to a preset anatomical reference structure as the reference scale direction, and superimposing it on the ultrasound cross-sectional image; connecting the origin and the target point to be measured to form a target vector; calculating the physical angle of the target vector relative to the reference scale direction; and using an angle-scale mapping algorithm to convert the physical angle into a clock scale value and display it on the ultrasound cross-sectional image.

Description

Technical Field

[0001] This invention relates to the field of ultrasound imaging medical technology, specifically to a method and system for measuring fetal biological angle ultrasound images based on a clock coordinate system. Background Technology

[0002] Currently, the mainstream method for measuring fetal bio-angles in clinical practice is still manual operation. Doctors need to locate multiple anatomical landmarks on the ultrasound image, manually draw two or more intersecting straight lines, and calculate their physical angles. Because the fetus's position in utero is random and the ultrasound view may be skewed, the manual calibration of the baseline inevitably introduces human error, resulting in poor repeatability of measurement results between different operators. Especially during real-time dynamic scanning, the operator needs to simultaneously control the probe and identify landmarks, placing a significant burden on the operator.

[0003] The way measurement results are expressed also limits the efficiency of clinical applications. Existing methods generally use physical angle values ​​as outputs, such as the normal reference value for the cardiac axis being 45°±20°. However, in actual diagnostic scenarios, ultrasound doctors are more accustomed to using clockwork orientation to describe the spatial relationship of fetal anatomical structures, such as the apex of the heart pointing towards the 4 o'clock position. Immediately matching abstract angle values ​​with the actual position of the fetus places a significant cognitive burden on inexperienced doctors, easily affecting the speed and accuracy of interpretation.

[0004] Furthermore, while existing ultrasound equipment is equipped with auxiliary tools such as protractors, these tools are statically superimposed on the screen, and their reference direction is fixed to the screen coordinate system rather than the fetus's own anatomical coordinate system. When the fetus rotates in utero or the doctor adjusts the probe angle, the correspondence between the reference direction on the screen and the fetus's actual longitudinal axis immediately becomes invalid. Doctors must re-establish the reference and recalibrate the baseline, significantly increasing the workload and raising the risk of missed diagnoses.

[0005] In recent years, automated angle measurement technologies based on deep learning have gradually emerged, alleviating the burden of manual operation to some extent. However, most existing technologies follow the traditional logic of identifying marker points, fitting axes, and calculating angles, resulting in physical angle values ​​as outputs, failing to fundamentally improve the intuitiveness of the results. Furthermore, these solutions are mostly designed for single measurement scenarios, lacking versatility across sections and multiple locations, and their robustness is limited when section quality is poor. More importantly, existing automated solutions have not established a dynamic reference coordinate system bound to the fetus's own anatomical structure, thus failing to maintain the stability of the measurement benchmark under conditions of continuous changes in fetal position.

[0006] Chinese patent application CN116309528A discloses a method, device, and computer equipment for processing fetal cardiac ultrasound images. It discloses a technical solution that automatically extracts the atrial septum, ventricular septum, spine, and thoracic cavity contours in a four-chamber view using a deep learning segmentation model. The solution obtains the axis pointing towards the apex of the heart and the axis connecting the spine and sternum through skeleton extraction and straight line fitting, and then automatically calculates the angle between the two axes to obtain the cardiac axis. This solution has the technical effect of eliminating subjective errors caused by manual tracing and improving the accuracy of cardiac axis measurement. However, it still has problems such as the measurement results being output only from a physical angle and lacking clinical intuitiveness, the lack of a dynamic coordinate reference system that is updated in real time with changes in fetal position, and the limitation to the single measurement scenario of cardiac axis, which cannot be transferred to other biological angles such as facial angles for evaluation.

[0007] European patent application document EP3510933A1 discloses an ultrasound diagnostic device and its control method. It discloses a technical solution that automatically identifies the intersection of the fetal spine center and the heart's cross point and connects them to generate a body axis. At the same time, it determines the heart axis based on the position of the ventricular septum and displays the angle between the two axes and an angle indicator with color grading on the ultrasound image. This achieves the technical effect of automatically detecting the heart axis angle and assisting doctors in intuitively judging whether the heart is abnormal through a visual indicator. However, it still has problems such as the angle indicator still being based on physical angles and not being transformed into a clinically used orientation description system; the displayed axis being a static annotation based on the current frame and lacking the ability to dynamically compensate for changes in fetal position; and the reference point being fixed at the line connecting the heart's cross point and the spine and unable to be flexibly switched to adapt to various biological angle measurement scenarios such as facial angles. Summary of the Invention

[0008] The purpose of this invention is to provide: A method and system for measuring fetal bio-angle ultrasound images based on a clock coordinate system, and related technologies, to solve technical problems such as cumbersome operation, unintuitive results, and significant influence from changes in body position in existing fetal bio-angle measurements, or combinations thereof.

[0009] Terminology Explanation: Unless otherwise defined, all technical terms in this document have the same meanings as commonly understood by one of ordinary skill in the art to which the subject matter of the claims pertains. Unless otherwise stated, all patents, patent inventions, and publications cited in this document are incorporated herein by reference in their entirety. If multiple definitions exist for terms in this document, the definitions in this chapter shall prevail.

[0010] It should be understood that the above brief description and the following detailed description are exemplary and for illustrative purposes only, and do not limit the subject matter of the invention in any way. In this invention, the singular is used in conjunction with the plural unless otherwise specifically stated. It should also be noted that, unless otherwise stated, the use of “or” or “or” means “and / or”. Furthermore, the use of the term “comprising” and other forms such as “including,” “containing,” and “contains” are not limiting.

[0011] Unless specifically defined herein, the use of various commercially available products herein employs standard techniques, or is carried out in accordance with methods known in the art or the description of this invention. The techniques and methods described herein can generally be implemented according to conventional methods well known in the art, based on the descriptions in the various summary and more specific documents cited and discussed in this specification.

[0012] The term "clock coordinate system" used in this article refers to a method of describing orientation using a clock face as a reference. It divides the 360° angular space in a plane into 12 intervals, each interval corresponding to 30°, with the 12 o'clock position as the starting reference direction, and the positions sequentially marked clockwise from 1 o'clock to 12 o'clock. In this invention, the clock coordinate system is constructed with the fetus's own anatomical reference point as the center and a specific anatomical reference direction as the 12 o'clock position, used to intuitively map physical angle values ​​to clock-like orientation readings familiar to clinicians.

[0013] In a first aspect, the present invention provides a method for measuring fetal biological angle ultrasound images based on a clock coordinate system, comprising the following steps: S1, acquire fetal ultrasound cross-sectional images and determine anatomical reference points and target points to be measured; S2, using the anatomical reference point as the origin and the direction from the origin to the preset anatomical reference structure as the reference scale direction, construct a virtual clock coordinate system and superimpose the virtual clock coordinate system onto the ultrasound section image; S3, connect the origin and the target point to be measured to form a target vector, and calculate the physical angle θ between the target vector and the reference scale direction; S4. The physical angle θ is converted into a corresponding clock scale value using an angle-scale mapping algorithm, and the clock scale value is displayed on the ultrasonic cross-sectional image.

[0014] Furthermore, the angle-scale mapping algorithm in S4 follows the formula: Clock Pos = (θ ÷ 30°) modulo operation 12; where θ is the clockwise physical angle, with a value range of 0°-360°, and Clock Pos This is the output clock scale value; when Clock... Pos When the calculation result is 0, assign it the value 12.

[0015] Furthermore, S4 also includes precision compensation processing: based on Clock Pos The decimal part of the calculation result is used to further map the clock scale value into a clock orientation expression that includes minute precision.

[0016] Furthermore, it also includes a dynamic compensation step: updating the coordinate system rotation angle in real time based on the displacement vector and rotation angle of the anatomical reference point.

[0017] Furthermore, the image recognition algorithm in S1 is an image recognition model; the reference scale direction is defined as the 12 o'clock direction of the virtual clock coordinate system.

[0018] Furthermore, after S1 and before S2, an image quality assessment step is included: using a quality assessment model to score the standardization of the ultrasound cross-sectional image, and only when the score is higher than a preset threshold, S2 and subsequent steps are executed.

[0019] Furthermore, it also includes a clinical warning step: comparing the clock scale value with the normal scale range in a preset clinical standard database, and outputting a warning signal on the display interface of the ultrasound section image when the clock scale value deviates from the normal scale range.

[0020] Furthermore, the determination of the target point to be measured in S1 includes the following steps: A rotating ray is emitted from the origin to determine the angular range of the heart, and the extreme distance point is selected as the target point to be measured; the angle range is 0°-360°.

[0021] Furthermore, in S4, the physical angles between the first boundary direction and the second boundary direction relative to the reference scale direction are calculated respectively, and converted into clock scale values ​​by the angle-scale mapping algorithm respectively, forming a clock scale interval of the angular distribution range of the fetal target anatomical structure and displayed on the ultrasound section image.

[0022] Furthermore, step S1 further includes the step of determining at least two directional anchor points, wherein the directional anchor points include at least two of the following types: The geometric center point determined based on the overall region of the target anatomical structure, the forward extreme point of the target anatomical structure in the ultrasound section image, the boundary point corresponding to the maximum or minimum angle within the angular distribution range of the target anatomical structure, the location point determined based on the functional connection region inside the target anatomical structure, the direction corresponding location point determined based on the blood flow or motion information of the target anatomical structure, and the feature point determined based on the regional morphological features after segmenting the target anatomical structure region. In step S3, the origin is connected to each of the direction anchor points to form multiple spatial direction lines, and the physical angle between each spatial direction line and the reference scale direction is calculated. In step S4, the physical angle between each spatial direction line is converted into the corresponding clock scale value through the angle-scale mapping algorithm, and the spatial configuration features of the target anatomical structure are constructed based on the angular relationship between the multiple spatial direction lines.

[0023] Furthermore, the spatial configuration features include at least one of the following: an angle vector composed of the angle values ​​of each spatial direction line relative to the reference scale direction, the relative angle values ​​between each spatial direction line, and an angle combination pattern formed by the clock scale values ​​of multiple spatial direction lines. It also includes comparing the spatial conformation features with a pre-established reference conformation interval or reference pattern. When the spatial conformation features exceed the reference conformation interval or the matching degree with the reference pattern is lower than a preset threshold, the target anatomical structure is determined to be an abnormal conformation and an early warning signal is output.

[0024] Secondly, the present invention provides: a fetal bio-angle ultrasound imaging measurement system based on a clock coordinate system, using the method described in the first aspect, comprising: The system includes: an image acquisition module for acquiring real-time fetal ultrasound cross-sectional images; a feature recognition module for automatically locating anatomical reference points and target points in the ultrasound cross-sectional images using image recognition algorithms; a coordinate system construction module for generating a virtual clock coordinate system with the anatomical reference points as the origin and the direction pointing from the origin to a preset anatomical reference structure as the reference scale direction, and driving the virtual clock coordinate system to rotate synchronously according to the real-time position changes of the anatomical reference points; a scale calculation module for calculating the physical angle θ between the target point and the reference scale direction, and converting the physical angle θ into clock scale values ​​using an angle-scale mapping algorithm; and a display module for overlaying the virtual clock coordinate system and the clock scale values ​​on the ultrasound diagnostic interface in real time.

[0025] Furthermore, the angle-scale mapping algorithm in the scale calculation module follows the formula: Clock Pos =(θ / 30°) mod 12; where θ is the physical included angle, Clock Pos This is the output clock scale value.

[0026] Furthermore, it also includes an early warning module, which is used to compare the clock scale value with the normal scale range in a preset clinical standard database, and output an early warning signal through the display module when the clock scale value deviates from the normal scale range.

[0027] The present invention has at least the following beneficial effects: 1. This invention is the first to dynamically bind a clock coordinate system to the fetus's own anatomical reference, so as to keep the measurement results constant under the change of body position. By converting physical angles into clock position readings, the measurement results directly correspond to the position description habits used by ultrasound doctors in daily life. Doctors do not need to mentally calculate and match abstract angle values ​​with fetal position in real time scanning, which significantly reduces the cognitive load in the interpretation process and improves screening efficiency.

[0028] 2. The virtual clock coordinate system constructed in this invention uses the fetus's own anatomical structure as the anchor point. When the fetus rotates in the uterus or the probe angle changes, the system drives the clock face to rotate synchronously in real time to compensate, so that the clock scale output always remains constant relative to the fetus's longitudinal axis. There is no need to manually reset the reference, which solves the problem of reference failure of existing static reference tools under the condition of fetal position change.

[0029] 3. This invention automatically locates anatomical reference points and target points through image recognition algorithms, replacing traditional manual calibration and stringing operations. It eliminates subjective errors caused by differences in manual point selection between different operators, reduces the need for doctors' experience and qualifications, and makes measurement simpler and results more accurate. Attached Figure Description

[0030] Figure 1 The flowchart illustrates a method for measuring fetal biological angle ultrasound images based on a clock coordinate system, as provided in this invention.

[0031] Figure 2 This is a schematic diagram of a fetal bio-angle ultrasound imaging measurement system based on a clock coordinate system provided by the present invention.

[0032] Figure 3 This is a schematic diagram of a model constructed based on multi-directional anchor points and multiple spatial direction lines based on the rotation of the spinal pole in one embodiment of the present invention.

[0033] Figure 4 This is a schematic diagram of the normal spatial configuration of a breech fetus in one embodiment of the present invention.

[0034] Figure 5 This is a schematic diagram of the abnormal spatial configuration of a breech fetus in one embodiment of the present invention.

[0035] Figure 6 This is a schematic diagram of the normal spatial configuration of a head-down fetus in one embodiment of the present invention.

[0036] Figure 7 This is a schematic diagram of an abnormal spatial configuration of a head-positioned fetus in one embodiment of the present invention.

[0037] Figure 8This is a schematic diagram of fetal facial contour angle measurement in one embodiment of the present invention. Detailed Implementation

[0038] The following non-limiting embodiments are intended to enable those skilled in the art to gain a more comprehensive understanding of the present invention, but do not limit the invention in any way. The following content is merely an exemplary description of the scope of protection claimed by the present invention, and those skilled in the art can make various changes and modifications to the present invention based on the disclosed content, and such changes should also fall within the scope of protection claimed by the present invention.

[0039] The present invention will be further described below by way of specific embodiments. Unless otherwise specified, all instruments, devices, equipment, reagents, products, etc., used in the embodiments of the present invention are obtained through conventional commercial means.

[0040] Unless otherwise stated, conventional methods within the scope of the art shall be used.

[0041] Example 1 like Figure 1 As shown, this invention provides a method for measuring fetal biological angle ultrasound images based on a clock coordinate system, comprising the following steps: S1, acquire fetal ultrasound cross-sectional images and determine anatomical reference points and target points to be measured; S2, with the anatomical reference point as the origin and the direction from the origin to the preset anatomical reference structure as the reference scale direction, a virtual clock coordinate system is constructed, and the virtual clock coordinate system is superimposed on the ultrasound section image; S3, connect the origin and the target point to be measured to form the target vector, and calculate the physical angle θ between the target vector and the reference scale direction; S4 uses an angle-scale mapping algorithm to convert the physical angle θ into the corresponding clock scale value, and displays the clock scale value on the ultrasonic cross-sectional image.

[0042] Specifically, in step S1, the system acquires standard ultrasound cross-sectional images of the fetus, including but not limited to the four-chamber view and the midsagittal view. The image recognition algorithm automatically locates key anatomical structures within the cross-section, identifying the anatomical reference point (the origin of the coordinate system) and the target points to be measured. For example, in cardiac axis measurement, the anatomical reference point is the center of the fetal spine cross-section, and the target point is the apex of the heart. Similarly, in facial angle measurement, the anatomical reference point can be a feature point on the brainstem line or hard palate line, and the target point is a facial feature point such as the tip of the mandible. By automatically identifying and replacing manual calibration, subjective errors between different operators are eliminated.

[0043] In step S2, the system uses the identified anatomical reference point as the central origin of the virtual clock, and defines the direction from the origin to the preset anatomical reference structure as the reference scale direction of the clock. Taking the heart axis measurement scenario as an example, the reference scale direction is the direction of the line connecting the center of the spine to the sternum, which is defined as the 12 o'clock position of the virtual clock. The system overlays a virtual clock disk layer on the ultrasound image, with the center of the disk coinciding with the anatomical reference point, and the 12 o'clock position aligned with the reference scale direction. Because the coordinate system is anchored to the fetus's own anatomical structure rather than a fixed screen orientation, the measurement results under different positions are comparable.

[0044] In step S3, the system connects the origin and the target point to form a target vector. The clockwise physical angle θ between the reference scale direction and the target vector is calculated in the image coordinate system, with a value ranging from 0° to 360°. Taking the spinal axis measurement as an example, the origin is the center of the spine's cross-section, the target point is the apex of the heart, and the target vector is the direction of the line connecting the center of the spine to the apex of the heart.

[0045] In step S4, the system converts the physical angle θ into a clock scale value using an angle-scale mapping algorithm and simultaneously displays the physical angle value and the corresponding clock orientation reading on the ultrasound image. For example, when the physical angle θ is 135°, the corresponding clock scale value is 4.5, which is displayed as 4:30. Converting abstract angle values ​​into clock orientation expressions familiar to doctors reduces the cognitive load during real-time scanning.

[0046] In one specific embodiment of this example, the angle-scale mapping algorithm in step S4 follows the formula: Clock Pos = (θ / 30°) mod 12. Where θ is the physical angle, ranging from 0° to 360°, and Clock... Pos This is the output clock scale value. Since the clock face is divided into 360° intervals and 12 hour markers, with each 30° physical angle corresponding to one hour, θ / 30° converts the angle to hours. The modulo 12 operation ensures the result cycles within the clock's time period. When Clock... Pos When the calculation result is 0, it is assigned the value 12, corresponding to the 12 o'clock position. For example, when θ is 0°, Clock... Pos The calculation result is 0, so it is assigned a value of 12, indicating that the target point is located at the 12 o'clock position; when θ is 90°, Clock... Pos The value of 3 indicates that the target point is located at the 3 o'clock position.

[0047] In one specific embodiment of this example, step S4 further includes precision compensation processing: according to Clock PosThe decimal part of the calculation result further maps the clock scale values ​​into a clock orientation expression that includes minute precision. Specifically, Clock Pos The integer part corresponds to the hour, and the fractional part multiplied by 60 corresponds to the minute. For example, Clock Pos When the value is 4.5, the integer part is 4, and the fractional part 0.5 multiplied by 60 equals 30, which is mapped to the 4 o'clock and 30-minute direction. Precision compensation processing ensures that the measurement results maintain intuitiveness while also possessing fine-grained angular discrimination capabilities.

[0048] In one specific embodiment of this method, the method further includes a dynamic compensation step: when displacement or rotation of the anatomical reference point is detected in the ultrasound cross-sectional image, the positional change of the anatomical reference point and the phase change of the reference scale direction are calculated in real time. Based on this, the virtual clock coordinate system is driven to rotate synchronously, ensuring that the clock scale value remains constant relative to the fetus's longitudinal axis. Specifically, when the fetus rotates in utero or the probe angle changes, the position of the anatomical reference point in the image changes accordingly. The system re-identifies the reference point position in real time and calculates the phase shift of the reference scale direction, driving the virtual clock face to perform corresponding angle rotation compensation to ensure that the 12 o'clock position always coincides with the fetus's longitudinal axis. Therefore, regardless of the fetus's movement, the clock scale output always reflects the true orientation of the target point relative to the fetus, avoiding reference failure and recalibration problems caused by changes in body position.

[0049] In one specific embodiment of this example, the image recognition algorithm in step S1 is a deep learning-based image recognition model, and the reference scale direction is defined as the 12 o'clock direction in the virtual clock coordinate system. The deep learning model, trained on a large number of labeled ultrasound images, extracts and locates anatomical landmarks from the cross-sectional images, exhibiting better robustness in non-standard cross-sectional conditions compared to traditional image processing methods. Fixing the reference scale direction to the 12 o'clock direction ensures that clock readings in different measurement scenarios follow a unified starting phase convention, facilitating clinical comparison and standardized evaluation.

[0050] In one specific embodiment of this example, an image quality assessment step is included after step S1 and before step S2: the standardization of the ultrasound cross-sectional image is scored using a quality assessment model, and step S2 and subsequent steps are executed only when the score is higher than a preset threshold. Non-standard cross-sections may lead to deviations in the identification of anatomical reference points, thus affecting the accuracy of subsequent angle calculations. By adding a quality screening step before coordinate system construction, it is ensured that the images entering the measurement process meet clinical requirements, thereby improving the reliability of the measurement results.

[0051] In one specific embodiment of this method, the method further includes a clinical warning step: comparing the clock scale value with the normal scale range in a preset clinical standard database; when the clock scale value deviates from the normal scale range, a warning signal is output on the display interface of the ultrasound cross-sectional image. Taking atrial axis measurement as an example, the clock scale range corresponding to a normal atrial axis is 4 o'clock to 5 o'clock. When the clock scale value calculated by the system falls outside the above range, the interface automatically issues a warning using color indicators or other methods, prompting doctors to pay attention and further confirm, which helps reduce the risk of missed diagnosis.

[0052] Example 2 like Figure 2 As shown, the present invention also provides a fetal biological angle ultrasound imaging measurement system based on a clock coordinate system, which, using the above method, includes an image acquisition module, a feature recognition module, a coordinate system construction module, a scale calculation module, and a display module.

[0053] The image acquisition module acquires real-time fetal ultrasound cross-sectional images. The feature recognition module automatically locates anatomical reference points and target points in the ultrasound cross-sectional images using image recognition algorithms. The coordinate system construction module generates a virtual clock coordinate system with the anatomical reference point as the origin and the direction pointing from the origin to a preset anatomical reference structure as the reference scale direction, and drives the virtual clock coordinate system to rotate synchronously according to the real-time position changes of the anatomical reference point. The scale calculation module calculates the physical angle θ between the target point and the reference scale direction, and converts the physical angle θ into a clock scale value using an angle-scale mapping algorithm. The display module overlays the virtual clock coordinate system and clock scale values ​​on the ultrasound diagnostic interface in real time.

[0054] Specifically, the image acquisition module connects to the ultrasound probe or receives existing ultrasound image data, acquiring real-time or stored fetal ultrasound cross-sectional images and transmitting them to the feature recognition module. Upon receiving the images, the feature recognition module automatically locates the coordinates of the anatomical reference point and the target point, transmitting the location results to the coordinate system construction module and the scale calculation module. The coordinate system construction module generates a virtual clock face based on the position of the anatomical reference point and adjusts the clock face's rotation angle in real-time as the fetal position changes, maintaining consistency between the reference direction and the fetal longitudinal axis. The scale calculation module performs angle-scale mapping calculations based on the angle between the target vector and the reference direction, outputting the clock scale value. The display module overlays the virtual clock coordinate system layer and the clock scale reading onto the ultrasound diagnostic interface for real-time reference by the physician. These modules work together to achieve a complete automated measurement process from image acquisition to result display.

[0055] In one specific embodiment of this example, the angle-scale mapping algorithm in the scale calculation module follows the following formula: Clock Pos= (θ / 30°) mod 12, where θ is the physical angle, and Clock Pos This is the output clock scale value. The scale calculation module receives the target point coordinates output by the feature recognition module and the reference direction information output by the coordinate system construction module. After calculating the physical angle, it executes the above mapping formula and transmits the calculation result to the display module for presentation.

[0056] In one specific embodiment of this example, the system further includes an early warning module. This module compares the clock scale value with a preset clinical standard database of normal scale intervals and outputs an early warning signal via a display module when the clock scale value deviates from the normal range. The early warning module receives the output from the scale calculation module, matches it with the built-in clinical standard database, and triggers an early warning signal when the scale value exceeds the normal range. This warning is then displayed to the doctor on the interface using color indicators or similar methods.

[0057] Example 3 The following description uses fetal heart axis measurement as a complete example, taking a doctor's ultrasound examination of the fetus in early pregnancy as an example.

[0058] The image acquisition module acquires real-time ultrasound images of the fetal four-chamber view. The feature recognition module uses a deep learning model to automatically identify the center of the spinal cross-section as the anatomical reference point and the apex of the heart as the target point. The coordinate system construction module uses the center of the spinal cross-section as the origin, defines the direction from the spine to the sternum as the 12 o'clock direction, and overlays a virtual clock disk on the ultrasound image. The scale calculation module calculates the clockwise physical angle θ between the target vector pointing from the origin to the apex of the heart and the 12 o'clock direction. Assuming θ is 135°, it is calculated using the formula Clock. Pos = (135 / 30) mod12 = 4.5, mapped to the 4 o'clock and 30-minute direction. The display module synchronously displays the physical angle 135° and the clock orientation 4 o'clock and 30-minute direction on the ultrasound image.

[0059] Since the clock scale corresponding to a normal cardiac axis is between 4 and 5 o'clock, and 4.5 falls within the normal range, the warning module does not issue a warning. If the fetus subsequently rotates within the uterus, causing the spine position to shift in the image, the coordinate system construction module detects the shift in real time and drives the virtual clock face to rotate synchronously, ensuring that the 12 o'clock position is always aligned with the line connecting the spine and sternum. The clock scale value recalculated by the scale calculation module still accurately reflects the position of the apex of the heart relative to the fetus itself.

[0060] Example 4 See Figures 4-7The following uses fetal heart spatial configuration assessment as a typical application scenario of the clock coordinate system measurement method of this invention, and combines polar coordinate scanning and multi-directional anchor point technology to provide a detailed description of the complete implementation process.

[0061] In step S1, the image acquisition module acquires a cross-sectional ultrasound image of the fetal chest cavity using an ultrasound examination device. The acquired ultrasound image includes at least the fetal spinal structure and the overall structure of the fetal heart, preferably a cross-sectional image near the four-chamber view, but it is not required to strictly follow the standard four-chamber view; it is sufficient to clearly show the outline of the spine and heart. If the image quality is poor, grayscale enhancement, noise reduction, or edge enhancement can be performed on the image to improve the contrast between the fetal heart structure and surrounding tissues. The feature recognition module automatically identifies the position of the fetal spine in the ultrasound image using an image recognition algorithm. By calculating the geometric center of the spinal echo region, the center point of the spine is determined as the anatomical reference point. As the most stable anatomical structure in the fetal body with the most significant echo characteristics, the position of the spine does not change significantly with abnormalities in the internal structure of the heart, making it suitable as a unified spatial reference benchmark.

[0062] In step S2, the coordinate system construction module uses the center point of the spine as the origin of the virtual clock coordinate system and the direction from the origin to the preset anatomical reference structure (such as the direction of the sternum) as the reference scale direction, defines it as the 12 o'clock direction of the virtual clock, and superimposes a virtual clock disk layer on the ultrasound image, with the center of the disk coinciding with the center point of the spine.

[0063] After step S2 is completed and before step S3 is executed, the system performs a polar coordinate rotation scan centered on the origin to determine the angular distribution range and main orientation of the fetal heart within the thoracic cavity. Specifically, the system emits a rotating ray from the origin and rotates gradually within a predetermined angular range with a fixed angular step size. The angular step size can be set to 1°, 2°, or other suitable angular resolution according to accuracy requirements. During the ray rotation, the system monitors the intersection between each rotating ray and the fetal heart structure in real time. When the rotating ray first intersects the outer contour of the fetal heart, the system records the corresponding angular position as the first boundary direction of the angular distribution of the fetal heart. As the rotating ray continues to rotate, when the ray no longer intersects the outer contour of the fetal heart, the system records the corresponding angular position as the second boundary direction. The fan-shaped area jointly defined by the first and second boundary directions represents the angular distribution range of the fetal heart within the thoracic cavity.

[0064] After determining the angular distribution range, the system further analyzes the intersection of each rotating ray with the cardiac structure within this range. Specifically, the system calculates the distance from the intersection point of each ray with the cardiac structure to the origin, and selects the angular position corresponding to the rotating ray whose distance reaches its extreme value as the predominant orientation of the fetal heart. In most cases, the predominant orientation coincides with the direction pointed to by the apex of the heart, but this method does not require explicit identification of the apex of the heart; it can be determined solely through the extreme value analysis of the ray intersection distance. The system uses the predominant orientation position as the target point to be measured.

[0065] In step S3, the system connects the origin and the target point to form a target vector, and calculates the clockwise physical angle θ between the target vector and the reference scale direction (12 o'clock direction), with a value ranging from 0° to 360°. Simultaneously, the system calculates the corresponding physical angles between the first boundary direction and the second boundary direction relative to the reference scale direction.

[0066] In step S4, the scale calculation module calculates the physical angle θ corresponding to the main orientation using the formula Clock. Pos = (θ / 30°) mod 12 is converted to clock scale values. The physical angles between the first and second boundary directions are also mapped to clock scale values ​​according to the same formula, thus forming a clock scale interval for the angular distribution range of the heart. The display module synchronously displays the virtual clock coordinate system, the clock scale value of the heart's principal orientation, and the angular distribution interval defined by the two boundary scale values ​​on the ultrasound image. For example, if the physical angle corresponding to the principal orientation is 150°, then Clock... Pos = (150 / 30) mod 12 = 5, which is displayed as "5 o'clock direction"; if the first boundary direction corresponds to 120° and the second boundary direction corresponds to 210°, then the angular distribution range is displayed as "4 o'clock to 7 o'clock".

[0067] In one specific embodiment of this example, based on determining the angular distribution range of the heart, the system further determines multiple directional anchor points and constructs multiple spatial direction lines to form a more complete spatial conformation description. The methods for determining the directional anchor points include the following: The first method calculates the geometric center point based on the overall fetal heart region and uses it as the directional anchor point. The second method identifies the anterior extreme point of the fetal heart closest to the chest wall in the transverse section of the thoracic cavity and uses it as the directional anchor point. The third method selects the heart boundary point corresponding to the maximum or minimum angle within the angular distribution range and uses it as the directional anchor point. The fourth method determines the location point as the directional anchor point based on the functional connection regions of the fetal heart's internal structures (such as the junction of the four chambers). The fifth method determines the location point corresponding to the main blood flow direction based on cardiac blood flow information acquired by color Doppler ultrasound. The sixth method, after image segmentation of the heart region, determines feature points as directional anchor points based on the morphological features of the segmented region (such as the endpoint of the long axis and the direction of centroid offset).

[0068] In practical applications, at least two different types of anchor points can be selected from the above-mentioned directional anchor points to participate in the construction of spatial configuration.

[0069] The system connects the origin to anchor points in each direction, forming multiple spatial direction lines. These direction lines are analogous to the hour hands on a clock face. The scale calculation module calculates the physical angle between each direction line and the reference scale direction, and converts it into the corresponding clock scale value using a mapping formula. Thus, each anchor point obtains a clear scale position in the clock coordinate system.

[0070] The system constructs spatial conformation features of the fetal heart based on the angular relationships between multiple spatial direction lines. Specific forms of these features include: angle vectors composed of the angle values ​​of each direction line relative to a reference scale direction; relative angle values ​​between each direction line; the distribution interval defined by the start and end angles of the angular distribution range; and angle combination patterns formed by the clock scale values ​​of multiple direction lines. These angle combination patterns can be matched with a pre-established reference pattern of a normal fetal heart to determine whether the spatial conformation is normal. These feature parameters can be used individually or combined to form a feature parameter set, which can then be used as input to subsequent classification or recognition models.

[0071] The early warning module compares the calculated spatial conformation features with a pre-established reference conformation interval. The reference conformation interval is established based on statistical analysis of a large amount of normal fetal echocardiography data, including the normal range of clock scale values ​​for each directional line and the normal range of relative angles between directional lines. When any parameter in the spatial conformation features exceeds the corresponding reference interval, or when the matching degree between the angle combination pattern and the reference pattern is lower than a preset threshold, the early warning module outputs an early warning signal on the ultrasound diagnostic interface via the display module, prompting the physician to pay attention to possible spatial conformation abnormalities in the fetal heart.

[0072] In one specific embodiment of this example, the feature recognition module uses a convolutional neural network to automatically identify the position of the spine, the boundary of the heart, and various directional anchor points. The training data consists of multiple sets of fetal thoracic transverse ultrasound images with labeled spine position, heart outline boundary, and directional anchor point coordinates. The labels include the geometric center of the spine, the tangent directions of the left and right edges of the heart, the location of the apex of the heart, and the direction of the main blood flow outlet. The model is trained through supervised learning to learn the spatial relationship between the spine and heart structures under different fetal positions and gestational weeks. After training, it can automatically output the coordinates of the spine center and the coordinates of each directional anchor point from the input images. The coordinate information output by the model is transmitted to the coordinate system construction module and the scale calculation module, which are used to establish a clock coordinate system and calculate spatial conformation features, respectively. When the fetus rotates in utero, the coordinate system construction module detects the displacement of the spine center point in the image in real time and recalculates the phase of the reference scale direction, driving the virtual clock face to rotate synchronously, ensuring that the clock scale output of all directional lines remains constant relative to the fetus's own longitudinal axis.

[0073] The method in this embodiment does not rely on the identification of fine anatomical structures inside the heart such as the ventricular septum. Instead, it uses the spine as a stable spatial reference benchmark and combines it with a clock scale system constructed from multi-directional anchor points. This allows for stable assessment of the spatial conformation of the heart even when the fetal heart is not yet fully developed in early pregnancy or when complex heart malformations prevent the ventricular septum from being identified. At the same time, it provides structured and easy-to-learn spatial feature parameters for artificial intelligence models.

[0074] Example 5 See Figure 8 The following description uses fetal facial contour angle measurement as a complete example, taking a doctor's ultrasound examination in early pregnancy to screen for micrognathia as an example. This example illustrates the specific application of the clock coordinate system measurement method of the present invention in facial angle assessment scenarios, further demonstrating the versatility of the method across sections and measurement sites.

[0075] In step S1, the image acquisition module acquires real-time ultrasound images of the fetus in the midsagittal plane using an ultrasound examination device. The acquired ultrasound images at least include the anatomical structures of the fetal face, such as the maxilla, mandible, and hard palate, and preferably a standard midsagittal section that clearly shows the anterior borders of the maxilla, mandible, and hard palate. If the image quality is poor, grayscale enhancement, noise reduction, or edge enhancement can be performed to improve the contrast between the facial bony structures and the surrounding soft tissues.

[0076] The feature recognition module uses image recognition algorithms to automatically identify the locations of the anterior borders of the fetal maxilla and mandible in ultrasound images. The anterior border of the maxilla is a characteristic point where the fetal maxilla intersects with the anterior border of the face. Because it is located on the midline of the face and has a relatively stable echo representation in the midsagittal plane, it is suitable as a spatial reference point for facial angle measurement. The system determines the anterior border of the maxilla as the anatomical reference point and the anterior border of the mandible as the target point to be measured.

[0077] In step S2, the coordinate system construction module uses the anterior edge of the maxilla as the origin of the virtual clock coordinate system. The system identifies the direction of the hard palate line in the ultrasound image and determines the direction perpendicular to the hard palate line upwards (i.e., pointing towards the top of the fetal skull) as the reference scale direction, defining it as the 12 o'clock position of the virtual clock. The system overlays a virtual clock disk layer on the ultrasound image, with the center of the disk coinciding with the anterior edge of the maxilla, and the 12 o'clock position aligned with the direction perpendicular to the hard palate line upwards.

[0078] In this embodiment, the direction perpendicular to the hard palate line is chosen as the reference scale direction because the hard palate line is a relatively constant linear anatomical landmark in the midsagittal plane of the fetal face, and its direction reflects the spatial orientation of the maxillary plane. Using the direction perpendicular to the hard palate line as the 12 o'clock direction ensures that the reference direction of the clock coordinate system is consistent with the anatomical longitudinal axis of the fetal face, making it easier for doctors to intuitively judge the degree of deviation of the mandible from the midline of the face based on the clock scale.

[0079] In step S3, the system connects the origin (the anterior edge of the maxilla) with the target point (the anterior edge of the mandible) to form a target vector. The system then calculates the clockwise physical angle θ between the reference scale direction (12 o'clock direction) and the target vector in the image coordinate system, with a value ranging from 0° to 360°. The physical angle θ reflects the angular position of the anterior edge of the mandible relative to the anterior edge of the maxilla in the clockwork coordinate system, and its magnitude directly corresponds to the spatial relationship between the mandible and the vertical axis of the face.

[0080] In step S4, the scale calculation module calculates the physical angle θ using the formula Clock. Pos = (θ ÷ 30°) modulo operation 12 to convert to clock scale value. When Clock Pos When the calculation result is 0, it is assigned the value 12. The display module synchronously displays the virtual clock coordinate system, physical angle values, and corresponding clock orientation readings on the ultrasound image.

[0081] The following explanation uses the measurement of facial contour angles in a normal fetus as an example. In a normally developing fetus, the anterior border of the mandible is roughly located slightly below and directly in front of the face relative to the anterior border of the maxilla. Assuming the physical angle θ calculated by the system is 180°, then Clock... Pos= (180 ÷ 30) modulo operation 12 = 6, displayed as "6 o'clock direction", indicating that the anterior edge of the mandible is located directly below the anterior edge of the maxilla, that is, the facial contour has a normal upper and lower alignment relationship.

[0082] Let's take a micrognathia screening scenario as an example. Micrognathia is characterized by underdeveloped fetal mandibles, with the anterior border of the mandible significantly retracted relative to the anterior border of the maxilla. When micrognathia is present, the anterior border of the mandible is no longer directly below the anterior border of the maxilla, but shifts posteriorly (i.e., towards the deeper part of the face), causing a change in the direction of the target vector, and the corresponding physical angle θ deviates from the normal value. Assuming the physical angle θ calculated by the system is 210°, then Clock... Pos = (210 ÷ 30) modulo operation 12 = 7, displayed as "7 o'clock position". Compared with the 6 o'clock position of a normal fetus, the reading of the 7 o'clock position intuitively reflects the trend of the mandible shifting backward, suggesting that there may be micrognathia.

[0083] The early warning module compares the calculated clock scale value with the normal scale range in the preset clinical standard database of facial contour angles. Taking facial contour angle as an example, the preset clock scale range corresponding to the anterior edge of the mandible in a normal fetus is between 5:30 and 6:30. When the clock scale value calculated by the system deviates from this range, the early warning module outputs an early warning signal on the ultrasound diagnostic interface through the display module, reminding the doctor through color indicators or text prompts that there may be an abnormality in the position of the mandible, and that further confirmation is needed to determine whether there is micrognathia or other facial developmental abnormalities.

[0084] In one specific embodiment of this method, a dynamic compensation step is also included. During ultrasound examination, the fetal face may shift or rotate in the image due to fetal movement or probe angle adjustment. The coordinate system construction module detects the positional changes of the anterior edge point of the maxilla in the image in real time and recalculates the phase shift of the hard palate line direction and the corresponding reference scale direction, driving the virtual clock face to perform corresponding angle rotation compensation to ensure that the 12 o'clock direction always remains consistent with the direction perpendicular to the hard palate line upwards. Thus, regardless of how the fetal face rotates in utero, the clock scale output always reflects the true orientation of the anterior edge point of the mandible relative to the facial anatomical coordinate system, avoiding measurement reference failure and recalibration problems caused by changes in fetal position.

[0085] In one specific embodiment of this example, step S4 further includes precision compensation processing: according to Clock Pos The decimal part of the calculation result further maps the clock scale values ​​into a clock orientation expression that includes minute precision. For example, if the physical angle θ is 195°, then Clock... Pos= (195 ÷ 30) modulo operation 12 = 6.5, the integer part is 6, the decimal part 0.5 multiplied by 60 gets 30, which is mapped to "6:30 direction". Precision compensation processing enables facial angle measurement to have a finer granular angle discrimination ability while maintaining the intuitiveness of clock orientation, which helps doctors to accurately judge the subtle abnormalities of facial contour angles.

[0086] In one specific embodiment of this example, the system also supports determining multiple facial orientation anchor points within the same midsagittal plane to construct a more comprehensive description of facial spatial conformation. Facial orientation anchor points may include, but are not limited to, bony landmarks such as the tip of the nasal bone, the anterior border of the frontal bone, and the distal end of the mandible. The system connects the origin to each facial orientation anchor point to form multiple spatial orientation lines, calculates the physical angle of each orientation line relative to a reference scale direction, and converts it into clock scale values. Through the angular relationships between multiple spatial orientation lines, the system constructs the spatial conformation features of the fetal face, including an angle vector composed of the clock scale values ​​of each orientation line and the relative angle values ​​between each orientation line. The facial spatial conformation features are compared with a pre-established normal facial reference conformation range. When the facial spatial conformation features exceed the reference range, the system outputs a warning signal to assist doctors in comprehensively judging whether there are developmental abnormalities in the fetal face.

[0087] This embodiment illustrates that the clock coordinate system measurement method of the present invention can flexibly adapt to different ultrasound sections and different biological angle measurement scenarios by changing the definitions of the anatomical reference point, the target point to be measured, and the reference scale direction. In the heart axis measurement scenario, the anatomical reference point is the center of the spine, and the reference scale direction is the direction of the line connecting the spine and the sternum; in the facial contour angle measurement scenario, the anatomical reference point is the anterior edge of the maxilla, and the reference scale direction is the direction perpendicular to the hard palate line upwards. Both scenarios use a unified coordinate system construction logic, angle-scale mapping algorithm, and dynamic compensation mechanism, demonstrating the cross-section and cross-site versatility of the method of the present invention.

[0088] Finally, it should be noted that the above content is only used to illustrate the technical solution of the present invention, and is not intended to limit the scope of protection of the present invention. Simple modifications or equivalent substitutions made by those skilled in the art to the technical solution of the present invention do not depart from the essence and scope of the technical solution of the present invention.

Claims

1. A method for measuring fetal biological angle ultrasound images based on a clock coordinate system, characterized in that: Includes the following steps, S1: Acquire fetal ultrasound cross-sectional images to determine anatomical reference points and target points to be measured; S2: Using the anatomical reference point as the origin and the direction from the origin to the preset anatomical reference structure as the reference scale direction, a virtual clock coordinate system is constructed, and the virtual clock coordinate system is superimposed on the ultrasound section image; S3: Connect the origin and the target point to be measured to form a target vector, and calculate the physical angle θ between the target vector and the reference scale direction; S4: The physical angle θ is converted into a corresponding clock scale value using an angle-scale mapping algorithm, and the clock scale value is displayed on the ultrasonic cross-sectional image.

2. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: The angle-scale mapping algorithm in S4 follows the following formula: Clock Pos = (θ ÷ 30°) modulo 12; Where θ is the clockwise physical angle, ranging from 0° to 360°. Pos This is the output clock scale value; when Clock... Pos When the calculation result is 0, assign it the value 12.

3. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 2, characterized in that: The S4 also includes precision compensation processing: based on Clock Pos The decimal part of the calculation result is used to further map the clock scale value into a clock orientation expression that includes minute precision.

4. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: It also includes a dynamic compensation step: updating the coordinate system rotation angle in real time based on the displacement vector and rotation angle of the anatomical reference point.

5. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: The image recognition algorithm in S1 is an image recognition model; the reference scale direction is defined as the 12 o'clock direction of the virtual clock coordinate system.

6. A method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to any one of claims 1-5, characterized in that: The step between S1 and S2 includes an image quality assessment step: using a quality assessment model to score the standardization of the ultrasound cross-sectional image, and executing S2 and subsequent steps only when the score is higher than a preset threshold.

7. A method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to any one of claims 1-5, characterized in that: It also includes a clinical warning step: comparing the clock scale value with the normal scale range in a preset clinical standard database, and outputting a warning signal on the display interface of the ultrasound section image when the clock scale value deviates from the normal scale range.

8. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: The determination of the target point to be measured in S1 includes the following steps: A rotating ray is emitted from the origin to determine the angular range of the target's anatomical structure, and the extreme distance point is selected as the target point to be measured; the angle range is 0°-360°.

9. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: In step S4, the physical angles between the first boundary direction and the second boundary direction are calculated relative to the reference scale direction, and are converted into clock scale values ​​by the angle-scale mapping algorithm, forming a clock scale interval of the angular distribution range of the fetal target anatomical structure and displayed on the ultrasound section image.

10. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 1, characterized in that: S1 further includes the step of determining at least two directional anchor points, wherein the directional anchor points include at least two of the following types: The geometric center point determined based on the overall region of the target anatomical structure, the forward extreme point of the target anatomical structure in the ultrasound section image, the boundary point corresponding to the maximum or minimum angle within the angular distribution range of the target anatomical structure, the location point determined based on the functional connection region inside the target anatomical structure, the direction corresponding location point determined based on the blood flow or motion information of the target anatomical structure, and the feature point determined based on the regional morphological features after segmenting the target anatomical structure region. In step S3, the origin is connected to each of the direction anchor points to form multiple spatial direction lines, and the physical angle between each spatial direction line and the reference scale direction is calculated. In step S4, the physical angle between each spatial direction line is converted into the corresponding clock scale value through the angle-scale mapping algorithm, and the spatial configuration features of the target anatomical structure are constructed based on the angular relationship between the multiple spatial direction lines.

11. The method for measuring fetal bio-angle ultrasound images based on a clock coordinate system according to claim 10, characterized in that: The spatial configuration features include at least one of the following: an angle vector composed of the angle values ​​of each spatial direction line relative to the reference scale direction, the relative angle values ​​between each spatial direction line, and an angle combination pattern formed by the clock scale values ​​of multiple spatial direction lines. It also includes comparing the spatial conformation features with a pre-established reference conformation interval or reference pattern. When the spatial conformation features exceed the reference conformation interval or the matching degree with the reference pattern is lower than a preset threshold, the target anatomical structure is determined to be an abnormal conformation and an early warning signal is output.

12. A fetal bio-angle ultrasound imaging measurement system based on a clock coordinate system, using the method described in any one of claims 1-11, characterized in that: include, The image acquisition module is used to acquire real-time ultrasound cross-sectional images of the fetus; The feature recognition module is used to determine the anatomical reference point and the target point to be measured; The coordinate system construction module is used to generate a virtual clock coordinate system with the anatomical reference point as the origin and the direction from the origin to the preset anatomical reference structure as the reference scale direction, and drive the virtual clock coordinate system to rotate synchronously according to the real-time position change of the anatomical reference point. The scale calculation module is used to calculate the physical angle θ between the target point to be measured and the reference scale direction, and to convert the physical angle θ into a clock scale value using an angle-scale mapping algorithm; The display module is used to overlay the virtual clock coordinate system and the clock scale value on the ultrasound diagnostic interface in real time. The angle-scale mapping algorithm in the scale calculation module follows the following formula: Clock Pos = (θ ÷ 30°) modulo operation 12; where θ is the clockwise physical angle, Clock Pos This is the output clock scale value; It also includes an early warning module, which is used to compare the clock scale value with the normal scale range in a preset clinical standard database, and output an early warning signal through the display module when the clock scale value deviates from the normal scale range.

Images