Fetal heart rate monitoring

Abstract

The present invention relates to an ultrasound fetal heart rate monitoring system. An ultrasound probe placed on the maternal abdomen provides sensor data. The ultrasound probe first operates in an imaging mode to acquire an ultrasound image from the ultrasound probe. A trained object detection model is applied to the ultrasound image to determine whether the fetal heart is located within a target area of ​​the ultrasound image that can be addressed by the Doppler mode of the ultrasound probe. If the fetal heart is located within the target area, the ultrasound probe is switched to a Doppler mode, a Doppler ultrasound signal is acquired from the ultrasound probe, and the fetal heart rate is calculated from the Doppler ultrasound signal.

Description

Technical Field

[0001] The present invention relates to an ultrasonic fetal heart rate monitoring system and a corresponding computer-implemented ultrasonic fetal heart rate monitoring method. The present invention further relates to a computer-readable medium.

Background Art

[0002] Complications during childbirth can affect the safety of both the mother and the newborn. Worldwide, there are approximately 500,000 maternal deaths each year due to childbirth complications. Approximately 7 million women have long-term major problems after childbirth, and 50 million women face adverse health effects after childbirth.

[0003] Fetal heart rate is important information for evaluating the health of the fetus during childbirth. Currently, for this purpose, typically, a fetal monitor based on Doppler ultrasound is used. Using the Doppler effect, the fetal heart rate can be calculated based on the measured frequency shift. The fetal heart rate is measured using a plurality of ultrasonic transducer elements and based on the power of the Doppler signals received in these elements.

[0004] Unfortunately, current fetal monitoring is complex and time-consuming to use, particularly because it requires the manual (re)positioning of the fetal monitor transducer. Placing the fetal heart within the ultrasound beam is crucial for the reliability of fHR (fetal heart rate) measurement. In clinical practice, clinicians palpate the maternal abdomen to identify fetal position and then fix the ultrasound transducer in the position on the maternal abdomen where the best fHR signal can be obtained. Finding the optimal transducer position is based on listening to the intensity of the Doppler sound output and / or on signal quality indicators. Displacement of the ultrasound transducer, or displacement of the fetal heart outside the ultrasound beam, can result in loss of the fHR signal. Therefore, obstetricians often need to repeat the monotonous and tedious procedure of positioning the ultrasound transducer to avoid prolonged loss of the fHR signal. Furthermore, it is a disadvantage that those using the equipment need to be well-trained, especially because there is a risk that maternal heart rate (e.g., captured from the umbilical cord or other maternal blood vessels) may be misidentified as the fetal heart rate by the fetal monitor.

[0005] European Patent EP 3454758 B1 discloses a method for determining the optimal position of an ultrasound transducer during fetal health monitoring. An ultrasound device is used having a central piezoelectric element and additional piezoelectric elements arranged in a circular pattern around the piezoelectric element. The device supports position assistance mode and fetal heart rate mode. In position assistance mode, one element of the transducer is used to generate an ultrasound signal, and the ultrasound echo of this signal is received by each element to determine the respective Doppler signal. The user is informed which transducer has the strongest Doppler signal intensity, which helps the user reposition the device so that the central element has the strongest signal. In fetal heart rate mode, all piezoelectric elements are used for both transmitting and receiving to generate an overall Doppler signal of the fetal heart rate. [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] For example, it would be beneficial to provide a system for fetal heart rate monitoring that better helps inexperienced users, such as those monitoring at home, to ensure that the ultrasound probe used is correctly positioned. [Means for solving the problem]

[0007] According to a first aspect of the present invention, an ultrasound fetal heart rate monitoring system and a corresponding method are provided as defined by claims 1 and 14, respectively. According to a further aspect of the present invention, a computer-readable medium is provided as defined by claim 15.

[0008] Various embodiments relate to monitoring fetal heart rate by acquiring a Doppler ultrasound signal using an ultrasound probe. The Doppler ultrasound signal exhibits a frequency shift across the measured volume. Based on this frequency shift, the fetal heart rate can be determined. In particular, as is known in itself, the fetal heart rate can be determined by determining the velocity and / or direction of movement of the heart wall. In most cases, the Doppler ultrasound signal is acquired in so-called pulsed (PW) Doppler mode. In PW Doppler mode, multiple pulses are transmitted, and the relative phase changes of these transmitted pulses are used to derive the frequency shift, thereby deriving the velocity and / or direction. As an alternative to using the PW Doppler method, the continuous wave (CW) Doppler method can also be used. CW Doppler allows for the continuous transmission and reception of sound waves, recording the velocity of each along a defined path.

[0009] To ensure that the Doppler ultrasound signal actually represents a measurement of the fetal heart and not, for example, the pulsation of maternal blood vessels, the inventors of this application envision using an ultrasound probe that can also operate in an imaging mode to acquire ultrasound images. In particular, unlike probes typically used for Doppler ultrasound, this probe may have a phased array of transducers, for example, a 1D or 2D phased array. The ultrasound image data is input into an object detection model to determine whether the fetal heart is located within a target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe. The imaging mode can be, for example, a so-called B-mode (luminance mode), also known as a 2D mode. For example, a one-dimensional phased array of transducers can be used to scan a plane through the measurement volume, thereby acquiring a 2D ultrasound image, in other words, a 2D ultrasound scan frame.

[0010] While ultrasound imaging itself is known, its operation, including data acquisition and interpretation, generally requires a high level of training, and current form factors are not suitable for monitoring purposes; therefore, it is not currently widely used during active labor. However, the inventors of this application have found that ultrasound imaging can still be advantageously used in the context of Doppler-based fetal heart rate monitoring by determining whether the fetal heart is located within a target region.

[0011] A trained object detection model is applied to ultrasound images acquired in imaging mode. This object detection model is configured to locate the fetal heart in the ultrasound image, for example, by outputting the location of the fetal heart in the ultrasound image if detected, or by outputting that no fetal heart was detected. If the detected location falls within a specific target region, the ultrasound probe is switched to Doppler mode to measure the fetal heart rate, for example by using one of the phased array transducers also used in imaging mode as described above.

[0012] The method offered has several advantages. By using a trained object detection model applied to ultrasound images, the fetal heart can be located with high reliability. As a result, when switching to Doppler mode when the fetal heart is recognized, the measured fetal heart rate is also more likely to be accurately measured. By using ultrasound images, the imaging capabilities of the ultrasound sensor can be utilized. This newly available visual information can be used to utilize state-of-the-art object detection models, such as convolutional neural networks. Object detection techniques can provide excellent performance and can be run in real time or near real time on ordinary hardware. One reason for the improved accuracy is that the object detection model operates on ultrasound images, not on the Doppler signal. This means that, in order to locate the fetal heart, for example, even in the middle of labor, the model can use other parts of the maternal abdomen and / or fetal anatomical structure. For example, the fetal spine or fetal femur can be seen in ultrasound images, unlike Doppler data. Even if the object detection model was trained only to recognize the fetal heart, it can implicitly use this other information as well.

[0013] Furthermore, since the recognition is visual and not based on Doppler signals, the risk of confusing the fetal heart with other sources that generate Doppler signals is eliminated. Relying solely on the strongest Doppler signal intensity does not guarantee that the fetal heart is being measured, nor does it guarantee that other anatomical structures that produce a Doppler shift at the desired baseband are not being measured. Visual recognition is independent of the Doppler shift produced by the fetal heart itself or any other blood vessels, thus improving its accuracy and, therefore, its safety.

[0014] Another advantage is that localization can be achieved even when the fetal heart is not in the beam's focus. For example, the measurement volume of an ultrasound probe in Doppler mode can be a cylindrical area about 10 cm wide. Providing guidance based on Doppler ultrasound is effective when the fetal heart is close to its measurement volume, for example, within 20 cm. Furthermore, while Doppler-based techniques have a particularly high risk of recognizing other anatomical parts besides the fetal heart, or otherwise providing ambiguous guidance, image-based localization remains effective. For example, image-based localization is effective based on first placing the probe anywhere on the abdomen. For example, the probe could be placed at two locations within a relatively short time period, and these locations would then indicate the best position for measuring the fetal heart, even if they are not where the fetal heart itself is located.

[0015] Furthermore, the use of image detection models applied in imaging mode also has many additional advantages, relating to the fact that the system can output ultrasound images and / or locations while in imaging mode. These outputs can be used in various ways, as described elsewhere, for example, to indicate the images and / or locations and / or to focus Doppler ultrasound, to provide guidance by detecting other parts of the fetal anatomical structure even when the fetal heart is not visible.

[0016] Compared to using the probe only in imaging mode, using Doppler mode to measure fetal heart rate offers numerous advantages. One advantage is the availability of current reliable measurement techniques for Doppler mode. Furthermore, Doppler mode can use lower power than imaging mode. For example, using Doppler mode requires 20 mW / cm². 2While the imaging mode uses less than that, the system uses more. As a result of reduced power consumption, the system can be made more suitable for continuous monitoring; for example, the system can be used for monitoring for at least 1 hour, at least 4 hours, or at least 10 hours. In particular, during such long-term use, less power is required, for example, 20 mW / cm². 2 Using less power helps prevent damage to bones and skin. Lower power also reduces the heat generated by the device, making it more comfortable on the abdomen and improving the lifespan of the device's components.

[0017] Using the same transducer, both ultrasound images and Doppler ultrasound signals can be captured. For example, the Doppler ultrasound signal can be captured by one of the transducers also used for imaging. In this way, relatively little hardware is required, and since the same transducer is used in both cases, it is guaranteed that the position of the ultrasound image in imaging mode coincides with the area that the ultrasound sensor can handle in Doppler mode.

[0018] The target region for switching to Doppler mode can coincide with the entire ultrasound image, but is preferably a sub-region of the entire ultrasound image, for example, a sub-region excluding the boundary region of the ultrasound image, e.g., a sub-region excluding at most or at least 10%, at most or at least 20%, or at most or at least 30% of the imaged region. In this way, when the probe is switched to Doppler mode, then moved slightly so that the signal is lost, and then switched back to imaging mode, the fetal heart is prevented from going out of range of the ultrasound sensor. The region that can be handled by the ultrasound probe in Doppler mode often coincides with the region captured by the ultrasound image, for example, the entire image can be handled in Doppler mode. Also, in such cases, the target region for which the device switches to Doppler mode is typically a sub-region of the entire region that is imaged and / or Doppler-measured.

[0019] Optionally, when in Doppler mode, a signal quality index for the Doppler ultrasound signal may be determined, for example, using known techniques. If this signal quality index does not meet a predetermined quality threshold, the ultrasound probe may be returned to imaging mode, and the object detection model may be repeatedly applied, for example, until the location of the fetal heart is identified. Thus, the accuracy of the measured fetal heart rate is further ensured; that is, in imaging mode, the fetal heart rate is not measured until the fetal heart is within the desired target region, and in Doppler mode, the signal quality index can ensure sufficient quality for accurate measurement of the heart rate. The quality threshold is calibrated so that the signal quality is generally sufficient, at least when the fetal heart is within the target region. However, it is possible to have a stricter definition of the target region, for example, the signal quality may be sufficient even outside the target region. However, the opposite situation, where the signal quality is insufficient even when the fetal heart is within the target region, is generally undesirable.

[0020] In any case, it is preferable that the quality threshold is calibrated so that the threshold is approximately in the region captured by the ultrasound image, including, for example, a safety margin. In this way, when the threshold is reached, it is possible to automatically switch to an imaging mode for recognizing the fetal heart, and in some cases, to automatically switch back to Doppler mode using the updated position where the Doppler ultrasound beam is focused, for example by beam steering, without requiring the probe to be repositioned.

[0021] Optionally, signal quality indicators may be output in a way that is perceptible to the user. In this way, the user is informed whether the probe is in a good position on the abdomen and can try to move the probe to improve signal quality. It is particularly convenient for the user to use visual indicators, such as one or more lights, on the ultrasound probe itself.

[0022] Optionally, the ultrasound probe has one or more transducers in the form of piezoelectric transducers and / or capacitive ultrasonic transducers (CMUTs). For example, these transducers can form a phased array of multiple transducers, e.g., a linear phased array, enabling the generation of a 2D ultrasound image. These transducers can also form a grid of transducers, e.g., a matrix array, enabling the generation of a 3D ultrasound image.

[0023] Optionally, the Doppler ultrasound beam may be focused to the position where it is placed. In other words, the window for Doppler measurement, including depth, can be set according to the position. Various ultrasound probes, particularly those based on transducer arrays, support the use of beamforming to focus the Doppler ultrasound beam to a specific position. Focusing is particularly advantageous in combination with the use of object detection to determine the position of the fetal heart, since the ultrasound beam is focused to that specific location. For example, the fetal heart does not need to be placed in a cylinder directly beneath the ultrasound probe, as in some prior art examples. This makes the ultrasound probe particularly easy to use, for example, with little or no training, because it requires little precision in the placement of the ultrasound probe.

[0024] This technique offers several advantages compared to using depth selection. Focusing the Doppler ultrasound beam can include setting not only the depth but also the horizontal and vertical positions perpendicular to that depth. This has the advantage that simply setting the depth cannot handle situations where multiple signal sources are present in that depth region. Furthermore, while depth selection requires a good initial signal to focus on the depth, the technique provided can provide guidance information even when there is no fetal heart in the imaged region. This technique also improves upon the slow and iterative windowing of depth selection and avoids the problem of the window becoming completely open when the signal-to-noise ratio deteriorates.

[0025] Optionally, an ultrasonic image and / or the position of the fetal heart in this ultrasonic image are shown on the display. For example, the ultrasonic image, or for example, a graphic image of the maternal abdomen such as an avatar, may be shown together with a square, cross or the like used to indicate the detected position. This can serve several purposes. Since the user can see on the screen where the fetal heart is placed both with respect to the image generated by the ultrasonic probe and with respect to the probe itself, it can be useful for guiding the ultrasonic probe.

[0026] Optionally, showing a graphic image instead of, for example, the ultrasonic image itself can enable easier interpretation and is preferable, for example, for legal reasons when the user is not permitted to view and / or make judgments with the ultrasonic image itself, or, for example, when the mother does not want this image to be seen. The ultrasonic image is processed in the background and is not stored or displayed to the user. Optionally, a target area for switching to Doppler mode is also visualized to provide guidance to the user when a switch to Doppler mode is expected.

[0027] When showing the ultrasonic image itself, displaying the position where the fetal heart is detected can also be used as a reliability mechanism by enabling the user to check whether the system has correctly recognized the fetal heart. For example, when the system switches to Doppler mode and / or when the focus of the Doppler ultrasonic beam is adjusted using the recognized fetal heart, the user can verify whether this is based on a correct recognition of the heart.

[0028] Optionally, in addition to the fetal heart, the object detection model is configured to locate one or more fetal body parts of the fetus. These body parts can include the fetal head, fetal spine, and / or fetal femur. The fact that ultrasound images are used instead of Doppler signals allows these body parts to be recognized as well. Outputting these additional localizations is advantageous because this information can make it easier to place the ultrasound probe in the correct position, especially in situations where the fetal heart itself cannot be seen within the image.

[0029] One way to output this additional localization is to show it on a display. For example, the ultrasound image can be shown along with the localization of any fetal body part configured to be located by the object detection model. These localizations can help a user, such as a particularly inexperienced user like a nurse or midwife, determine the orientation of the fetus and thereby place the probe in order to operate it appropriately.

[0030] Another method that uses at least the localization of the fetal heart and optionally also the localization of any other fetal body part located by the object detection model is to determine guidance information for guiding the ultrasound probe to the fetal heart. This guidance information can be output to the display. Interestingly, this guidance information can propose moving the probe to different positions on the maternal abdomen rather than simply suggesting a rotation or change in the angle of the probe with respect to the maternal abdomen.

[0031] In particular, this is based on the location of one or more of the fetal head, fetal spine, and fetal femur. Especially when the fetal spine is detected but the fetal head is not, the guidance information can be used to determine guidance information for guiding the ultrasound probe to the expected position of the fetal head. This has the advantage of successful placement without relying too much on the initial placement of the probe, as the fetal spine can be oriented at any position in the uterus. Furthermore, the position of the fetal spine provides very useful information for the overall anatomical structure of the fetus. Given the position of the spine, it is possible to suggest the position of the fetal heart and position the ultrasound probe to capture the fetal heart. Also, locating the fetal head is beneficial because this fetal head can be recognized with a fairly high degree of reliability, and when the ultrasound probe is placed at the position of the fetal head and both the fetal head and fetal spine are detected, guidance to the fetal heart can be suggested with particularly high reliability. Then, guidance information for guiding the ultrasound probe to the expected position of the fetal heart is determined.

[0032] To determine guidance information, for example, an algorithm can be used that internally reconstructs the geometric shape of the fetus based on the localization. For example, a machine learning-enabled guidance model trained to output guidance based on a labeled dataset can also be used.

[0033] Guidance information can be presented in various ways, for example, visualized as suggested positions on a grid, such as a 3x3 grid or similar, which is extended onto an image of the maternal abdomen. Examples are shown herein. Guidance information can be combined, if necessary, with visualizations of determined fetal anatomical structures and / or actual images from an ultrasound probe. Guidance information may be presented without displaying (and possibly without storing) actual ultrasound images, which, as mentioned elsewhere, makes the visualization easier to interpret and is useful in situations where displaying ultrasound images is undesirable.

[0034] Guidance based on the localization of body parts is a particularly effective guidance mechanism that can determine the overall position of the fetus, and has the added advantage of determining not only the position of the fetal heart itself. Furthermore, this type of guidance is well known to those familiar with clinical practice, where the person placing the ultrasound probe typically follows a protocol in which the clinician palpates the maternal abdomen in a certain pattern, e.g., upper → lower → left → right, in order to reveal the anatomical structure of the fetus. Protocols like those used today can actually take 15-20 minutes, even for experienced individuals. The guidance technology provided makes it possible to significantly reduce this time and allows the probe to be used even by inexperienced or unqualified users (nurses, midwives) during the prenatal period. The system aims to guide the probe to the head first, then to the heart, once the spine is detected, but in the meantime, if a sufficiently reliable heart has already been detected, it is possible to switch directly to Doppler mode when detecting it. However, it is also possible to detect only the fetal heart and / or switch to Doppler mode after both the spine and head have been detected, for example, to allow for the complete determination of the fetal orientation before switching modes.

[0035] Optionally, the orientation of the ultrasound probe is used to determine guidance information. For example, the ultrasound probe may have an inertial measuring device (IMU) that provides the orientation angle of the ultrasound probe. This orientation may be input into an object detection model to determine the position of the fetal heart and / or other body parts, and may be used to determine guidance information. Interestingly, given a spherical shape for the maternal abdomen, the orientation of the ultrasound probe can provide a fairly accurate estimate of the position of the ultrasound probe on the abdomen, and is used not only as input to the object detection model, but also as input to the object detection model, particularly to determine guidance, as it represents the current position of the probe.

[0036] The above embodiments relate to systems and methods for using machine learning-capable object detection models and / or guidance models. Systems for training object detection models and / or guidance models for these uses, as well as corresponding computer implementation methods, are also envisioned. This training can be performed using techniques known in themselves. For example, gradient-based training, such as stochastic gradient descent, can be performed using the Adam optimizer disclosed in Kingma and Ba, “Adam: A Method for Stochastic Optimization” (available at https: / / arxiv.org / abs / 1412.6980 and incorporated herein by reference). As is known, such optimization methods are heuristic and / or reach local optima.

[0037] Those skilled in the art will understand that two or more of the above-described embodiments, implementations, and / or optional aspects of the present invention can be combined in any way that is deemed useful.

[0038] Modifications and variations for any system and / or any computer-readable medium correspond to the modifications and variations described in the corresponding computer implementations and can be performed by those skilled in the art in accordance with this specification. [Brief explanation of the drawing]

[0039] These and other aspects of the present invention will be further revealed and described with reference to embodiments described as examples with reference to the following description and accompanying drawings. [Figure 1] Figure 1 shows an ultrasound fetal heart rate monitoring system. [Figure 2A] Figure 2A shows a detailed example of how to monitor fetal heart rate. [Figure 2B] Figure 2B shows a detailed example of how to switch the ultrasound probe from imaging mode to Doppler mode. [Figure 3]Figure 3 shows a detailed example of locating the fetal heart. [Figure 4A] Figure 4A shows a detailed example of how to determine guidance information. [Figure 4B] Figure 4B shows a detailed example of how to determine guidance information. [Figure 4C] Figure 4C shows a detailed example of how to determine guidance information. [Figure 5] Figure 5 shows a computer-implemented ultrasound fetal heart rate monitoring method. [Figure 6] Figure 6 shows a computer-readable medium containing data.

[0040] Please note that the diagram is merely a general representation and is not drawn to scale. In drawings, elements corresponding to elements that have already been described may have the same reference number. [Modes for carrying out the invention]

[0041] Figure 1 shows the ultrasound fetal heart rate monitoring system 100.

[0042] System 100 has a data interface 120 for accessing model data 030 representing a trained object detection model. The object detection model is configured to locate a fetal heart in an ultrasound image. The model data 030 may have a set of trained parameters, for example, at least 10,000 or at least 100,000 parameters. For example, the model may be a neural network, also known as an artificial neural network. In particular, the model may be a deep neural network or a convolutional neural network. In the case of a neural network, the set of parameters 030 has the weights of the nodes of the neural network. For example, the number of layers in the model may be at least 5 or at least 10, and the number of nodes and / or weights may be at least 1,000 or at least 10,000. The data interface 120 is for accessing additional information, such as that described herein, for example, model data representing a trained guidance model.

[0043] Model data representing object detection models and / or guidance models are obtained, for example, by training the models by system 100 itself or by another system having a similar hardware architecture, in particular by a system having the data interface and processor system described herein.

[0044] For example, as shown in Figure 1, the data interface 120 can be configured by a data storage interface that can obtain data 030 from data storage 021. For example, the data storage interface 120 may be a memory interface or persistent storage interface, such as a hard disk or SSD interface, or a personal LAN or WAN network interface, such as a Bluetooth®, Zigbee®, or Wi-Fi interface, or an Ethernet® or optical fiber interface. The data storage 021 may be the internal data storage of the system 100, such as a hard drive or SSD, or an external data storage, such as a network-accessible data storage. In some embodiments, each piece of data is obtained from different data storage, for example, via different subsystems of the data storage interface 120. Each subsystem may be of the type described above relative to the data storage interface 120.

[0045] The system 100 may further have a sensor interface 160 for acquiring sensor data 124 from the ultrasound probe 071. The ultrasound probe 071 is configured to be placed in the maternal abdomen. The ultrasound probe 071 is operable in imaging mode and Doppler mode. Typically, the ultrasound probe 071 operates in imaging mode or Doppler mode, but not in both modes simultaneously. The ultrasound probe 071 has one or more transducers. For example, the transducer may be a piezoelectric transducer, or the transducer may be a CMUT transducer. The use of a CMUT transducer is preferred because it has lower manufacturing costs.

[0046] The transducers may be arranged in a configuration suitable for 2D and / or 3D ultrasound imaging phased arrays, for example, in a 1D or 2D phased array configuration. For example, the number of transducers in a phased array can be at least 2, at least 10, or at least 50. For example, the transducers may be arranged in a 1D phased array configuration, for example, a linear phased array having at least 2, at least 10, or at least 50 transducers. As another example, the transducers may be arranged in a 2D phased array configuration, for example, a matrix phased array. A matrix phased array can have at least 10 or at least 50 transducers in both directions and be arranged, for example, in a rectangular pattern. Such a transducer pattern is used for imaging, while it can also be used in Doppler mode, for example, by using one specific transducer to generate one Doppler signal, or by using multiple respective transducers to generate multiple respective Doppler signals. In particular, transducers typically do not use a circular arrangement of piezoelectric elements with a single piezoelectric element in the center, as this arrangement does not work well for imaging. The ultrasonic probe 071 can operate at frequencies of at least 1 MHz and / or up to 5 MHz, for example.

[0047] In imaging mode, the ultrasound probe 071 can provide sensor data 124 representing an ultrasound image. The ultrasound image can represent the amplitude of the echoes of the ultrasound signal transmitted by the transducer. In particular, the imaging mode can be the so-called B-mode. For example, a two-dimensional image can be acquired using a one-dimensional phased array. It is also possible to acquire a three-dimensional image using a two-dimensional phased array. For example, it is also possible to use a one-dimensional phased array to acquire a three-dimensional image, as disclosed in MNSenlik and H. Koymen, "Radiation Impedance of an Array of Circular Capacitive Micromachined Ultrasonic Transducers", IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 57, no. 4, April 2010.

[0048] In Doppler mode, a Doppler ultrasonic signal representing the velocity of the volume being measured is acquired. For example, the Doppler mode may be a pulsed (PW) Doppler mode, in which ultrasound is transmitted and received alternately, enabling the measurement of the velocity at a specified depth. By setting the direction of the signal as it is transmitted, for example by using beamforming, the Doppler ultrasonic beam is focused at a specific location. Continuous wave (CW) Doppler mode can also be used. In this mode, the direction of the signal is typically still controlled, but it does not allow selection of a specific depth.

[0049] Typically, since imaging mode and Doppler mode use different ultrasonic signals, both imaging mode and Doppler mode operate with the same set of transducers, but not simultaneously. For example, in type B imaging mode, the beam can be transmitted by separate transducers with different phase differences to scan in a specific direction. In Doppler mode, for example, pulses may be transmitted by only one transducer, or the same pulse may be transmitted simultaneously by multiple transducers. Therefore, the ultrasonic probe 071 is configured to operate in imaging mode or Doppler mode via control signals 124 transmitted from system 100 to the ultrasonic probe 071 via sensor interface 160.

[0050] The Doppler ultrasound beam may be focused to a specific location, in particular, a location where the fetal heart is located by an object detection model such as that described herein. The beam is focused by beamforming, which typically involves signal processing, for example, in the ultrasound probe 071 itself or by the processor subsystem 140.

[0051] The ultrasonic probe 071 may further include an orientation sensor, such as an IMU sensor. The processor subsystem 140 is configured to acquire orientation data 124 indicating the orientation of the ultrasonic probe, for example, having the angle of the sensor 071, via the sensor interface 160. This orientation data is used to determine guidance information, as described herein.

[0052] System 100 further includes a processor subsystem 140 configured to operate the ultrasound probe 071 in imaging mode and acquire an ultrasound image 124 from the ultrasound probe 071 while System 100 is operating. The processor subsystem 140 is further configured to apply an object detection model 030 to the ultrasound image 124 to determine whether the fetal heart is located within a target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe 071. If the fetal heart is located within the target region, the processor subsystem 140 is further configured to switch the ultrasound probe 071 to Doppler mode, acquire a Doppler ultrasound signal 124 from the ultrasound probe, and calculate the fetal heart rate from this Doppler ultrasound signal 124.

[0053] System 100 may further have an output interface 180. The output interface 180 is used for various outputs, including fetal heart rate calculated from the Doppler ultrasound signal, ultrasound image, the position of the fetal heart in the ultrasound image, a signal quality index of the Doppler ultrasound signal, and / or guidance information. For example, the output interface is an output interface to rendering devices such as a display, light source, speaker, or vibration motor, which may be used to generate a perceptible output signal that is generated based on the output information. For example, as shown, the output interface may be for a display 190. In Doppler mode, for example, the fetal heart rate 192 is displayed and / or a warning is issued if the fetal heart rate is abnormal and / or a signal quality index of the Doppler ultrasound signal is output. In imaging mode, for example, the ultrasound image, the position of a identified fetal body part, and / or guidance information are shown. As another example, the output interface may be configured by a data interface 120, which in these embodiments is an input / output ("IO") interface through which an output, such as a fetal heart rate, is stored in data storage 021. This output interface may also provide an output, such as the determined fetal heart rate, for further processing, for example, the fetal heart rate may be provided to a fetal heart rate monitoring module operated by a processor subsystem 140.

[0054] The processor subsystem 140 may be further configured to determine a signal quality index of the Doppler ultrasound signal 124. If this signal quality index does not meet a predetermined quality threshold, the ultrasound probe 071 is returned to imaging mode via the control signal 124. Instead of using the signal quality index to control the mode of the ultrasound probe, or in addition to that, the processor subsystem 140 may be configured to output the signal quality index in a way that is perceptible to the user. This can be done via the output interface 180, for example, on a display 190, and / or via the ultrasound probe's visual indicator 072, for example, by one or more lights, for example, using the color, intensity and / or several lights of the visual indicator 072 used to indicate signal quality.

[0055] The system 100, ultrasound sensor 071, and / or display 190 can be configured in various ways. For example, the system 100 and probe 071 may form a single device, or a processor subsystem 140 that determines the fetal heart rate by applying an object detection model may be included in or fixedly connected to the ultrasound sensor 071 or the ultrasound sensor 071. The system 100 may also be implemented as a fetal monitor (connectable to or connected to the probe 071) with a built-in or externally connectable display 190. The system 100 may also be implemented on the cloud, and data may be provided to a mobile device that displays the fetal heart rate on the display 190, for example. The system 100 itself may be located on a mobile device, for example, by providing its functionality as a mobile application.

[0056] More generally, each system described herein, including but not limited to system 100 in Figure 1, can be embodied as a single device or apparatus, such as a workstation or server, or in a single device or apparatus. This device or apparatus may be an embedded device. The device or apparatus may have one or more microprocessors that run appropriate software. For example, the processor subsystem of each system may be embodied by a single central processing unit (CPU), but may be embodied by a combination or system of such CPUs and / or other types of processing units. The software may be downloaded and / or stored in corresponding memory, such as volatile memory, such as RAM, or non-volatile memory, such as flash memory. Alternatively, the processor subsystem of each system may be implemented in the device or apparatus in the form of programmable logic, for example, as a field-programmable gate array (FPGA). Generally, each functional unit of each system can be implemented in the form of a circuit. Each system may be implemented in a distributed manner, including different devices or apparatus, such as a distributed local server or a cloud-based server.

[0057] System 100 may instead, or in addition to, be a training system for training an object detection model. Such a training system does not need to have a sensor interface 160 and / or an output interface 180, and for example, a data interface 120 can be used to access the training data and / or model being trained.

[0058] Figure 2A shows a detailed, but non-limiting, example of how to monitor fetal heart rate using an ultrasound fetal heart rate monitoring system as described herein. The figure shows an overall flowchart illustrating several actions taken to detect the fetal heart and plot the fetal heart rate in real time, for example, on a mobile application. According to the figure, the fetal heart is automatically detected in real time from the 2D ultrasound scan frame, and once the fetal heart is detected, the system can switch from 2D image mode to pulse wave mode to plot the fetal heart rate.

[0059] In positioning operation POS210, the ultrasound probe of the fetal heart rate monitoring system may be positioned on the patient's abdomen, for example, by a midwife.

[0060] In the fixed operation FIX220, the ultrasound probe may be repositioned on the abdomen to fix its position and / or to acquire a trace. This can be done, for example, by a midwife.

[0061] In detection operation DET230, the fetal heart is detected and displayed, for example, on a screen connected to a fetal heart rate monitoring system. One or more piezoelectric / CMUT sensors can be used to capture 2D ultrasound data from the initial position. An object detection algorithm can detect fetal anatomical structures (e.g., fetal heart, fetal head, fetal spine, and / or fetal femur) on the ultrasound image acquired from the initial position and guide the user to the next position, for example, by suggesting where the probe should be moved. An algorithm may be provided using an IMU sensor to detect the fetal heart at the angle of the sensor. If the fetal heart has not yet been detected, the process may return to the fixing operation FIX, which fixes the ultrasound probe in place. The system can further repeat detection operation DET. Optionally, as described elsewhere, if the fetal heart has not yet been detected, information to help position the ultrasound probe correctly may be provided to the user, for example, in the form of the ultrasound probe and / or the position of the fetal organs it is placed on, and / or in the form of guidance information.

[0062] When a fetal heart is detected, the probe switches to Doppler mode, allowing for continuous calculation of the fetal heart rate. In particular, the MON240 monitoring mode enables Doppler mode, allowing the fetal heart rate monitoring system to calculate the fetal heart rate from the Doppler ultrasound signal. While in Doppler mode, the signal quality index of the Doppler ultrasound signal can be determined as a known indicator. The signal quality index may be output to the user, for example, in the form of one or more lights on the ultrasound probe itself or another visual indicator.

[0063] In interpretation operation INT250, the trace and / or fetal heart rate are interpreted, for example, by a midwife and / or gynecologist.

[0064] In the out-of-field operation OOV260, the fetal heart is determined to be outside the field of view, for example, because the signal quality of the Doppler ultrasound signal is insufficient to reliably determine the fetal heart rate, for example, because it does not meet a predetermined quality threshold. In this case, the ultrasound probe can be fixed again, for example, by returning the probe to imaging mode. Interestingly, switching back to imaging mode allows the user to help refocus on the fetal heart so that fetal heart rate monitoring continues. The quality threshold is defined as the point at which the fetal heart is still generally within the range of the ultrasound image captured in imaging mode. In this way, when the signal quality falls below the threshold, the fetal heart is indicated and / or guidance information based on the position of the fetal heart is determined, making refocusing relatively easy.

[0065] Figure 2B shows a detailed, but non-limiting, example of how to switch an ultrasound probe from imaging mode to Doppler mode.

[0066] In acquisition operation Aq221, the ultrasound probe operates in imaging mode to acquire an ultrasound image UI222 from the ultrasound probe. The ultrasound image can be, for example, at least 32×32 pixels, at least 128×128 pixels, or at least 512×512 pixels. This image may be grayscale.

[0067] In the model application operation Appl223, the object detection model OD224 is applied to the ultrasound image UI, and localization information LOC225 can be determined, indicating whether the fetal heart is located within the target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe.

[0068] As is well known, an object detection model OD can output localization information LOC in the form of class probabilities corresponding to one or more types of objects to be detected, combined with, for example, the location of the recognized object, such as the position of the recognized object and optionally its size. This location can be represented, for example, as one or more coordinates in an ultrasound image, a bounding box, etc. Various object detection models are known and can be used.

[0069] In one embodiment, the object detection model comprises a neural network, particularly a deep neural network. In particular, the inventors of the present application have achieved good results using the YOLO object detection model from J. Redmon et al., “You Only Look Once: Unified, Real-Time Object Detection” (available at https: / / arxiv.org / abs / 1506.02640, which is incorporated herein by reference). YOLO is a region-proposed network that applies a single neural network to the entire image to detect a specific object. The network divides the image into multiple regions and predicts bounding boxes and probabilities for each region. YOLO divides the input image into an S×S grid. If the center of an object falls within a grid cell, that grid cell is responsible for detecting that object. Each grid cell predicts B bounding boxes and confidence scores for these boxes. The confidence scores reflect whether the boxes truly contain the object of interest and also reflect the prediction accuracy. Each bounding box has five predictions x, y, w, h and confidence scores. The (x,y) coordinates represent the center of the box relative to the grid cell boundaries. The width (w) and height (h) are predicted for the entire image. The confidence level of the YOLO output is a regression trained to output the degree of overlap (IoU) between the output bounding box and the ground truth bounding box. Other models based on the same principle can also be used. As a specific example, the YOLOv3-tiny model can be applied to a 416x416 image.

[0070] Specifically, the object detection model OD may be trained and applied to 2D fetal heart four-chamber images. Positive training examples may include axial images of the fetal heart and can be collected and curated, for example, from a cine loop of 2D fetal scans. Planes that do not contain the fetal heart are used as negative training examples. The training dataset can be created using these images by labeling them as fetal heart images and annotating them with the bounding box of the fetal heart, or labeling them as non-fetal heart images. These images may be preprocessed to remove text annotations as needed. Annotations can be made, for example, using the DarkLabel tool. Training can be performed using the Darknet deep learning framework. The inventors of this application have found that training with 4000 iterations using a batch size of 64 images yields good results.

[0071] For example, the full training and testing process may include starting with data, annotation, cleaning of annotated data, training data based on, for example, patient IDs, splitting into validation and test data, training multiple models on the training data by varying hyperparameters, validating the models on validation data, finalizing the model with the highest validation accuracy, testing the final model on unknown data, and calculating evaluation metrics.

[0072] Based on the localization information (LOC), it can be determined whether the fetal heart is located within a target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe. For example, this is the case when the fetal heart is recognized by the model with at least a given threshold confidence, and the location where this fetal heart is recognized is within the target region. The target region can correspond, for example, to the ultrasound image UI itself or its subregions. If the fetal heart is located within the target region, the ultrasound probe can be switched to Doppler mode Sw227, as discussed elsewhere, to begin monitoring the fetal heart rate, and optionally using the localization information (LOC) to focus the Doppler ultrasound beam to the localized location. Acquired action Aq and subsequent steps may be repeated unless the fetal heart is found to be within the target region. Furthermore, the localization information (LOC) may be used in various ways, in particular, as discussed in more detail elsewhere, to determine guidance information to help the user guide the ultrasound probe to the fetal heart by outputting and / or using the location.

[0073] Figure 3 shows a detailed but non-limiting example of localizing the fetal heart. In this example, an object detection model is used to localize the fetal heart in the ultrasound image 310 by outputting a bounding box 320 where the fetal heart is detected. The bounding box 320 indicating the location of the fetal heart in the ultrasound image 300 and ultrasound image 310 are shown in this example on the user interface 300 shown on the display. As shown, the object detection model can optionally output a confidence value of the detection, which is also shown on the display, for example, as a label (0.985 in this example), as the color or line type of the bounding box, etc. If the fetal heart is not detected, for example, if the confidence value does not exceed a threshold, the bounding box is typically not shown. The object detection model is configured to detect one or more additional fetal anatomical structures, such as the fetal head, fetal spine, and fetal femur. If detected, localization to these additional anatomical structures is also shown on the ultrasound image 310.

[0074] Figures 4A–4C show a detailed but non-limiting example of determining guidance information for guiding the ultrasound probe to the fetal heart.

[0075] As discussed with respect to Figure 2B, for example, guidance information is determined from the position of one or more fetal body parts recognized by an object detection model in the ultrasound image from the ultrasound sensor while the ultrasound sensor is in imaging mode. The guidance information is further based on the orientation of the ultrasound probe, which is determined by the IMU sensor or other orientation sensors. Interestingly, the position of the IMU sensor on the abdomen can be determined based on the orientation of the IMU sensor and using the geometric shape of the abdomen; for example, if the IMU sensor is oriented vertically, the IMU sensor is placed in the center of the abdomen, and if it is tilted to the left, the IMU sensor is placed on the left side of the abdomen, and so on. Other methods are also possible for determining the position of the ultrasound sensor on the abdomen, for example, based on the ultrasound image, and these can similarly be used to determine the guidance information.

[0076] Figures 4A–4C show fetal heart rate monitoring systems 410, 420, and 430 equipped with displays, on which guidance information is displayed in the form of suggested positions 411, 412, and 413 where the ultrasound probe should be moved. In this example, the suggested position is determined by the granularity of the suggested cell in a grid representing the abdomen. In this particular example, a 3x3 grid is used. The position is highlighted on a visualization of the maternal abdomen on which this location is magnified. This visualization can be a standard image or figure of the maternal abdomen. Thus, the guidance can be used without the need to store or display the ultrasound image used to determine this guidance, which is beneficial as will be mentioned elsewhere. The highlighting of the position may indicate the progress of the guidance. For example, on the fetal monitor shown in Figures 4A–4C, the color of the highlighted positions 411–431 in the examples of Figures 4A and 4B indicates that the fetal heart has not yet been detected, while the highlighting color in Figure 4C indicates that the fetal heart has been detected. Visualizations 440, 450, and 460 show alternative position highlighting in the form of proposed grid cell boundary styles to which the probe should be moved. As illustrated, cell boundaries can have different colors and / or thicker boundaries than other cells.

[0077] Visualizations 440–460 also show that the estimated position of the fetus in the abdomen is visualized. This can be based on standard images or illustrations of the fetus. For example, the entire fetus can be visualized, as illustrated in Figure 460, or a portion of the fetus can be shown, as shown in Figures 440 and 450, where this portion corresponds to a known position of the fetal anatomical structure, for example, based on the position where the ultrasound probe has been moved and / or the fetal anatomical structure that has been located so far. An exemplary guidance procedure is illustrated. Using this guidance procedure, a user, for example, a caregiver, can position the probe toward the fetal heart much faster (reducing the need for manual palpation through automated object detection) while following a familiar existing protocol.

[0078] The guidance may begin by suggesting an initial position for placing the ultrasound probe. As shown in Figure 4A, this initial position is, for example, the left center of the abdomen. This position is often used in current protocols because it typically allows detection of the fetal vertebrae, enabling easier extrapolation of the rest of the anatomical structure.

[0079] If the fetal vertebrae cannot be detected, alternative initial positions, such as right-center, upper-center, or lower-center, may be proposed.

[0080] If the fetal spine is detected but the fetal head has not yet been detected, guidance information is determined to guide the ultrasound probe to the expected position of the fetal head. For example, the expected position of the fetal head is determined under the assumption that the fetus is placed head down and reaches the proposed position 421 in Figure 4b.

[0081] If the fetal head is not detected, an alternative expected position is determined and proposed, for example, based on the assumption that the fetus is positioned head-up.

[0082] If the fetal spine and fetal head are detected, guidance information can be determined to direct the ultrasound probe to the expected location of the fetal heart. Once the positions of the fetal spine and fetal head are identified, the expected location of the fetal heart can be precisely determined based on the normal fetal geometric shape. This is illustrated by proposal 431 in Figure 4C, in which the fetal heart is proposed to be located in the central cell of the grid.

[0083] If the fetal heart is not detected, an alternative location for the fetal heart can be determined, or the process can be restarted, for example, by detecting the spine.

[0084] If a fetal heart is detected, the system can switch to Doppler mode to monitor the fetal heart, as described. This can also occur while the guidance is still searching for the fetal spine or fetal head. In this case, it is possible to switch directly to Doppler mode, but if necessary, the guidance process can also be continued first to locate the fetal spine and fetal head, and to determine the complete position and orientation of the fetus. This information can be output to the user, for example, or it can be used automatically, for example, using a trained machine learning model, to predict the angle at which the fetal head will progress during labor and / or delivery.

[0085] Several methods are possible to determine guidance information. One possibility is to use a clear geometric model of the fetal anatomical structure. This model is relatively simple; for example, the fetus is modeled as a straight line representing the spine and ends with a sphere representing the fetal head. The position of the fetal heart is determined relative to the determined positions of the straight line and sphere. An alternative is to use a machine learning-capable guidance model trained to output guidance information, for example, cells in a grid as shown in Figures 4A-4C. For example, this guidance model has separate models used to determine the positions of the fetal spine, fetal head, and fetal heart, respectively, in a subsequent step following the locating of the fetal heart. Another option is to apply a trained feature extractor to extract one or more features from the ultrasound image UI and compare these extracted features with a trained fetal image dataset for which guidance information is available. For example, a general-purpose feature extractor can be used in this case.

[0086] Figure 5 shows a block diagram of a computer implementation method 500 for ultrasound fetal heart rate monitoring. Method 500 corresponds to the operation of system 100 in Figure 1. However, this is not limiting, and Method 500 may be performed using a different system, apparatus or device. Method 500 may have a step 510 in an operation titled “Obtain Sensor Data” to obtain sensor data from an ultrasound probe positioned on the maternal abdomen. Method 500 may have a step 520 in an operation titled “Access Object Detection Model” to access model data representing a trained object detection model. The object detection model is configured to locate the fetal heart in the ultrasound image. Method 500 may have a step 530 in an operation titled “Ultrasound Imaging” to operate the ultrasound probe in imaging mode to obtain an ultrasound image from the ultrasound probe. Method 500 may include a step 540 in an operation titled “DETECT HEART” to determine whether the fetal heart is located in a position on the ultrasound image that can be handled by the Doppler mode of the ultrasound probe. Method 500 may include a step 550 in an operation titled “HEART DETECTED?” to determine whether the fetal heart is located within the target region. Method 500 may include a step 560 in an operation titled “SWITCH TO DOPPLER” to switch the ultrasound probe to Doppler mode if the fetal heart has been detected. Method 500 may further include a step 570 in an operation titled “OBTAIN DOPPLER SIGNAL” to acquire a Doppler ultrasound signal from the ultrasound probe. Method 500 may further include a step 580 of calculating the fetal heart rate from the Doppler ultrasound signal in an operation titled “Monitor Fetal Heart Rate (MONITOR FHR)”. Generally, the operation of Method 500 in Figure 5 may be performed in any suitable order, for example, sequentially, simultaneously, or in combination thereof, provided that a specific order is required by the input / output relationship, for example, where applicable. The method may be combined with further steps, for example, the object detection model and / or guidance model described herein may be trained before application. A separate computer implementation method for training the object detection model and / or guidance model is also conceivable.

[0087] The method may be implemented on a computer as a computer implementation method, either as dedicated hardware or as a combination of both. As shown in Figure 6, computer instructions, such as executable code, are stored on a computer-readable medium 600, for example, in the form of a series of machine-readable physical marks 610 and / or as a series of elements having different electrical, magnetic, or optical properties or values. The medium 600 may be temporary or non-temporary. Examples of computer-readable mediums include memory devices, optical memory devices, integrated circuits, servers, online software, etc. Figure 6 shows an optical disc 600. Alternatively, the computer-readable medium 600 may have data 610 representing trained object detection models and / or guidance models for use with the techniques described herein.

[0088] Examples, embodiments, or optional features, whether non-limiting or not, should not be understood as limiting the claimed invention.

[0089] The embodiments described above are illustrative, not limiting, of the invention, and it should be noted that those skilled in the art can design many alternative embodiments without departing from the scope of the appended claims. No reference numeral placed between parentheses in a claim should be construed as limiting that claim. The use of the verb “have” and its conjugations does not exclude the existence of elements or stages other than those described in the claim. The absence of a statement that an element is multiple does not exclude the existence of multiple elements. The expression “at least one of” following a list or set of elements represents a selection of all or some of the elements from that list or set. For example, the expression “at least one of A, B, and C” should be understood as including only A, only B, only C, both A and B, both A and C, both B and C, or all of A, B, and C. The invention is implemented by hardware having several distinct elements and by a appropriately programmed computer. In a claim for an apparatus listing several means, some of these means may be embodied by one of the same items of hardware. The mere fact that certain means are described in mutually different dependent claims does not indicate that combinations of these means cannot be used advantageously.

Claims

1. An ultrasound fetal heart rate monitoring system, A sensor interface for acquiring sensor data from an ultrasound probe placed in the mother's abdomen. A data interface for accessing model data representing a trained object detection model, wherein the object detection model is configured to locate the fetal heart in an ultrasound image, and Processor subsystem The processor subsystem has, In order to acquire an ultrasound image from the ultrasound probe, the ultrasound probe is operated in imaging mode. The object detection model is applied to the ultrasound image to determine whether the fetal heart is located within the target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe, and If the fetal heart is located within the target region, the ultrasound probe is switched to Doppler mode, a Doppler ultrasound signal is acquired from the ultrasound probe, and the fetal heart rate is calculated from the Doppler ultrasound signal. It is configured in such a way, The processor subsystem is further configured to determine guidance information for guiding the ultrasound probe to the fetal heart from the location of one or more body parts of the fetus located by the object detection model, and to output the determined guidance information to a display. Ultrasound fetal heart rate monitoring system.

2. The system according to claim 1, wherein the system is configured to operate the ultrasound probe in either the imaging mode or the Doppler mode, but not in both modes simultaneously.

3. The system according to claim 1 or 2, wherein the processor subsystem is further configured to determine a signal quality index of the Doppler ultrasonic signal, and if the signal quality index does not meet a predetermined quality threshold, to return the ultrasonic probe to the imaging mode.

4. The system according to claim 3, wherein the processor subsystem is configured to output the signal quality index to the user.

5. The system according to claim 1, wherein the ultrasonic probe has one or more piezoelectric transducers and / or one or more capacitive ultrasonic transducers (CMUTs).

6. The system according to claim 1, wherein the processor subsystem is further configured to focus a Doppler ultrasonic beam emitted from the ultrasonic probe onto the target region.

7. The system according to claim 1, further comprising an output interface to a display, wherein the processor subsystem is configured to show the position of the fetal heart in the ultrasound image on the display.

8. The system according to claim 7, wherein the object detection model is further configured to locate one or more of the fetal head, the fetal spine, and the fetal femur, and the processor subsystem is configured to display the locates on the display.

9. The system according to claim 1, wherein the object detection model is configured to locate one or more of the fetal head, the fetal spine, and the fetal femur, and the processor subsystem is configured to determine the guidance information using the locateging.

10. The aforementioned processor subsystem is If the fetal spine is detected and the fetal head is not detected, guidance information is determined to guide the ultrasound probe to the expected position of the fetal head, and / or When the fetal spine and head are detected, guidance information is determined to guide the ultrasound probe to the expected position of the fetal heart. The system according to claim 9, configured as described above.

11. The system according to claim 1, wherein the ultrasound image is not displayed on the display.

12. The system according to claim 1, wherein the processor subsystem is further configured to acquire orientation data indicating the orientation of the ultrasound probe via the sensor interface, and to use the orientation data to determine the position of the fetal heart and / or the guidance information.

13. A computer-implemented ultrasound fetal heart rate monitoring method is: A step of acquiring sensor data from an ultrasound probe placed in the mother's abdomen. A step of accessing model data representing a trained object detection model, wherein the object detection model is configured to locate the fetal heart in an ultrasound image, To acquire an ultrasound image from the ultrasound probe, the ultrasound probe is operated in imaging mode. The steps include: applying the object detection model to the ultrasound image to determine whether the fetal heart is located within the target region of the ultrasound image that can be handled by the Doppler mode of the ultrasound probe; and If the fetal heart is located within the target region, the ultrasound probe is switched to Doppler mode, a Doppler ultrasound signal is acquired from the ultrasound probe, and the fetal heart rate is determined from the Doppler ultrasound signal. The method has, The steps include determining guidance information for guiding the ultrasound probe to the fetal heart based on the location of one or more body parts of the fetus identified by the object detection model, and outputting the determined guidance information to a display. A computer-implemented ultrasound fetal heart rate monitoring method further comprising the following.

14. A temporary or non-temporary computer-readable medium having data representing an instruction to cause the processor system to perform the computer-implemented ultrasound fetal heart rate monitoring method according to claim 13, when executed by the processor system.

Images