Device for monitoring fetal heart rate

Abstract

An apparatus for monitoring a fetal heart rate includes a memory including instruction data representing a set of instructions, and a processor in communication with the memory and configured to execute the set of instructions, which when executed by the processor causes the processor to: i) obtain ultrasound image data from an ultrasound transducer, ii) determine a location of the fetal heart within the ultrasound image data, iii) send a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, where the determined location of the fetal heart is used to set a target location for the Doppler mode, and iv) monitor the fetal heart rate using the ultrasound data obtained in the Doppler mode.

Description

[Technical field]

[0001] The present disclosure relates to ultrasound systems. In particular, but not by way of limitation, the present disclosure relates to ultrasound systems, devices, and methods for monitoring the heart rate of a fetus. [Background technology]

[0002] During labor and delivery, some of the most common obstetric procedures include fetal heart rate (FHR) monitoring and monitoring of maternal contractions. These measures allow healthcare professionals to assess the fetal status and determine if intervention is required. Optimal FHR monitoring requires a collaborative perinatal team to perform the monitoring correctly, interpret it appropriately, and communicate findings effectively and in a timely manner to all members of the care team when high-risk patterns are detected. Therefore, proper FHR acquisition and monitoring play a critical role in FHR measurement.

[0003] Fetal heart rate monitors generally include a transducer that is placed on the mother's abdomen and held in place with a strap or belt. The placement of the transducer is still generally performed using acoustic feedback resulting from Doppler shift. Correct and optimal placement of the transducer is essential, since other sources of Doppler shift may be present in the detection range of the ultrasound transducer. For example, blood flow in large maternal vessels is a major cause of erroneous measurements. The heart rate calculation also changes suddenly from fetal heart rate to maternal heart rate, depending on which signal is received stronger. The superposition of the two signals, which occurs quite frequently, also results in a gradual change in the heart rate value, which is not recognized at first glance. Furthermore, the ultrasound sensor may not be able to maintain focus on the fetal heart due to fetal or maternal movements.

[0004] Thus, FHR measurements in many current fetal monitors rely on the clinical expertise of a medical professional in placing and maintaining the fetal monitor transducer over the fetal heart. Therefore, a high level of training is typically required both to optimally place the transducer and to maintain it in the correct position (as the fetus / mother moves). Thus, in some situations, the training and understanding of the clinician (e.g., operating personnel) may not be sufficient to optimally place the transducer. Summary of the Invention [Problem to be solved by the invention]

[0005] As discussed above, ultrasound transducers for use in fetal monitoring during labor are difficult to position correctly to monitor the fetal heart rate. Furthermore, the ultrasound transducer requires constant monitoring to maintain correct positioning as the fetus and / or mother move. A frequent problem with the use of such monitors is that the heart rate obtained is from the mother and not the fetus. Such problems are further compounded in multiple pregnancies, where fetal movement can cause the fetal heart rate monitor to begin monitoring the wrong fetus over time. Furthermore, in the case of multiple fetuses, it is difficult to ensure that fetal heart rate measurements are consistently attributed to the correct fetus.

[0006] Furthermore, transducer placement using only Doppler shift as feedback requires real-time transmission of acoustics. For wireless systems, this limits the choice of wireless technology, as current protocols often have a delay of about 100 ms that is unacceptable when placing transducers on an acoustic feedback basis. It is an object of the present disclosure herein to provide an improved fetal monitoring system that addresses some of these issues. [Means for solving the problem]

[0007] According to a first aspect, an apparatus for monitoring a first fetal heartbeat is provided. The apparatus includes a memory including instruction data representing a set of instructions, and a processor in communication with the memory and configured to execute the set of instructions. The set of instructions, when executed by the processor, causes the processor to obtain ultrasound image data from an ultrasound transducer, determine a location of the first fetal heart in the ultrasound image data, send a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, where the determined location of the first fetal heart is used to set a target location for the Doppler mode, and monitor the first fetal heartbeat using the ultrasound data obtained in the Doppler mode.

[0008] According to a second aspect, there is provided an ultrasound system including the apparatus of the first aspect, and further including an ultrasound transducer.

[0009] According to a third aspect, there is a computer-implemented method of monitoring a first fetal heartbeat, the method comprising the steps of: i) obtaining ultrasound image data from an ultrasound transducer, ii) determining a location of the first fetal heart within the ultrasound image data, iii) sending a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, where the determined location is used to set a target location for the Doppler mode, and iv) monitoring the first fetal heartbeat using the ultrasound data obtained in the Doppler mode.

[0010] According to a fourth aspect there is provided a computer program product including a computer readable medium having computer readable code embodied therein, the computer readable code being configured, when executed by a suitable computer or processor, to cause the computer or processor to perform the method of the third aspect.

[0011] Thus, the devices, systems, and methods herein allow for "acoustic-free" placement of the ultrasound transducer on the fetus, whereby imaging modes are used to correctly place the transducer, rather than relying solely on acoustic feedback. "Acoustic-free" placement with an indicator of where to move the transducer is also advantageous in areas where multiple women are monitored in one room. Furthermore, the systems and methods herein allow the fetus to be monitored even when the fetus changes position. It also ensures that the fetal heart is being monitored (as opposed to, e.g., the mother's heart). In embodiments where more than one fetus is monitored (as described below), the methods and devices herein ensure that each fetus is monitored correctly, e.g., without confusion / overlap between the FHRs of different fetuses.

[0012] These and other aspects will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0013] Example embodiments will now be described, by way of example only, with reference to the following drawings, in which: [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 illustrates an apparatus according to some embodiments of the present disclosure. [Diagram 2] FIG. 1 illustrates an ultrasound system in accordance with certain embodiments of the present disclosure. [Diagram 3] FIG. 1 illustrates an example apparatus according to some embodiments of the present disclosure. [Figure 4] FIG. 1 illustrates an example output of a YOLO network in accordance with certain embodiments herein. [Diagram 5] FIG. 1 illustrates Doppler gating. [Figure 6] FIG. 2 illustrates an example process according to some embodiments of the present disclosure. [Figure 7] FIG. 1 illustrates another example process according to some embodiments herein. [Figure 8] FIG. 13 illustrates an example output of a YOLO network for an image containing two fetal hearts. [Figure 9a] FIG. 1 illustrates exemplary landmarks of the fetal heart. [Figure 9b] FIG. 1 illustrates exemplary landmarks of the fetal heart. [Figure 10] FIG. 1 illustrates an example method according to some embodiments of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] As mentioned above, fetal heart rate monitors that use acoustic feedback resulting from Doppler shifts to detect and monitor fetal heart rate are difficult to place correctly and difficult to maintain in the correct position when the fetus and / or mother move. These problems are compounded in pregnancies involving multiple fetuses, where there is more than one fetus. It is an objective of the embodiments herein to improve this situation and provide improved fetal heart rate monitoring during labor.

[0016] 1, an apparatus 100 for monitoring a first fetal heartbeat is shown according to some embodiments of the present disclosure. Typically, the apparatus forms part of a computing device or system, such as, for example, a laptop, desktop computer, or other computing device. In some embodiments, the apparatus 100 forms part of a distributed computing arrangement or cloud.

[0017] In general, the apparatus is adapted or configured to perform any of the method embodiments described herein, such as method 1000 described below with respect to FIG.

[0018] The apparatus includes a memory 104 including instruction data representing a set of instructions 106, and a processor 102 (e.g., processing circuitry or logic) in communication with the memory and configured to execute the set of instructions 106. Briefly, the set of instructions 106, when executed on the processor 102, causes the processor 102 to: i) obtain ultrasound image data from an ultrasound transducer, ii) determine a location of a first fetal heart within the ultrasound image data, iii) send a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, where the determined location of the first fetal heart is used to set a target location for the Doppler mode, and iv) monitor the first fetal heartbeat using the ultrasound data obtained in the Doppler mode.

[0019] As mentioned above, a major problem with current ultrasound transducers used in obstetric fetal monitoring is that it is difficult to correctly position and maintain the ultrasound transducer in the correct position to monitor the fetal heart. Fetal or maternal movements, ambient noise, maternal heartbeat, and other sounds, e.g., from blood moving through the maternal veins and arteries, all result in inaccurate positioning and monitoring of the fetal heart. The device proposed herein uses the imaging mode of the transducer to determine the location of the fetal heart. The determined location is then used to set a target location in the Doppler mode (e.g., used to set the Doppler gate), thereby monitoring the fetal heartbeat. This ensures that it is the fetal heart that is monitored, as opposed to, e.g., any of the other aforementioned movements. This therefore provides a more robust fetal monitoring.

[0020] More specifically, the processor 102 includes one or more processors, processing units, multi-core processors, or modules configured or programmed to control the device 100 in the manner described herein. In certain implementations, the processor 102 includes multiple software and / or hardware modules, each configured or intended to perform an individual step or multiple steps of the methods described herein. The processor 102 includes one or more processors, processing units, multi-core processors, and / or modules configured or programmed to control the device 100 in the manner described herein. In some implementations, for example, the processor 102 includes multiple (e.g., used simultaneously) processors, processing units, multi-core processors, and / or modules configured for distributed processing. Those skilled in the art will recognize that such processors, processing units, multi-core processors, and / or modules may be located in different locations and perform different steps of the methods described herein and / or different portions of a single step.

[0021] The memory 104 is configured to store program code executed by the processor 102 to perform the methods described herein. Alternatively or in addition, the one or more memories 104 may be external to (i.e., separate or remote from) the apparatus 100. For example, the one or more memories 104 may be part of another device. The memory 104 is used to store the acquired ultrasound data and / or any other information or data received, calculated, or determined by the processor 102 of the apparatus 100 or from any interface, memory, or device external to the apparatus 100. The processor 102 is configured to control the memory 104 to store the acquired ultrasound data and / or any other information or data received, calculated, or determined by the processor.

[0022] In some embodiments, memory 104 includes multiple sub-memories, each capable of storing instruction data, such as at least one sub-memory storing instruction data representing at least one instruction of the set of instructions, while at least one other sub-memory stores instruction data representing at least one other instruction of the set of instructions.

[0023] It will be appreciated that Fig. 1 shows only the components required to illustrate this aspect of the disclosure, and that in an actual implementation, device 100 may include additional components in addition to those shown. For example, device 100 further includes a display. The display includes, for example, a computer screen and / or a mobile phone or tablet screen. The device further includes a user input device, such as a keyboard, mouse, or other input device, that allows a user to interact with the device, for example, to provide initial input parameters used in the methods described herein. Device 100 includes a battery or other power source for powering device 100, or a means for connecting device 100 to a mains power source.

[0024] In some embodiments, the device 100 is included in an ultrasound system, for example an ultrasound system that can include the device 100 described above and further include a transducer that can be used to record a sequence of ultrasound image data.

[0025] In some embodiments, the ultrasound system further includes a display. The processor is adapted to send instructions to the display to cause the display to display the ultrasound image data and / or the ultrasound data acquired in Doppler mode. The processor is further adapted to send instructions to the display to cause the display to provide an indication of whether the ultrasound transducer is operating in imaging mode or Doppler mode.

[0026] In some embodiments described below, the ultrasound system includes an array of transducers arranged in a wearable "patch." The patch includes drive electronics for individually actuating each transducer in the array. The transducers are further fitted with sensors (e.g., inertial measurement unit IMU sensors, etc.) for determining the orientation of each transducer.

[0027] 2 illustrates an example ultrasound system 200. Ultrasound system 200 includes apparatus 100 as described above. In other embodiments, components of ultrasound system 200 are adapted to perform method 1000 as described below.

[0028] The system 200 includes an array transducer probe 4 having a transducer array 6 for transmitting ultrasound waves and receiving echo information. The transducer array 6 includes a CMUT transducer, a piezoelectric transducer formed from a material such as PZT or PVDF, or any other suitable transducer technology. In this example, the transducer array 6 is a two-dimensional array of transducers 8 capable of scanning either a 2D plane or a three-dimensional volume of an area of ​​interest. In another example, the transducer array is a 1D array.

[0029] The transducer array 6 is coupled to a microbeamformer 12 that controls reception of signals by the transducer elements. The microbeamformer can at least partially beamform signals received by sub-arrays, commonly referred to as "groups" or "patches," of the transducers, as described in U.S. Patent No. 5,997,479 (Savord et al.), U.S. Patent No. 6,013,032 (Savord), and U.S. Patent No. 6,623,432 (Powers et al.).

[0030] In an alternative embodiment, instead of the microbeamformer 12, the transducer array is directly manipulated by a main system beamformer (not shown in FIG. 2).

[0031] The system 200 further includes a transmit / receive (T / R) switch 16, which couples to the microbeamformer 12 and switches the array between transmit and receive modes.

[0032] Transmission of ultrasound beams from the transducer array 6 is directed by a transducer controller 18, which is coupled to the microbeamformer by the T / R switch 16, and a main transmit beamformer (not shown), which receives input from user operation of a user interface or control panel 38. The controller 18 includes transmit circuitry configured to drive the transducer elements of the array 6 (directly or via the microbeamformer) during transmit mode.

[0033] It should be noted that in an alternative embodiment in which the transducer array is directly operated by the main system beamformer instead of the microbeamformer 12, the T / R switch 16 protects the main beamformer 20 from high energy transmit signals.

[0034] In a typical line-by-line imaging sequence, the beamforming system in the probe operates as follows: During transmit, the beamformer (either a microbeamformer or a main system beamformer depending on the implementation) activates a transducer array or a subaperture of a transducer array. A subaperture is a one-dimensional line of transducers or a two-dimensional patch of transducers within a larger array. In transmit mode, the focusing and steering of the ultrasound beam generated by the array or a subaperture of the array is controlled as described below.

[0035] Upon receiving backscattered echo signals from the subject, the received signals undergo receive beamforming (as described below) to align the received signals, and if a subaperture is used, the subaperture is then shifted, for example by one transducer element, The shifted subaperture is then activated, and this process is repeated until all of the transducer elements of the transducer array have been activated.

[0036] For each line (or subaperture), the total receive signal used to form the associated line of the final ultrasound image is the sum of the voltage signals measured by the transducer elements of the given subaperture during the receive period. Following the beamforming process, the resulting line signals are commonly referred to as radio frequency (RF) data. Each line signal (RF data set) generated by the various subapertures then undergoes additional processing to generate a line of the final ultrasound image. The change in amplitude of the line signal over time contributes to the change in brightness of the ultrasound image with depth, with high amplitude peaks corresponding to bright pixels (or groups of pixels) in the final image. Peaks appearing near the beginning of the line signal represent echoes from shallow structures, while peaks appearing later and later in the line signal represent echoes from structures at increasing depths within the subject.

[0037] One of the functions controlled by the transducer controller 18 is the direction in which the beam is steered and focused. The beam is steered straight ahead from the transducer array (perpendicular to the transducer array) or at a different angle for a wider field of view. The steering and focusing of the transmit beam is controlled as a function of the transducer element activation times.

[0038] In general ultrasound data acquisition, two methods are distinguished: plane wave imaging and "beam steering" imaging. The two methods are distinguished by the presence of beamforming in the transmit mode ("beam steering" imaging) and / or in the receive mode (plane wave imaging and "beam steering" imaging).

[0039] Turning first to the focusing function, by activating all of the transducer elements simultaneously, the transducer array creates a plane wave that diverges as it travels through the subject. In this case, the beam of ultrasound remains unfocused. By introducing position-dependent time delays in the activation of the transducers, it is possible to focus the wavefront of the beam to a desired point, called a focal zone. A focal zone is defined as a point where the lateral beam width is less than half the transmit beam width. In this way, the lateral resolution of the final ultrasound image is improved.

[0040] For example, if the transducer elements are activated in succession starting from the outermost elements and ending with the central element of the transducer array with a time delay, a focal zone is formed at a given distance from the probe in line with the central element. The distance of the focal zone from the probe varies according to the time delay between each subsequent repetition of the transducer element activation. After the beam passes through the focal zone, it starts to diverge, forming a far-field imaging region. It should be noted that in focal zones located very close to the transducer array, the ultrasound beam diverges rapidly in the far field, which causes beam width artifacts in the final image. In general, the near field located between the transducer array and the focal zone shows little detail due to the large overlap of the ultrasound beams. Therefore, changing the location of the focal zone will change the quality of the final image quite significantly.

[0041] It should be noted that in transmit mode, only one focal point is defined, unless the ultrasound image is divided into multiple focal zones (each of which has a different transmit focal point).

[0042] In addition, when receiving echo signals from inside the subject, it is possible to perform the reverse of the above process to perform receive focusing. In other words, the input signal is received by the transducer elements, subjected to an electronic delay time, and then passed to the system for signal processing. The simplest example of this is called delay-and-sum beamforming. It is possible to dynamically adjust the receive focusing of the transducer array as a function of time.

[0043] Turning now to the function of beam steering, by properly applying time delays to the transducer elements, it is possible to impart a desired angle to the ultrasound beam as it exits the transducer array. For example, by activating the transducers on a first side of the transducer array, followed by the remaining transducers in a sequence ending on the opposite side of the array, the wavefront of the beam is angled toward the second side. The magnitude of the steering angle relative to the normal of the transducer array depends on the magnitude of the delay time between the activation of subsequent transducer elements.

[0044] Furthermore, the steering beam can be focused, and the total time delay applied to each transducer element is the sum of both the focusing delay time and the steering delay time, in which case the transducer array is called a phased array.

[0045] In the case of CMUT transducers that require a DC bias voltage for operation, the transducer controller 18 is coupled to control a DC bias control 45 for the transducer array. The DC bias control 45 sets the DC bias voltage applied to the CMUT transducer elements.

[0046] For each transducer element of the transducer array, an analog ultrasound signal, commonly called channel data, enters the system through a receive channel. In the receive channel, a partially beamformed signal is created by the microbeamformer 12 from the channel data and then passed to the main receive beamformer 20, which combines the partially beamformed signals from the individual patches of the transducer into a fully beamformed signal, called radio frequency (RF) data. The beamforming performed at each stage may be as described above or may include additional functions. For example, the main beamformer 20 has 128 channels, each of which receives partially beamformed signals from patches of tens or hundreds of transducer elements. In this way, the signals received by the thousands of transducers of the transducer array efficiently contribute to a single beamformed signal.

[0047] The beamformed received signals are coupled to a signal processor 22, which processes the received echo signals in various ways, such as bandpass filtering, decimation, separation of I and Q components, and harmonic signal separation, which acts to separate linear and nonlinear signals to allow identification of nonlinear (harmonic of fundamental frequency) echo signals returning from tissue and microbubbles. The signal processor also performs additional signal enhancements such as speckle reduction, signal combining, and noise removal. The bandpass filters in the signal processor may be tracking filters whose passbands slide from higher to lower frequency bands as echo signals are received from increasing depths, thereby blocking higher frequency noise from greater depths where anatomical information is typically lacking.

[0048] The transmit and receive beamformers can be implemented with different hardware and have different functions. Naturally, the receiver beamformer is designed to take into account the characteristics of the transmit beamformer. In Fig. 2, for simplicity, only the receiver beamformers 12, 20 are shown. In a complete system, there is also a transmit chain with a transmit microbeamformer and a main transmit beamformer.

[0049] The function of the microbeamformer 12 is to perform an initial combination of signals to reduce the number of analog signal paths, which is typically performed in the analog domain.

[0050] Final beamforming occurs in the main beamformer 20, typically after digitization.

[0051] The transmit and receive channels use the same transducer array 6 with a fixed frequency band. However, the bandwidth occupied by the transmit pulses may vary depending on the transmit beamforming used. The receive channels either capture the entire transducer bandwidth (which is the conventional approach) or use bandpass processing to extract only the bandwidth containing the desired information (e.g., harmonics of the main harmonic).

[0052] The RF signals are then coupled to a B-mode (i.e., intensity mode or 2D imaging mode) processor 26 and a Doppler processor 28. The B-mode processor 26 performs amplitude detection on the received ultrasound signals to image structures within the body, such as organ tissues and blood vessels. In the case of line-by-line imaging, each line (beam) is represented by an associated RF signal, whose amplitude is used to generate an intensity value to be assigned to a pixel of the B-mode image. The exact location of a pixel in the image is determined by the location of the associated amplitude measurement along the RF signal and the line (beam) number of the RF signal. B-mode images of such structures are formed in harmonic or fundamental imaging modes, or a combination of both, as described in U.S. Pat. No. 6,283,919 (Roundhill et al.) and U.S. Pat. No. 6,458,083 (Jago et al.). The Doppler processor 28 processes the temporally distinct signals resulting from tissue motion and blood flow to detect material motion, such as the flow of blood cells, within the image field. The Doppler processor 28 typically includes a wall filter with parameters set to pass or block echoes returning from selected types of material within the body.

[0053] The structural and motion signals produced by the B-mode and Doppler processors are coupled to the scan converter 32 and the multiplanar reformatter 44. The scan converter 32 arranges the echo signals in the spatial relationship in which they were received and in the desired image format. In other words, the scan converter functions to convert the RF data from a cylindrical coordinate system to a Cartesian coordinate system that is suitable for displaying ultrasound images on the image display 40. For B-mode imaging, the brightness of a pixel at a given coordinate is proportional to the amplitude of the RF signal received from that location. For example, the scan converter arranges the echo signals in a two-dimensional (2D) sector format or a pyramidal three-dimensional (3D) image. The scan converter overlays the B-mode structural image with colors that correspond to the motion at the point in the image field. The Doppler estimated velocity produces a given color. The combined B-mode structural image and color Doppler image show the tissue motion and blood flow within the structural image field. The multiplanar reformatter converts echoes received from points in a common plane within a volumetric region of the body into an ultrasound image of that plane, as described in U.S. Patent No. 6,443,896 (Detmer). The volume renderer 42 converts the echo signals of the 3D data set into a projected 3D image viewed from a given reference point, as described in U.S. Patent No. 6,530,885 (Entrekin et al.).

[0054] The 2D or 3D images are coupled from the scan converter 32, the multiplanar reformatter 44, and the volume renderer 42 to the image processor 30 for further enhancement, buffering, and temporary storage for display on the image display 40. The imaging processor is adapted to remove certain imaging artifacts from the final ultrasound image, such as, for example, acoustic shadowing caused by, for example, strong attenuators or refraction, back enhancement caused by, for example, weak attenuators, reverberation artifacts when, for example, highly reflective tissue interfaces are located nearby, etc. In addition, the image processor is adapted to handle certain speckle reduction functions to improve the contrast of the final ultrasound image.

[0055] In addition to being used for imaging, the blood flow values ​​produced by the Doppler processor 28 and the tissue structure information produced by the B-mode processor 26 are coupled to a quantification processor 34. The quantification processor is used to make measurements within the images. The quantification processor receives input from a user control panel 38, as described in more detail below.

[0056] Output data from the quantification processor is coupled to a graphics processor 36 for reproducing measurement graphics and values ​​by images on a display 40 and for audio output from the display device 40. The graphics processor 36 also generates graphic overlays for display with the ultrasound images. These graphic overlays include standard identification information such as the patient name, the date and time of the image, imaging parameters, etc. For these purposes, the graphics processor receives inputs such as the patient name from a user interface 38. The graphic overlays further include information such as whether the transducer is monitoring in Doppler mode or imaging mode. The user interface is also coupled to a transmit controller 18 for controlling the generation of ultrasound signals from the transducer array 6 and thus the images produced by the transducer array and the ultrasound system. The transmit control function of the controller 18 is only one of the functions performed. The controller 18 also takes into account the operating mode (given by the user) and the corresponding required transmitter configuration and bandpass configuration in the receiver analog-to-digital converter. The controller 18 can be a state machine with fixed states.

[0057] The user interface is also coupled to a multiplanar reformatter 44 for selection and control of multiple MPR image planes that are used to perform quantification procedures in the image field of the MPR image.

[0058] Those skilled in the art will recognize that the details provided above and the components shown in FIG. 2 are merely examples, and that an ultrasound system may have different components than those shown in FIG.

[0059] Returning now to FIG. 1 and referring to the functionality of the processor 102, generally, one or more transducers in imaging mode are used to localize the fetus (e.g., by moving the probe over the surface of the mother's stomach). Once the fetus is located, the processor commands the probe to switch to Doppler mode for subsequent monitoring of the localized fetal heart rate. In this way, it is ensured that it is the fetal heart that is monitored, as opposed to, for example, the maternal heart or other components of the maternal vascular system. In this way, this facilitates improved placement of the fetal heart monitor.

[0060] More specifically, as described above, the processor 102 is arranged to obtain, in step i), ultrasound image data from an ultrasound transducer.

[0061] Typically, the method is performed in real time, e.g. as part of a live ultrasound examination, e.g. for obstetric monitoring during labour. For example, the processor is adapted to obtain ultrasound image data from a probe, e.g. in an imaging mode operation or a B-mode operation. The ultrasound image data is obtained from a probe placed on the abdomen of the fetus's mother to image the mother's uterus. Thus, in step i), the obtained ultrasound image data includes ultrasound image data of the uterus.

[0062] In some embodiments, a single ultrasound transducer is used, for example, the ultrasound transducer forms part of a probe. Such a probe is moved (e.g., by a clinician) over the mother's abdomen during imaging mode, and once the location of the heart has been located in the image data, it is switched to Doppler mode. Once in Doppler mode, the probe is held in place, for example, using a conventional strap or belt.

[0063] In other embodiments, the ultrasonic transducer forms part of (e.g. is one of) an array of transducers, such as an array of capacitive micromachined ultrasonic transducers CMUT, the array of transducers being arranged, for example, as a wearable ultrasonic patch.

[0064] The transducers in such an array include, for example, capacitive micromachined ultrasound transducers (CMUTs) arranged in a 2D matrix. The CMUT sensors can share drive electronics. In this way, instead of trying to locate the fetal heart by moving a small handheld ultrasound probe over the area of ​​the mother's abdomen, a large field of view ultrasound system is formed to image a large volume of the mother's abdomen. A large number of CMUT sensors are mounted on flexible interconnects and multiplexed to the drive electronics by utilizing a DC bias electrode on each device. Each CMUT array is actuated by a DC voltage (also called a bias voltage). As soon as the DC voltage is applied to a single CMUT, the single CMUT transmits and receives ultrasound. By decoupling the DC bias voltage from the DC bias electrode, the ultrasound sensor remains floating and therefore does not affect the operation of the selected sensor of the set. The drive electronics are for driving and reading out the selected set of CMUT cells.

[0065] Each ultrasound transducer is further associated with a sensor for determining an orientation of the respective ultrasound transducer, e.g., a position sensor. In some embodiments, each transducer is associated with (e.g., attached to) a respective inertial measurement unit (IMU) sensor. In such embodiments, the processor is further adapted to obtain data from a plurality of IMU sensors, each IMU attached to a respective transducer in the array of transducers. The processor is then adapted to determine an orientation of each respective transducer using data from the IMU sensors. The processor, for example, further determines an absolute position (e.g., fetal presentation) of the first fetus from data from the IMU sensors, as described in more detail below.

[0066] An exemplary transducer array 300 showing an array of transducers 302 according to an embodiment herein is shown in FIG. 3. In this embodiment, n×n 2D CMUT sensors are placed on a flex PCB 302 (20 mm apart). In this embodiment, a 9-axis IMU sensor 306 is placed above every CMUT sensor. The IMU sensor provides absolute orientation of the CMUT sensors on the flex. A CMUT control system 308 controls when each CMUT in the array is activated (e.g., the CMUT control system 308 activates each CMUT in sequence, as described below). In this embodiment, the CMUT array flex is connected to an image acquisition system 310 consisting of a microbeamformer supporting B-mode imaging and continuous wave Doppler mode. Raw RF and IMU data are sent from the transducers 302 in the array to the image acquisition system 310. The data is analyzed by a host system 312 (such as device 100), which analyzes the acquired ultrasound images from the image acquisition system to perform steps ii)-iv) as described below.

[0067] In an embodiment in which an ultrasound image is acquired from an array of ultrasound transducers (e.g., the CMUT sensor of FIG. 3) in step i), a sequence of ultrasound images is obtained from the ultrasound transducers by activating the array of ultrasound transducers in sequence one after the other.

[0068] Referring back to the processor of FIG. 1, after obtaining the ultrasound image data from the ultrasound transducer, the processor performs step ii) to determine the location of the first fetal heart within the ultrasound image data.

[0069] In some embodiments, step ii) is performed using machine learning. For example, the processor is adapted to determine the location of the first fetal heart in the ultrasound image data using a first machine learning model trained to identify the location of the fetal heart in the ultrasound image. For example, the heart detection is performed using deep learning techniques. For example, a neural network such as a convolutional neural network is used. Those skilled in the art will be familiar with machine learning models and how to train them to determine the location of objects in an image (e.g., via backpropagation and gradient descent). Those skilled in the art will recognize that these are merely examples and that other types of models may also be trained to detect the fetal heart in an ultrasound image.

[0070] As an example, You Only Look Once (YOLO) based heart detection is used to locate the heart. YOLO is described in the paper entitled "Yolov3: An incremental improvement" arXiv preprint arXiv:1804.02767, 2018 by J. Redmon and A. Farhadi. YOLO is a region proposal network that applies a single neural network to the entire image to detect specific objects. The network divides the image into regions and predicts a bounding box and probability 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 the detection of that object. Each grid cell predicts bounding boxes and confidence scores for those boxes. These scores reflect the reliability of the object model contained in the box and the accuracy of the box containing the predicted object. Each bounding box consists of five predictions: x, y, w, h, and confidence. The (x,y) coordinates represent the center of the box relative to the bounds of the grid cell. The width (w) and height (h) are predicted relative to the entire image.

[0071] In an embodiment where a sequence of ultrasound images is acquired from an array of ultrasound sensors, each ultrasound image in the sequence is analyzed in turn to determine whether the image contains an image of the fetal heart. For example, the orientation of the ultrasound (e.g., CMUT) sensor is detected relative to the orientation of the mother using an orientation sensor (such as the IMU sensor described above) integrated into the ultrasound transducer. Automatic detection of the fundus end of the uterus versus the cervical end of the uterus is performed using deep learning techniques. Deep learning techniques are also used to detect the fetal spine and fetal head. This allows detection of the fetal position. When a sequence of images is acquired, one ultrasound transducer in the array is activated in sequence, one at a time, across the grid. This is repeated until the heart is detected in one of the images (e.g., from one of the transducers).

[0072] The confidence level of the YOLO output is a regression that is trained to output the intersection of the union (IOU) between the output bounding box and the ground truth bounding box. An example output is shown in Figure 4, which shows an ultrasound image 400 of a fetal heart 402 and a bounding box 404 on the fetal heart 402 output by the YOLO model, the image originating from a single CMUT sensor.

[0073] In general, the YOLO model is used to determine the location of the fetal heart in an ultrasound image, e.g., to obtain the depth of the heart. In embodiments where an array of transducers is used, the YOLO model is used to determine which transducer in the array of transducers has within its field of view the fetal heart.

[0074] For example, the output of step ii) typically includes an indication of the depth of the fetal heart, the scan line number, and / or an indication of the ultrasound transducer at which the fetal heart was detected (e.g., in instances where an array of transducers is used).

[0075] 1, in step iii), the processor 102 is then caused to send a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, and the determined location (e.g., depth and / or scan line) of the first fetal heart is used to set a target location for the Doppler mode. For example, the location (e.g., depth and / or scan line) is used to set a Doppler gate for the Doppler mode.

[0076] In an embodiment in which YOLO is used, once the fetal heart is detected, the depth and angle 504 at which m-mode must be enabled is determined and passed to the imaging system.

[0077] The output of the YOLO model is used to set the location of the Doppler mode, e.g., the center of the bounding box predicted by the YOLO model is used to determine the point through which the scanline passes. The angle of the scanline is also determined. Figure 5 shows a calculated scanline 502 from the depth information, which is determined in a known manner. This information is then used to set the target location of the Doppler mode (e.g., to set the Doppler gate).

[0078] In step iv), the processor then monitors the first fetal heartbeat using the ultrasound data acquired in Doppler mode, which is performed in a known manner.

[0079] In the above-described embodiment in which ultrasound image data from each transducer in the array of transducers is obtained in step i), the processor selects a first ultrasound transducer from the array of transducers having a first view of the fetal heart and performs steps ii), iii), and iv) on the first ultrasound transducer and the ultrasound image data obtained using the first ultrasound transducer.

[0080] In other words, the processor i) obtains ultrasound image data (e.g., a sequence of ultrasound images as described above) from an array of ultrasound transducers and selects a first ultrasound transducer from the array of transducers having a view of the first fetal heart; ii) determines a location of the first fetal heart in the ultrasound image data from the first transducer; iii) sends a message to the first ultrasound transducer to instruct the first ultrasound transducer to switch to a Doppler mode, where the determined location of the first fetal heart is used to set a target location for the Doppler mode; and iv) monitors the heartbeat of the first fetal heart using the ultrasound data obtained in the Doppler mode from the first transducer.

[0081] For example, the processor switches the transducer in the array having an image of the heart to Doppler mode to perform fetal heart monitoring. In this manner, the array is used in imaging mode to determine the location of the fetal heart, and then the transducer having the best view of the heart is switched to Doppler mode for fetal monitoring. Thus, the sonographer / clinician does not need to manually determine the optimal position of the ultrasound transducer probe, but rather the optimally positioned transducer is selected in an automated fashion.

[0082] As mentioned above, fetal and / or maternal movement may mean that the placement of the ultrasound transducers may need to be updated over time, or in the case of an array of transducers, a different transducer in the array (e.g., a different or second transducer) may be better positioned to optimally monitor the fetal heart.

[0083] Therefore, in some embodiments, the processor further repeats steps i)-iv) to relocate the fetal heart. For example, steps i)-iv) are performed / repeated at periodic time intervals. This ensures that the location of the fetal heart is monitored periodically. In other words, the processor causes the transducer to switch between imaging mode + Doppler mode intermittently (e.g., every "n" minutes) to ensure that the fetal heart (or hearts in case of multiples, as described below) is visible and audible.

[0084] In another example, steps i)-iv) are alternatively or additionally repeated if the audibility of the first fetal heartbeat falls below a threshold audibility in the ultrasound data acquired in Doppler mode. In other words, if the quality of the heartbeat signal starts to deteriorate or deteriorates below a threshold, this triggers the device 100 to relocate the heart.

[0085] In other examples, steps i) to iv) are alternatively or additionally repeated if the signal-to-noise magnitude of the heartbeat falls below a threshold signal-to-noise level and / or if the monitored heartbeat is interrupted (or becomes ambiguous).

[0086] Upon switching to imaging+Doppler mode, the device relocates the fetal heart (e.g., using deep learning algorithms as described above) or reconfirms the position and depth of the fetal heart, and then switches to Doppler mode accordingly, thus ensuring that the heart is properly monitored throughout.

[0087] This process is illustrated in FIG. 6, which shows an example process 600 according to some embodiments herein. In this embodiment, there is an array of transducers mounted on a CMUT patch 602. The patch has driving electronics in the form of a CMUT selection PCB 604, and each transducer in the array has an IMU sensor 606 attached to it. The patch provides Euler angles and image data to the device 100. The device 100 obtains / receives ultrasound image data from each transducer in the array of transducers (e.g., step i above) (and may perform a check 608 that image data has been obtained) and analyzes the image data from each transducer to determine the location of the first fetal heart in the ultrasound image data. In this embodiment shown in FIG. 6, a YOLO network 610 is used to identify and determine the location of the heart (as described above). A pose estimate 612 is used to determine the orientation of the heart and / or to identify individual hearts in the case of multiple fetuses (this is described in more detail below with respect to FIGS. 7-9). The device 100 then selects a first ultrasound transducer from the array of transducers that has a view of the first fetal heart (as determined by YOLO). The device then sends a message 614 to the selected first ultrasound transducer 604 to instruct the selected first ultrasound transducer to switch to Doppler mode (e.g., activate the Doppler mode). As described above, the determined location of the first fetal heart is used to set a target location for the Doppler mode (e.g., set a Doppler gate for the Doppler mode). The location of the fetal heart is determined at least in part using data from the IMU sensor 606 on the transducer 602. In step 616, the first fetal heartbeat is then monitored 616 using ultrasound data obtained from the selected first transducer in the Doppler mode. The device monitors 618 the quality (e.g., signal to noise) of the Doppler signal, and if the signal falls below a threshold quality, instructions are sent to the transducer array to instruct the transducers to switch to imaging mode and acquire ultrasound image data.The process is then repeated. In this manner, imaging and Doppler modes are used to efficiently and reliably monitor the fetal heart without the need for input from the clinician.

[0088] As noted above, pregnancies involving multiple fetuses (e.g., twins, triplets, etc.) pose additional challenges as the heartbeat of each fetus needs to be monitored separately. The fetal heartbeats need to be separated and reliably monitored, for example, to attribute each signal to the correct fetus and without confusing one heartbeat with another. Fetal movement over time further complicates this.

[0089] In the case of a single transducer with a wide field of view, a single transducer is used to image the second (and subsequent) fetal hearts, e.g., steps ii)-iv) are performed on two or more hearts in the same image.

[0090] In the case of multiple pregnancies, the array of transducers as described above allows for continuous monitoring of each fetal heart. For example, the processor further selects a second ultrasound transducer from the array of transducers having a view of a second fetal heart and performs steps ii), iii), and iv) on the second ultrasound transducer and the ultrasound image data from the second ultrasound transducer of the second fetal heartbeat.

[0091] In other words, the processor is configured to: i) obtain ultrasound image data from an array of ultrasound transducers, select a first ultrasound transducer from the array of transducers having a view of a first fetal heart, and select a second ultrasound transducer from the array of transducers having a view of a second fetal heart; ii) determine a location of the first fetal heart in the ultrasound image data from the first transducer, and determine a location of the second fetal heart in the ultrasound image data from the second transducer; and iii) send a message to the first ultrasound transducer to instruct the first ultrasound transducer to switch to a Doppler mode, whereby the first fetal heart is viewed from the first fetal heart. sending a message to the second ultrasound transducer, where the determined location of the fetal heart is used to set a target location of a Doppler mode of the first ultrasound transducer, and sending a message to the second ultrasound transducer to instruct the second ultrasound transducer to switch to Doppler mode, where the determined location of the second fetal heart is used to set a target location of the second ultrasound transducer Doppler mode; and iv) monitoring the first fetal heartbeat using ultrasound data obtained from the first transducer in Doppler mode and monitoring the second fetal heartbeat using ultrasound data obtained from the second transducer in Doppler mode.

[0092] One of skill in the art will recognize that this process may be further generalized to a third fetus and / or subsequent fetuses.

[0093] One of the problems with current FHR monitors when monitoring multiple fetuses is that they cannot (easily) distinguish between multiple hearts, i.e., which FHR comes from which fetus. This is further complicated because twins tend to align their FHR. In an embodiment herein, landmark identification is proposed as a way to track each fetus to ensure that each heartbeat is consistently attributed to the correct fetus. For example, each fetal heart is slightly different, and therefore, based on the relative position of a set of landmarks for each fetal heart, the fetal heart of a first fetus is differentiated from the fetal heart of a second fetus in a twin pregnancy (also applicable to multiple fetuses). In this way, key salient points are used, so that the fetus is always tagged, even after transducer repositioning and / or fetal movement.

[0094] Thus, in some embodiments, the processor 102 is further adapted to determine a first plurality of landmarks of the first fetal heart in the ultrasound image data of the first fetus and a second plurality of landmarks of the second fetal heart in the ultrasound image data of the second fetus. The processor is then adapted to repeat steps i)-iv) on new ultrasound image data using the first plurality of landmarks for identifying the first fetus and the second plurality of landmarks for identifying the second fetus for continuous monitoring of the first fetal heartbeat and the second fetal heartbeat.

[0095] 7 illustrates a method 700 for identifying, tagging, and monitoring two (or more) fetal hearts using imaging and sensor information according to an exemplary embodiment of the present specification. In this embodiment, the fetal heartbeats are monitored by combining information from YOLO 704 (a real-time object detector) followed by information from a real-time deep learning based so-called pose estimator and information from multiple IMU (inertial measurement unit) sensors. The pose estimator is described, for example, in the paper entitled "Deep high-resolution representation learning for human pose estimation" by Ke Sun, Bin Xiao, Dong Liu, and Jingdong Wang, CoRR, abs / 1902.09212, 2019.

[0096] In this embodiment, YOLO 704 is performed on an ultrasound image 702 obtained from an array of transducers (e.g., according to step i) above). YOLO performs step ii) above (e.g., YOLO is used to determine the location of the first and second fetal hearts in the ultrasound image data). FIG. 8 shows an example ultrasound image 800 and example bounding boxes 802, 804 output by an example YOLO. With respect to FIG. 7, any bounding boxes 704 that contain detections above a certain confidence threshold are then cropped 706 and sent independently to a pose estimator 708 to estimate the pose of each fetal heart independently. The pose estimator 708 is configured to determine a first plurality of landmarks of the first fetal heart in the ultrasound image data of the first fetus and a second plurality of landmarks of the second fetal heart in the ultrasound image data of the second fetus.

[0097] In this example, the posture identification is performed using a deep neural network that detects (e.g., regresses) landmarks of the fetal heart. An example is shown in FIG. 9a, which shows a number of landmarks 904 in an image of a heart 900. In general, the landmarks include points in the cardiac anatomy and / or points on the abdominal anatomy, such as the spine. Examples of landmarks include, but are not limited to, the spine, Croox, tricuspid junction, apex, mitral junction, central LV, central RV, central LA, central RA, aorta, tricuspid v-midpoint, and mitral v-midpoint. The posture identifier is used to track several landmarks of the first (and / or second) fetal heart in real time, as shown in FIG. 9b. The relative orientation of the landmarks (e.g., relative to each other) is used to determine the orientation of the heart. As an example, the spine-cardiac crux-apex angle is used to determine the cardiac axis 906, as shown in FIG. 9b. The cardiac axis determines the absolute orientation of the fetus. Crookes' position (where the valves of the four chambers of the heart join) is used as the location of the center of the heart in ultrasound scans.

[0098] Sensor information is used in addition to image information. For example, in the embodiment of FIG. 7, IMU sensors 710 (as described above with respect to FIG. 3) attached to transducers in an array of transducers are used to determine the position and orientation of each respective transducer's fan beam shape 712. In this manner, the orientation and position 714 of each heart relative to the world coordinate system is determined. A message is sent to the transducers in view of each fetal heart to instruct each ultrasound transducer to switch to Doppler mode (using the determined location of each fetal heart to set a target location for the Doppler mode, e.g., by setting a Doppler gate for the Doppler mode), and thus the heartbeat of each heart is monitored (716) as described above (for steps iii and iv).

[0099] By using Doppler gating, two or more transducers can be used simultaneously in Doppler mode, thereby allowing the FHR of two or more fetuses to be monitored simultaneously.

[0100] As the fetus moves, the position and orientation of each heart is tracked by repeating the steps outlined above based on the previous position and orientation and using a nearest neighbor linear approximation.

[0101] Furthermore, subsequent acquisitions are performed with a limited FOV since it is known that the baby's position is primarily a gentle movement (unlike one or more limb movements in early pregnancy). If at any point the heart is not detected within the search window, the FOV is widened and steps i)-iv) are repeated for a new FOV, e.g., according to the "As Low As Reasonably Achievable" (ALARA) principle. In this way, two or more fetal hearts are reliably tagged and monitored during labor, even when the fetuses move.

[0102] 10, there is a computer-implemented method of monitoring a first fetal heartbeat. Briefly, in a first step 1002, the method includes i) obtaining ultrasound image data from an ultrasound transducer. In a second step 1004, the method includes ii) determining a location of the first fetal heart within the ultrasound image data. In a third step 1006, the method includes iii) sending a message to the ultrasound transducer to instruct the ultrasound transducer to switch to a Doppler mode, the determined location being used to set a target location for the Doppler mode. In a fourth step 1008, the method includes iv) monitoring the first fetal heartbeat using the ultrasound data obtained in the Doppler mode.

[0103] The steps of obtaining ultrasound image data from an ultrasound transducer have been described in detail above with respect to the apparatus 100 , and it will be understood that the details and embodiments described therein apply equally to the method 1000 .

[0104] The step of determining the location of the first fetal heart within the ultrasound image data has been described in detail above with respect to the apparatus 100, and it will be understood that the details and embodiments described therein apply equally to the method 1000.

[0105] The step of sending a message to the ultrasound transducer to instruct it to switch to Doppler mode, where the determined location is used to set a target location for the Doppler mode, has also been described above with respect to device 100, and it will be understood that the details and embodiments described therein apply equally to method 1000.

[0106] The step of monitoring the first fetal heart rate using ultrasound data obtained in Doppler mode has also been described above with respect to apparatus 100, and it will be understood that the details and embodiments described therein apply equally to method 1000.

[0107] In another embodiment, a computer program product is provided that includes a computer readable medium having computer readable code embodied therein, the computer readable code being configured, when executed by a suitable computer or processor, to cause the computer or processor to perform one or more of the methods described herein.

[0108] It will therefore be appreciated that the present disclosure also applies to computer programs adapted to put the embodiments into practice, in particular computer programs on or in a carrier. The programs may be in the form of object code, such as source code, object code, code intermediate sources and partially compiled forms, or in any other form suitable for use in the implementation of the methods according to the embodiments described herein.

[0109] It will be further recognized that such programs may have many different architectural designs. For example, the program code implementing the functions of the method or system may be subdivided into one or more subroutines. Many different ways of distributing functionality among these subroutines will be apparent to those skilled in the art. The subroutines may be stored together in one executable file to form a self-contained program. Such an executable file includes computer executable instructions, such as processor instructions and / or interpreter instructions (e.g., Java interpreter instructions). Alternatively, one or more or all of the subroutines may be stored in at least one external library file and linked, e.g., at run-time, either statically or dynamically, to the main program. The main program includes at least one call to at least one of the subroutines. The subroutines further include function calls to each other.

[0110] The carrier of a computer program may be any entity or device capable of carrying the program. For example, the carrier comprises a data storage such as a ROM, for example a CD ROM or a semiconductor ROM, or a magnetic recording medium, for example a hard disk. Furthermore, the carrier may be a transmissible carrier such as an electric or optical signal conveyed via an electric or optical cable or by radio or other means. When the program is embodied in such a signal, the carrier is constituted by such a cable or other device or means. Alternatively, the carrier may be an integrated circuit in which the program is embedded, the integrated circuit being adapted for, or used for, performing the relevant method.

[0111] Modifications to the disclosed embodiments can be understood and effected by those skilled in the art from a study of the drawings, the disclosure, and the appended claims, in practicing the principles and techniques described herein. In the claims, the words "comprise, include, have" do not exclude other elements or steps, and singular elements do not exclude a plurality. A single processor or other unit performs the functions of several items recited in the claims. The mere fact that certain means are recited in mutually different dependent claims does not indicate that a combination of these means cannot be used to advantage. A computer program may be stored or distributed on a suitable medium, such as an optical storage medium or a solid medium, provided together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be interpreted as limiting the scope.

Claims

1. 1. A device for monitoring a first fetal heart rate, the device comprising: a memory containing instruction data representing a set of instructions; a processor in communication with the memory for executing the set of instructions, the set of instructions, when executed by the processor, i) obtaining ultrasound image data from each ultrasound transducer forming part of an array of ultrasound transducers and selecting a first ultrasound transducer from the array of ultrasound transducers having a view of the first fetal heart; ii) determining a location of the heart of the first fetus within ultrasound image data acquired using the first ultrasound transducer; iii) sending a message to the first ultrasound transducer to instruct the first ultrasound transducer to switch to a Doppler mode, where the determined location of the heart of the first fetus is used to set a target location of the Doppler mode; iv) monitoring the heart rate of the first fetus using ultrasound data acquired in the Doppler mode; An apparatus for causing the processor to perform the steps of:

2. 2. The apparatus of claim 1, wherein the processor determines a location of the heart of the first fetus within the ultrasound image data using a first machine learning model trained to identify fetal heart locations in ultrasound images.

3. The processor further comprises: At periodic time intervals, if the audibility of the heartbeat of the first fetus is below a threshold audibility in the ultrasound data obtained in the Doppler mode; if the signal-to-noise magnitude of the heart beat falls below a threshold signal-to-noise level; and / or If the monitored heart rate is interrupted, The apparatus according to claim 1 or 2, wherein steps i) to iv) are repeated.

4. The processor further comprises: selecting a second ultrasound transducer from the array of ultrasound transducers having a second view of the fetal heart; 4. The apparatus of claim 1, further comprising: an ultrasound transducer for detecting a fetal heartbeat, the ultrasound transducer comprising: a second ultrasound transducer for detecting a fetal heartbeat; a second ultrasound transducer for detecting a fetal heartbeat;

5. The processor further comprises: determining a first plurality of landmarks of the heart of the first fetus in the ultrasound image data of the first fetus and a second plurality of landmarks of the heart of the second fetus in the ultrasound image data of the second fetus; 5. The apparatus of claim 4, further comprising: repeating steps i) to iv) on new ultrasound image data to identify the first fetus using the first plurality of landmarks and to identify the second fetus using the second plurality of landmarks for continuous monitoring of the heartbeat of the first fetus and the heartbeat of the second fetus.

6. 6. The apparatus of claim 5, wherein the first plurality of landmarks and / or the second plurality of landmarks are determined using a second machine learning model trained to identify landmarks based on pose estimation.

7. The apparatus of claim 1 , wherein the array of ultrasonic transducers comprises an array of capacitive micromachined ultrasonic transducers.

8. The processor further comprises: obtaining data from a plurality of inertial measurement unit sensors; each inertial measurement unit sensor is attached to a respective transducer in the array of ultrasonic transducers; 8. An apparatus according to claim 1, wherein the processor uses the data from the inertial measurement unit sensors to determine an orientation of each respective transducer.

9. The apparatus of claim 8 , wherein the processor is further configured to determine an absolute position or presentation of the first fetus from the data from the inertial measurement unit sensor.

10. 10. An ultrasound system including the device of any one of claims 1 to 9, further including the ultrasound transducer.

11. Further comprising a display; the processor sends instructions to the display to display the ultrasound image data and / or the ultrasound data obtained in the Doppler mode on the display; 11. The ultrasound system of claim 10, wherein the processor further sends instructions to the display to cause the display to provide an indication of whether the ultrasound transducer is operating in an imaging mode or a Doppler mode.

12. 1. A computer-implemented method for monitoring a first fetal heart rate, the computer-implemented method comprising: i) obtaining ultrasound image data from each ultrasound transducer forming part of an array of ultrasound transducers and selecting a first ultrasound transducer from the array of ultrasound transducers having a view of the first fetal heart; ii) determining a location of the heart of the first fetus within ultrasound image data acquired using the first ultrasound transducer; iii) sending a message to the first ultrasonic transducer to instruct the first ultrasonic transducer to switch to a Doppler mode, the determined location being used to set a target location for the Doppler mode; iv) monitoring the heart rate of the first fetus using ultrasound data acquired in the Doppler mode; 13. A computer-implemented method comprising:

13. 13. A computer readable medium having computer readable code embodied therein, the computer readable code being configured, when executed by a suitable computer or processor, to cause the computer or processor to perform the computer implemented method of claim 12.

Images