WO2025149396A1

WO2025149396A1 - Multiple pregnany/gestation detection using ultrasound imaging with blind sweep protocol

Info

Publication number: WO2025149396A1
Application number: PCT/EP2025/050003
Authority: WO
Inventors: Anuradha Anuradha; Subhendu Seth; Krishnan V MANIKANDA; Madhumita Gupta; Pallavi Vajinepalli
Original assignee: Koninklijke Philips NV
Current assignee: Koninklijke Philips NV
Priority date: 2024-01-12
Filing date: 2025-01-02
Publication date: 2025-07-17
Anticipated expiration: 2026-07-12

Abstract

A system includes a processor configured for communication with an ultrasound probe, where the processor is configured to: control the ultrasound probe to obtain a plurality of ultrasound image frames during a blind sweep protocol on a patient with a pregnancy; provide the plurality of ultrasound image frames as an input to a deep learning network trained to detect at least one of maternal anatomy or fetal anatomy; and generate, as an output of the deep learning network, a detection. The detection includes at least one of an inter-twin membrane within the plurality of ultrasound image frames, or a plurality of fetal anatomical parts within the plurality of ultrasound image frames. The processor is further configured to determine, using the detection, whether the pregnancy comprises a multiple gestation; and provide, to a display in communication with the processor, an output representative of the determination.

Description

MULTIPLE PREGNANY/GESTATION DETECTION USING ULTRASOUND IMAGING WITH BLIND SWEEP PROTOCOL

TECHNICAL FIELD

[0001] The subject matter described herein relates to devices, systems, and methods for using ultrasound data from a blind abdominal imaging sweep to detect multiple pregnancies/gestations.

BACKGROUND

[0002] Ultrasound imaging is often used for diagnostic purposes in an office or hospital setting, but may also be used in resource-constrained care settings (e.g., homes, accident sites, ambulances, mobile health facilities, etc.) by emergency personnel, home health nurses, midwives, etc., who may lack ultrasound expertise. To facilitate ultrasound image acquisition by untrained or minimally trained users, a “blind sweep” protocol is often employed, in which the user follows pre-determined probe paths (e.g., sweeping out a pattern on the patient’s abdomen) during imaging.

[0003] Ultrasound imaging is a vital component of high-quality obstetric care. For example, detection of multiple pregnancies through ultrasound can allow for appropriate referral for delivery care in highly resourced centers with providers trained to handle any associated risks and complications. However, in rural and under-resourced communities, the scarcity of ultrasound imaging results in a considerable gap in the healthcare of pregnant mothers.

[0004] Multiple gestation can constitute a significant risk to both mother and fetuses, and is therefore clinically categorized as a type of high-risk pregnancy. Both diagnosis and management of these pregnancies are a challenge. Antepartum complications - including preterm labor, preterm premature rupture of the membranes, intrauterine growth restriction (IUGR), intrauterine fetal demise, gestational diabetes, and preeclampsia - develop in over 80% of multiple pregnancies as compared with approximately 25% of singleton gestation. While ultrasound can be very useful in identifying multiple gestation access to ultrasound may be out of reach in low-resource settings, where the expertise required to identify multiple pregnancies is also limited. Access and skill to perform the ultrasound in low-resource setting is therefore deficient in many parts of the world.

[0005] The information included in this Background section of the specification, including any references cited herein and any description or discussion thereof, is included for technical reference purposes only and is not to be regarded as subject matter by which the scope of the disclosure is to be bound.

SUMMARY

[0006] Disclosed is an ultrasound blind sweep multiple pregnancy detection system that, following a blind sweep protocol (sweeps at particular locations along body without trying to find specific anatomy), detects anatomical features in the captured images and uses them to identify patients who may need to be referred for evaluation by human experts. For example, the system may detect multiple pregnancy (e.g., twins, triplets, etc.), also known as multiple gestation. A pregnant patient can be referred to an obstetrician trained to deal with multiple pregnancy/gestation based on the output of the system. A deep learning network, such as a convolutional neural network, is used to detect anatomy of the pregnant patient or the one or multiple fetuses inside the patient’s uterus (e.g., inter-twin membrane, fetal head, fetal heart, etc.). A benefit of the ultrasound blind sweep anatomy detection system is to improve the quality of care by using information obtained during the blind sweep protocol to identify medical conditions (e.g., multiple gestation/pregnancy, etc.) that may require expert care, compared to a single/singleton pregnancy (only one fetus).

[0007] A system of one or more computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination of them installed on the system that in operation causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.

[0008] One general aspect includes a system including a processor configured for communication with an ultrasound probe, where the processor is configured to: control the ultrasound probe to obtain a plurality of ultrasound image frames during a blind sweep protocol on a patient with a pregnancy; provide the plurality of ultrasound image frames as an input to a deep learning network trained to detect at least one of maternal anatomy or fetal anatomy; generate, as an output of the deep learning network, a detection of at least one of: an inter-twin membrane within the plurality of ultrasound image frames; or a plurality of fetal anatomical parts within the plurality of ultrasound image frames; determine, using the detection of at least one of the inter-twin membrane or the plurality of fetal anatomical parts, whether the pregnancy may include a multiple gestation; and provide, to a display in communication with the processor, an output representative of the determination of whether the pregnancy may include the multiple gestation. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.

[0009] Implementations may include one or more of the following features. In some aspects, the processor is configured to perform the determination of whether the pregnancy may include the multiple gestation based on at least one of: the detection of the inter-twin membrane; a co-existence of the plurality of fetal anatomical parts within a single ultrasound image frame; a geometric mapping of the plurality of fetal anatomical parts; or a geometric sequence of the plurality of fetal anatomical parts. In some aspects, to perform the determination of whether the pregnancy may include the multiple gestation, the processor is configured to combine results of at least two of: the detection of the inter-twin membrane; the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame; the geometric mapping of the plurality of fetal anatomical parts; or the geometric sequence of the plurality of fetal anatomical parts. In some aspects, the output of the deep learning network may include a plurality of detections of the inter-twin membrane in the plurality of ultrasound image frames, where, to perform the determination of whether the pregnancy may include the multiple gestation based on the detection of the inter-twin membrane in the plurality of ultrasound image frames, the processor is configured to: determine a quantity of the plurality of ultrasound image frames with the detection of the inter-twin membrane; compare the quantity to a threshold quantity; and determine that the pregnancy may include the multiple gestation when the quantity exceeds the threshold quantity. In some aspects, the output provided to the display may include: at least one ultrasound image frame with the detection of the inter-twin membrane; and a bounding box identifying the inter-twin membrane overlaid on the at least one ultrasound image frame. In some aspects, the output of the deep learning network for the single ultrasound image frame may include a plurality of detections of the plurality of fetal anatomical parts, where, to perform the determination of whether the pregnancy may include the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame, the processor is configured to: provide the plurality of detections as input to at least one of a statistical model or a rule-based expert system; and generate, as an output of at least one of the statistical model or the rule-based expert system, a determination that the single ultrasound image frame is indicative of the multiple gestation. In some aspects, to perform the determination of whether the pregnancy may include the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame, the processor is configured to: repeat the determination that the single ultrasound image frame is indicative of the multiple gestation for the plurality of ultrasound image frames; determine a quantity of the plurality of ultrasound image frames indicative of the multiple gestation; compare the quantity to a threshold quantity; and determine that the pregnancy may include the multiple gestation when the quantity exceeds the threshold quantity. In some aspects, the output provided to the display may include: the single ultrasound image frame; and a plurality of bounding boxes identifying the plurality of fetal anatomical parts overlaid on the single ultrasound image frame. In some aspects, the output of the deep learning network may include a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, where, to perform the determination of whether the pregnancy may include the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts, the processor is configured to: generate a spatial mapping of the plurality of fetal anatomical parts; and perform the determination that the pregnancy may include the multiple gestation using a rule-based expert system and the spatial mapping. In some aspects, to perform the determination of whether the pregnancy may include the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts, the processor is configured to: detect a plurality of regions with the plurality of fetal anatomical parts in the spatial mapping; determine at least one distance between the plurality of regions; provide, as input to the rule-based expert system, the plurality of regions and the at least one distance; and generate, as an output of the rulebased expert system, the determination that the pregnancy may include the multiple gestation. In some aspects, the output provided to the display may include: the spatial mapping. In some aspects, the output of the deep learning network may include a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, where, to perform the determination of whether the pregnancy may include the multiple gestation based on the geometric sequence of the plurality of fetal anatomical parts, the processor is configured to: generate a text sequence of anatomy tags representative of the plurality of fetal anatomical parts and the plurality of ultrasound image frames; provide the text sequence as an input to a sequence model; and generate, as an output of the sequence model, the determination that the pregnancy may include the multiple gestation. In some aspects, the plurality of ultrasound image frames may include a first sweep and a second sweep of the blind sweep protocol, where a physical location associated with the anatomy tags from an ending of a first portion of the first sweep in the text sequence proximate to the physical location associated with the anatomy tags from the beginning of a second portion of the second sweep in the text sequence. In some aspects, the system may include the ultrasound probe. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.

[0010] One general aspect includes a method which includes controlling, with a processor, an ultrasound probe in communication with the processor to obtain a plurality of ultrasound image frames during a blind sweep protocol on a patient with a pregnancy; providing, with the processor, the plurality of ultrasound image frames as an input to a deep learning network trained to detect anatomy of the patient; generating, with the processor and as an output of the deep learning network, a detection of at least one of: an inter-twin membrane within the plurality of ultrasound image frames; or a plurality of fetal anatomical parts within the plurality of ultrasound image frames; determining, with the processor, whether the pregnancy may include a multiple gestation, using the detection of at least one of the inter-twin membrane or the plurality of fetal anatomical parts; and providing, with the processor, an output representative of the determination of whether the pregnancy may include the multiple gestation to a display in communication with the processor. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.

[0011] Implementations may include one or more of the following features. In some aspects, determining whether the pregnancy may include the multiple gestation is based on at least one of: the detection of the inter-twin membrane; a co-existence of the plurality of fetal anatomical parts within a single ultrasound image frame; a geometric mapping of the plurality of fetal anatomical parts; or a geometric sequence of the plurality of fetal anatomical parts. In some aspects, the output of the deep learning network may include a plurality of detections of the inter-twin membrane in the plurality of ultrasound image frames, where determining whether the pregnancy may include the multiple gestation based on the detection of the inter-twin membrane in the plurality of ultrasound image frames may include: determining a quantity of the plurality of ultrasound image frames with the detection of the inter-twin membrane; comparing the quantity to a threshold quantity; and determining that the pregnancy may include the multiple gestation when the quantity exceeds the threshold quantity. In some aspects, the output of the deep learning network for the single ultrasound image frame may include a plurality of detections of the plurality of fetal anatomical parts, where determining whether the pregnancy may include the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame may include: providing the plurality of detections as input to at least one of a statistical model or a rule-based expert system; and generating, as an output of at least one of the statistical model or the rule-based expert system, a determination that the single ultrasound image frame is indicative of the multiple gestation. In some aspects, the output of the deep learning network may include a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, where determining whether the pregnancy may include the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts may include: generating a spatial mapping of the plurality of fetal anatomical parts; and performing the determination that the pregnancy may include the multiple gestation using a rule-based expert system and the spatial mapping. In some aspects, the output of the deep learning network may include a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, where determining whether the pregnancy may include the multiple gestation based on the geometric sequence of the plurality of fetal anatomical parts may include: generating a text sequence of anatomy tags representative of the plurality of fetal anatomical parts and the plurality of ultrasound image frames; providing the text sequence as an input to a sequence model; and generating, as an output of the sequence model, the determination that the pregnancy may include the multiple gestation. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.

[0012] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter. A more extensive presentation of features, details, utilities, and advantages of the ultrasound blind sweep multiple pregnancy detection system, as defined in the claims, is provided in the following written description of various aspects of the disclosure and illustrated in the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Illustrative aspects of the present disclosure will be described with reference to the accompanying drawings, of which:

[0014] Figure l is a schematic, diagrammatic representation of an ultrasound imaging system, according to aspects of the present disclosure.

[0015] Figure l is a schematic diagram of a processor circuit, according to aspects of the present disclosure.

[0016] Figure 3A is a set of schematic, diagrammatic, cross-sectional views of two fetuses surrounded by an amniotic sac and chorionic sac within a uterus of a patient, representing different types of twins, according to aspects of the present disclosure.

[0017] Figure 3B is a set of schematic, diagrammatic, cross-sectional views of two fetuses arranged in different positions within a uterus of a patient as seen by an ultrasound imaging plane, according to aspects of the present disclosure.

[0018] Figure 4 is a schematic, diagrammatic representation of a patient on whose abdomen the ultrasound blind sweep protocol will be followed, according to aspects of the present disclosure.

[0019] Figure 5 is a schematic, diagrammatic representation, in hybrid block diagram/flow diagram form, of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure.

[0020] Figure 6 is a schematic, diagrammatic representation, in block diagram form, of at least a portion of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure.

[0021] Figure 7A is a schematic, diagrammatic overview, in block diagram form, of a training mode for an untrained neural network, according to aspects of the present disclosure.

[0022] Figure 7B is a schematic, diagrammatic overview, in block diagram form, of an inference mode or clinical usage mode for the trained neural network, according to aspects of the present disclosure.

[0023] Figure 8 is a schematic, diagrammatic illustration, in block diagram form, of the detection of anatomy (e.g., the head, heart, abdomen, pelvis, placenta, membrane, etc.), according to aspects of the present disclosure.

[0024] Figure 9 is a schematic, diagrammatic representation, in flow diagram form, of an example inter-twin membrane detection method, according to aspects of the present disclosure. [0025] Figure 10 is an ultrasound image frame that includes an inter-twin membrane bounding box, indicating that an inter-twin membrane has been detected, according to aspects of the present disclosure.

[0026] Figure 11 is a schematic, diagrammatic representation, in hybrid flow diagram and block diagram form, of an example method for detection of multiple co-existing fetal parts within an image frame, within a sweep, or within all of the sweeps of a blind sweep protocol, according to aspects of the present disclosure.

[0027] Figure 12 is an ultrasound image frame that includes a single placenta bounding box and two separate, non-overlapping head bounding boxes, indicating that two fetal heads have been detected, according to aspects of the present disclosure.

[0028] Figure 13 is a schematic, diagrammatic representation, in hybrid block diagram / flow diagram form, of an example fetal parts geometric mapping method, according to aspects of the present disclosure.

[0029] Figure 14 is a schematic, diagrammatic view of a scanning process, according to aspects of the present disclosure.

[0030] Figure 15A is a graphical representation of a horizontal scan or horizontal sweep, according to aspects of the present disclosure.

[0031] Figure 15B is a graphical representation of a distance calculation for a horizontal sweep, according to aspects of the present disclosure.

[0032] Figure 16A is a graphical representation of a vertical scan or vertical sweep, according to aspects of the present disclosure.

[0033] Figure 16B is a graphical representation of a distance calculation for a vertical sweep, according to aspects of the present disclosure.

[0034] Figure 17 is a screen display of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure.

[0035] Figure 18 is a spatial mapping screen display of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure. [0036] Figure 19 is a schematic, diagrammatic representation, in hybrid block diagram / flow diagram form, of an example fetal parts geometric sequencing method, according to aspects of the present disclosure.

[0037] Figure 20 is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method, according to aspects of the present disclosure. [0038] Figure 21A is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method, according to aspects of the present disclosure.

[0039] Figure 21B is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method, according to aspects of the present disclosure.

[0040] Figure 22 is a is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method, according to aspects of the present disclosure.

DETAILED DESCRIPTION

[0041] In accordance with at least one aspect of the present disclosure, an ultrasound blind sweep multiple pregnancy detection system is provided which can identify anatomy of interest such as anatomy indicating two or more fetuses are present - based on a blind sweep protocol with an ultrasound imaging probe. The ultrasound blind sweep multiple pregnancy detection system presents a novel approach to imaging quality control by ensuring that features such as multiple leads, multiple spines, etc. are detected, and using the detected locations of these features to, for example, determine whether the pregnancy is multiple and thus potentially in need of expert care.

[0042] Ultrasound imaging is a vital component of high-quality obstetric care. In rural and under-resourced communities, the scarcity of ultrasound imaging results in a considerable gap in healthcare for pregnant mothers. Increased detection of pregnancy complications through ultrasound can allow for appropriate referral for delivery care in more- resourced centers with more highly trained providers. The present disclosure seeks to overcome this barrier to ultrasound access in a locally sustainable and resource-conscious way, using standardized blind-sweep scanning protocols combined with artificial intelligence, obviating the need for an interpreting provider (e.g., a radiologist or obstetrician) and an experienced sonographer in the remote location.

[0043] Monitoring intrauterine structures is essential for normal fetal development and perinatal outcome. Women with multiple pregnancy / multiple gestation may be at increased risk of maternal, fetal and postnatal adverse outcomes. The present disclosure provides a computer assisted simple triaging (CAST) setting for ultrasound systems, to assist novice users (e.g., midwives with minimal training) with an algorithm for obstetrical ultrasound screening for multiple gestation that pre-selects appropriate referral cases for experts/trained providers from community health, for delivery care in more resourced centers. Automated CAST can help rural health providers like midwives with minimal training/non-expertise to screen out benign cases (e.g., low-risk or non-emergent cases that may not need referral), to avoid additional steps in the workflow, and to increase exam capabilities in community medicine.

[0044] The obstetric blind sweep protocol can be taught to health care workers in a very short period, enabling them to acquire high-quality ultrasound images of pregnant mothers. There have been encouraging results on determination of gestational age and fetal presentation from this type of blind sweep protocol. A need exists for detection of multiple gestation from this automated process, as it is one type of high-risk pregnancy.

[0045] Users can be readily trained to perform blind sweeps. Combining this with an AI- guided automatic detection of more than a single pregnancy, and categorization of the pregnancy as multiple, can be a boon in these settings where the high-risk mothers can be referred for further diagnosis and tertiary care. One aim of the present disclosure is the identification of multiple pregnancies based on automatic detection of multiple coexistences of anatomies. These may for example be the same type e.g., head of both the fetuses, or may be different anatomies e.g., axial head and axial abdomen/heart. The multiple detection may be in a single frame, within a single sweep, or across multiple sweeps in the blind sweep protocol. Identification of fetal membranes may be statistically significant in their frequency of occurrence, which can be identified again by the Al solution (e.g., a detector module such as YOLO) to support the presence of twin gestation.

[0046] The present disclosure provides a solution to the problem of identifying the number of fetuses in the maternal womb at a time. Since multiple gestation is a high-risk pregnancy that can result in maternal fetal complications, it may be highly advantageous to diagnose and refer to higher centers for confirmation, determining the chronicity and amnionicity, and follow-up. While the obstetric sweep protocol takes the images of the fetus in a simplified manner which can be easily taught to users, the ALguided automatic detection of the fetus will let the user know the number of fetuses inside the mother’s uterus. Once the detection of multiple gestations is done, the user can refer the patient to higher centers for further management or follow-up of the patient. Diagnosis of multiple gestation may help manage maternal and fetal complications, if any, and thus reduce perinatal mortality and morbidity.

[0047] The anatomical data may be acquired from a 3x3 or 5x5 blind sweep protocol, which covers most of the uterus of the mother. From these images of these sweeps (cineloops or cine scans), the present disclosure provides a method to identify such multiple gestations. By screening out benign cases (e.g., low-risk, non-emergent cases that may not need referral), the system can avoid additional steps in the workflow , thus saving time and effort while freeing up experienced healthcare professionals, especially in poorer regions with clinician shortages.

[0048] Aspects of the present disclosure can include features described in U.S. Provisional Application No. 63/540,740, filed September 27, 2023, titled “Ultrasound Imaging With Ultrasound Probe Guidance In Blind Sweep Protocol” and/or U.S. Provisional Application No. 63/540,755, filed September 27, 2023, titled “Ultrasound Imaging with Follow Up Sweep Guidance After Blind Sweep Procedure”, which are incorporated by reference as though fully set forth herein.

[0049] The ultrasound blind sweep multiple pregnancy detection system provides multiple ways to detect multiple pregnancies from the blind sweep protocol.

[0050] In one aspect, frame-level anatomical coexistences can be used as a marker of multiple gestations. The ultrasound blind sweep multiple pregnancy detection system identifies these kinds of signature coexistence inside data acquired with blind sweep protocol. These coexistences can be two fetal anatomical parts (e.g., of which each fetus has only one), e.g., the head of both fetuses or the axial appearance of the head of one fetus against, e.g., the existence of the abdomen or heart of another and vice versa. Detection of these logical coexistences will help in ruling in multiple pregnancies.

[0051] In one aspect, detection of inter-twin membrane(s) can be used to rule in or rule out multiple gestation. Since fetal parts indicative of multiple gestation may not always coexist within a single frame, their presence can be at a distance from one another, which can be understood by looking at the entire inter/intra sweep data set. The presence of a fetal membrane (with a moderate prevalence) may be a secondary confirmatory feature in such situations. Clinically it can be highly visible, and may thus be targeted with any detector modules to isolate.

[0052] The identification of these conjugate clinical features has not previously been tackled in the algorithmic sense to rule in or rule out multiple gestation.

[0053] Recognition of multiple gestation relies on many factors. Duplicated /multiple anatomy detections of the same category in the same frame are one of the key factors to identify the multiplicity. The intervening membrane can be a diagnostic factor for separating two or multiple fetuses. However, the membranes may not be consistently found or identified in all twins. For example, monochori onic and monoamniotic twins do not have an intertwin membrane. There can also be technical issues such as sampling site, insonation angle, ultrasound setting, and the quality of the machine that affect visualization and the appearance of the membrane.

[0054] Despite all this, statistically, the incidence of dizygotic twins (all are almost invariably diamniotic dichorionic) is high in spontaneous pregnancies (about 76% of the twin pregnancies). Among monozygotic twin gestations, approximately one-third are dichorionic diamniotic, whereas almost two-thirds are monochorionic diamniotic. Thus, the probability of finding a separating membrane is high. Thus, combining the anatomy detector to recognize duplicated anatomy along with membrane detection can reduce false negatives.

The detector module can detect the head, heart, stomach, spine etc. as per the training given by suitable annotations. Whenever multiple heads, hearts or stomachs are detected in a single frame there can be a trigger to analyze the possibility of multiple gestations. Identification of an intertwin membrane provides an additional confidence-boosting factor for the identification of more than one gestation.

[0055] One aspect includes frame-level anatomical co-occurrences for identifying multiple gestations. The coexistence of significant fetal structures of different fetuses in a single frame is the confirmatory pattern of multiple gestation. Usually in clinical practice, these are regarded as confirmatory frame(s) to conclude multiple gestations.

[0056] Algorithm for arresting coexistence: Any detector module can be leveraged to identify different fetal parts, which in turn will help in identifying the logical coexistence of multiple structures.

[0057] In another aspect, the above structural coexistence can be learned via an Al algorithm either directly on B-mode images or on the detected space of anatomical parts (especially fetal parts). In inference mode, in real time or near-real time, the system is expected to classify the sweep/exam as multiple gestations giving higher importance to these structural coexistences (if any).

[0058] One aspect includes targeting Inter-twin membrane for identifying multiple gestations. Statistical significance: 76% of the total twin pregnancy is dizygotic (dichorionic- diamniotic, or fraternal). Hence, the incidence of a thicker membrane is high.

[0059] In monozygotic (e.g., identical twin) pregnancies, there is a roughly 20% chance to be di chori onic-di amniotic. Hence, the chance of a thin membrane is also possible. The remainder of twin or multiple pregnancies (roughly 19% of all cases) will not show any separating membrane. Thus, the presence of an identifiable membrane provides a handy way to identify 75-80% of the multiple gestation (MG) population, if detected properly.

[0060] One aspect includes the inter-twin membrane: a detector module can be leveraged to identify these unique appearances of inter-twin membranes, mostly suspended inside the amniotic fluid, which is statistically significant in frequency of occurrence.

[0061] The present disclosure aids substantially in the capture of high-quality ultrasoundbased diagnoses by minimally trained users, by automatically detecting multiple pregnancy / multiple gestation. Implemented on a processor in communication with an ultrasound probe, the ultrasound blind sweep multiple pregnancy detection system disclosed herein provides practical improvements in the quality of care available to patients in underserved areas. This improved imaging methodology transforms a process that is heavily reliant on professional experience into one that is accurate and repeatable even for minimally trained personnel, without the normally routine need to train clinicians such as emergency department personnel to recognize multiple gestation and other prenatal conditions. This unconventional approach improves the functioning of the ultrasound imaging system, by providing reliable, repeatable imaging and diagnosis in hospital, office, vehicle, field, and home settings, as well as referral recommendations for patients suspected to have a multiple pregnancy.

[0062] The ultrasound blind sweep multiple pregnancy detection system may be implemented as a process at least partially viewable on a display, and operated by a control process executing on a processor that accepts user inputs from a keyboard, mouse, or touchscreen interface, and that is in communication with one or more sensor probes. In that regard, the control process performs certain specific operations in response to different inputs or selections made at different times. Certain structures, functions, and operations of the processor, display, sensors, and user input systems provide novel features or aspects of the present disclosure.

[0063] These descriptions are provided for exemplary purposes only, and should not be considered to limit the scope of the ultrasound blind sweep multiple pregnancy detection system. Certain features may be added, removed, or modified without departing from the spirit of the claimed subject matter.

[0064] For the purposes of promoting an understanding of the principles of the present disclosure, reference will now be made to the aspects illustrated in the drawings, and specific language will be used to describe the same. It is nevertheless understood that no limitation to the scope of the disclosure is intended. Any alterations and further modifications to the described devices, systems, and methods, and any further application of the principles of the present disclosure are fully contemplated and included within the present disclosure as would normally occur to one skilled in the art to which the disclosure relates. In particular, it is fully contemplated that the features, components, and/or steps described with respect to one aspect may be combined with the features, components, and/or steps described with respect to other aspects of the present disclosure. For the sake of brevity, however, the numerous iterations of these combinations will not be described separately.

[0065] Figure l is a schematic, diagrammatic representation of an ultrasound imaging system 100, according to aspects of the present disclosure. The ultrasound imaging system 100 may for example be used to acquire ultrasound video sweeps, which can then be analyzed by a human clinician or an artificial intelligence to diagnose medical conditions. [0066] The ultrasound imaging system 100 is used for scanning an area or volume of a subject’s body. A subject may include a patient of an ultrasound imaging procedure, or any other person, or any suitable living or non-living organism or structure. The ultrasound imaging system 100 includes an ultrasound imaging probe 110 in communication with a host 130 over a communication interface or link 120. The probe 110 may include a transducer array 112, a beamformer 114, a processor circuit 116, and a communication interface 118. The host 130 may include a display 132, a processor circuit 134, a communication interface 136, and a memory 138 storing subject information.

[0067] In some aspects, the probe 110 is an external ultrasound imaging device including a housing 111 configured for handheld operation by a user. The transducer array 112 can be configured to obtain ultrasound data while the user grasps the housing 111 of the probe 110 such that the transducer array 112 is positioned adjacent to or in contact with a subject’s skin. The probe 110 is configured to obtain ultrasound data of anatomy within the subject’s body while the probe 110 is positioned outside of the subject’s body. In some aspects, the probe 110 can be an external ultrasound probe, a transthoracic probe, and/or a curved array probe. [0068] In other aspects, the probe 110 can be an internal ultrasound imaging device and may comprise a housing 111 configured to be positioned within a subject’s body. In some aspects, the probe 110 may be a curved array probe. Probe 110 may be of any suitable form for any suitable ultrasound imaging application including both external and internal ultrasound imaging.

[0069] For an ultrasound imaging device, the transducer array 112 emits ultrasound signals towards an anatomical object 105 of a subject and receives echo signals reflected from the object 105 back to the transducer array 112. The ultrasound transducer array 112 can include any suitable number of acoustic elements, including one or more acoustic elements and/or a plurality of acoustic elements. In some instances, the transducer array 112 includes a single acoustic element. In some instances, the transducer array 112 may include an array of acoustic elements with any number of acoustic elements in any suitable configuration. For example, the transducer array 112 can include between 1 acoustic element and 10000 acoustic elements, including values such as 2 acoustic elements, 4 acoustic elements, 36 acoustic elements, 64 acoustic elements, 128 acoustic elements, 500 acoustic elements, 812 acoustic elements, 1000 acoustic elements, 3000 acoustic elements, 8000 acoustic elements, and/or other values both larger and smaller. In some instances, the transducer array 112 may include an array of acoustic elements with any number of acoustic elements in any suitable configuration, such as a linear array, a planar array, a curved array, a curvilinear array, a circumferential array, an annular array, a phased array, a matrix array, a one-dimensional (ID) array, a 1.x dimensional array (e.g., a 1.5D array), or a two- dimensional (2D) array. The array of acoustic elements (e.g., one or more rows, one or more columns, and/or one or more orientations) can be uniformly or independently controlled and activated. The transducer array 112 can be configured to obtain one-dimensional, two- dimensional, and/or three-dimensional images of a subject’s anatomy. In some aspects, the transducer array 112 may include a piezoelectric micromachined ultrasound transducer (PMUT), capacitive micromachined ultrasonic transducer (CMUT), single crystal, lead zirconate titanate (PZT), PZT composite, other suitable transducer types, and/or combinations thereof.

[0070] The object 105 may include any anatomy or anatomical feature, such an abdomen of a pregnant patient, one or multiple fetuses inside the abdomen of the pregnant patient, etc. [0071] The beamformer 114 is coupled to the transducer array 112. The beamformer 114 controls the transducer array 112, for example, for transmission of the ultrasound signals and reception of the ultrasound echo signals. In some aspects, the beamformer 114 may apply a time-delay to signals sent to individual acoustic transducers within an array in the transducer 112 such that an acoustic signal is steered in any suitable direction propagating away from the probe 110. The beamformer 114 may further provide image signals to the processor circuit 116 based on the response of the received ultrasound echo signals. The beamformer 114 may include multiple stages of beamforming. The beamforming can reduce the number of signal lines for coupling to the processor circuit 116. In some aspects, the transducer array 112 in combination with the beamformer 114 may be referred to as an ultrasound imaging component.

[0072] The processor 116 is coupled to the beamformer 114. The processor 116 may also be described as a processor circuit, which can include other components in communication with the processor 116, such as a memory, beamformer 114, communication interface 118, and/or other suitable components. The processor 116 may include a central processing unit (CPU), a graphical processing unit (GPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a controller, a field programmable gate array (FPGA) device, another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein. The processor 116 may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. The processor 116 is configured to process the beamformed image signals. For example, the processor 116 may perform filtering and/or quadrature demodulation to condition the image signals. The processor 116 and/or 134 can be configured to control the array 112 to obtain ultrasound data associated with the object 105. [0073] The communication interface 118 is coupled to the processor 116. The communication interface 118 may include one or more transmitters, one or more receivers, one or more transceivers, and/or circuitry for transmitting and/or receiving communication signals. The communication interface 118 can include hardware components and/or software components implementing a particular communication protocol suitable for transporting signals over the communication link 120 to the host 130. The communication interface 118 can be referred to as a communication device or a communication interface module.

[0074] The communication link 120 may be any suitable communication link. For example, the communication link 120 may be a wired link, such as a universal serial bus (USB) link or an Ethernet link. Alternatively, the communication link 120 may be a wireless link, such as an ultra-wideband (UWB) link, an Institute of Electrical and Electronics Engineers (IEEE) 802.11 WiFi link, or a Bluetooth link.

[0075] At the host 130, the communication interface 136 may receive the image signals. The communication interface 136 may be substantially similar to the communication interface 118. The host 130 may be any suitable computing and display device, such as a workstation, a personal computer (PC), a laptop, a tablet, or a mobile phone.

[0076] The processor 134 is coupled to the communication interface 136. The processor 134 may also be described as a processor circuit, which can include other components in communication with the processor 134, such as the memory 138, the communication interface 136, an optional speaker 139, and/or other suitable components. The processor 134 may be implemented as a combination of software components and hardware components. The processor 134 may include a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a controller, an FPGA device, another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein. The processor 134 may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. The processor 134 can be configured to generate image data from the image signals received from the probe 110. The processor 134 can apply advanced signal processing and/or image processing techniques to the image signals. In some aspects, the processor 134 can form a three-dimensional (3D) volume image from the image data. In some aspects, the processor 134 can perform real-time processing on the image data to provide a streaming video of ultrasound images of the object 105. In some aspects, the host 130 includes a beamformer. For example, the processor 134 can be part of and/or otherwise in communication with such a beamformer. The beamformer in the in the host 130 can be a system beamformer or a main beamformer (providing one or more subsequent stages of beamforming), while the beamformer 114 is a probe beamformer or micro-beamformer (providing one or more initial stages of beamforming).

[0077] The memory 138 is coupled to the processor 134. The memory 138 may be any suitable storage device, such as a cache memory (e.g., a cache memory of the processor 134), random access memory (RAM), magnetoresistive RAM (MRAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), flash memory, solid state memory device, hard disk drives, solid state drives, other forms of volatile and non-volatile memory, or a combination of different types of memory.

[0078] The memory 138 can be configured to store subject information, measurements, data, or files relating to a subject’s medical history, history of procedures performed, anatomical or biological features, characteristics, or medical conditions associated with a subject, computer readable instructions, such as code, software, or other application, as well as any other suitable information or data. The memory 138 may be located within the host 130. Subject information may include measurements, data, files, other forms of medical history, such as but not limited to ultrasound images, ultrasound videos, and/or any imaging information relating to the subject’s anatomy. The subject information may include parameters related to an imaging procedure such as an anatomical scan window, a probe orientation, and/or the subject position during an imaging procedure. The memory 138 can also be configured to store information related to the training and implementation of machine learning algorithms (e.g., neural networks) and/or information related to implementing image recognition algorithms for detecting/ segmenting anatomy, image quantification algorithms, and/or image acquisition guidance algorithms, including those described herein.

[0079] The display 132 is coupled to the processor circuit 134. The display 132 may be a monitor or any suitable display. The display 132 is configured to display the ultrasound images, image videos, and/or any imaging information of the object 105.

[0080] The ultrasound imaging system 100 may be used to assist a sonographer in performing an ultrasound scan. The scan may be performed in a point-of-care setting. In some instances, the host 130 is a console or movable cart. In some instances, the host 130 may be a mobile device, such as a tablet, a mobile phone, or portable computer. During an imaging procedure, the ultrasound system can acquire an ultrasound image of a particular region of interest within a subject’s anatomy. The ultrasound imaging system 100 may then analyze the ultrasound image to identify various parameters associated with the acquisition of the image such as the scan window, the probe orientation, the subject position, and/or other parameters. The ultrasound imaging system 100 may then store the image and these associated parameters in the memory 138. At a subsequent imaging procedure, the ultrasound imaging system 100 may retrieve the previously acquired ultrasound image and associated parameters for display to a user which may be used to guide the user of the ultrasound imaging system 100 to use the same or similar parameters in the subsequent imaging procedure, as will be described in more detail hereafter.

[0081] In some aspects, the processor 134 may utilize deep learning-based prediction networks to identify parameters of an ultrasound image, including an anatomical scan window, probe orientation, subject position, identify and location of anatomical features, and/or other parameters. In some aspects, the processor 134 may receive metrics or perform various calculations relating to the region of interest imaged or the subject’s physiological state during an imaging procedure. These metrics and/or calculations may also be displayed to the sonographer or other user via the display 132.

[0082] In some aspects, the host 130 may also include a speaker 180. The speaker 180 may for example be used to provide advisory tones, beeps, or other auditory feedback to the user.

[0083] Before continuing, it should be noted that the examples described above are provided for purposes of illustration, and are not intended to be limiting. Other devices and/or device configurations may be utilized to carry out the operations described herein.

[0084] Figure l is a schematic diagram of a processor circuit 250, according to aspects of the present disclosure. The processor circuit 250 may be implemented in the ultrasound imaging system 100, or other devices or workstations (e.g., third-party workstations, network routers, etc.), or on a cloud processor or other remote processing unit, as necessary to implement the method. As shown, the processor circuit 250 may include a processor 260, a memory 264, and a communication module 268. These elements may be in direct or indirect communication with each other, for example via one or more buses.

[0085] The processor 260 may include a central processing unit (CPU), a digital signal processor (DSP), a controller, or any combination of general-purpose computing devices, reduced instruction set computing (RISC) devices, application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other related logic devices, including mechanical and quantum computers. The processor 260 may also comprise another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein. The processor 260 may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0086] The memory 264 may include a cache memory (e.g., a cache memory of the processor 260), random access memory (RAM), magnetoresistive RAM (MRAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), flash memory, solid state memory device, hard disk drives, other forms of volatile and nonvolatile memory, or a combination of different types of memory. In an aspect, the memory 264 includes a non-transitory computer-readable medium. The memory 264 may store instructions 266. The instructions 266 may include instructions that, when executed by the processor 260, cause the processor 260 to perform the operations described herein. Instructions 266 may also be referred to as code. The terms “instructions” and “code” should be interpreted broadly to include any type of computer-readable statement(s). For example, the terms “instructions” and “code” may refer to one or more programs, routines, subroutines, functions, procedures, etc. “Instructions” and “code” may include a single computer-readable statement or many computer-readable statements.

[0087] The communication module 268 can include any electronic circuitry and/or logic circuitry to facilitate direct or indirect communication of data between the processor circuit 250, and other processors or devices. In that regard, the communication module 268 can be an input/output (I/O) device. In some instances, the communication module 268 facilitates direct or indirect communication between various elements of the processor circuit 250 and/or the ultrasound imaging system 100. The communication module 268 may communicate within the processor circuit 250 through numerous methods or protocols. Serial communication protocols may include but are not limited to United States Serial Protocol Interface (US SPI), Inter-Integrated Circuit (I²C), Recommended Standard 232 (RS- 232), RS-485, Controller Area Network (CAN), Ethernet, Aeronautical Radio, Incorporated 429 (ARINC 429), MODBUS, Military Standard 1553 (MIL-STD-1553), or any other suitable method or protocol. Parallel protocols include but are not limited to Industry Standard Architecture (ISA), Advanced Technology Attachment (ATA), Small Computer System Interface (SCSI), Peripheral Component Interconnect (PCI), Institute of Electrical and Electronics Engineers 488 (IEEE-488), IEEE-1284, and other suitable protocols. Where appropriate, serial and parallel communications may be bridged by a Universal Asynchronous Receiver Transmitter (UART), Universal Synchronous Receiver Transmitter (USART), or other appropriate subsystem.

[0088] External communication (including but not limited to software updates, firmware updates, model sharing between the processor and central server, or readings from the ultrasound imaging system 100) may be accomplished using any suitable wireless or wired communication technology, such as a cable interface such as a universal serial bus (USB), micro USB, Lightning, or FireWire interface, Bluetooth, Wi-Fi, ZigBee, Li-Fi, or cellular data connections such as 2G/GSM (global system for mobiles) , 3G/UMTS (universal mobile telecommunications system), 4G, long term evolution (LTE), WiMax, or 5G. For example, a Bluetooth Low Energy (BLE) radio can be used to establish connectivity with a cloud service, for transmission of data, and for receipt of software patches. The controller may be configured to communicate with a remote server, or a local device such as a laptop, tablet, or handheld device, or may include a display capable of showing status variables and other information. Information may also be transferred on physical media such as a USB flash drive or memory stick.

[0089] Figure 3A is a set of diagrammatic views of two fetuses 200, 201 surrounded by a at least one amniotic sac or amnion 210 and at least one chorionic sac or chorion 315 within a uterus of a patient, representing different types of twins, according to aspects of the present disclosure. At least one placenta 240 is also visible. Automated detection of multiple pregnancy or multiple gestation (e.g., twins) is an object of the present disclosure.

[0090] In a first example 310, the multiple pregnancy/gestation is monochorionic (e.g., having only one chorion 315) and monoamniotic (e.g., having only one amniotic sac 210 that contains both fetuses 200, 201, and has a single placenta 240.

[0091] In a second example 320, the multiple pregnancy/gestation is monochorionic (e.g., having one chorion 315) but diamniotic (e.g., having two amniotic sacs 210, 211, separated by a fetal membrane 350, with each amniotic sac holding one fetus 200, 201), and a single placenta 240. Detection of the fetal membrane is one way to detect a multiple gestation or multiple pregnancy, as described below.

[0092] In a third example 330, the multiple pregnancy/gestation is dichorionic (e.g., having two chorions 315, 316, separated by respective membranes 350) and diamniotic (e.g., having two amniotic sacs 210, 211, with each sac holding one twin 200, 201), and two placentas 240 that are fused together.

[0093] In a fourth example 340, the multiple pregnancy/gestation is dichorionic (e.g., having two separate chorions 315, 316, separated by respective membranes 350) and diamniotic (e.g., having two separate amniotic sacs 200, 201, each holding one twin 200, 201), and two separate placentas 240, 241. This type of twin pregnancy is associated with a lower rate of complications than monochorionic or monoamniotic pregnancies.

[0094] 70% of twin pregnancies are dizygotic (e.g., resulting from two fertilized eggs), and dizygotic pregnancies are believed to be always dichorionic and diamniotic. 30% of pregnancies are monozygotic (e.g., resulting from a single fertilized egg), and 20% of these pregnancies will also be dichorionic and diamniotic. Thus, detection of the fetal membrane 350 separating the two amniotic sacs can be effective in detecting up to 76% of twin pregnancies. The remaining 24% of detections may rely on other features of the pregnancy such as multiple occurrences of a unique anatomical feature (e.g., two heads, two hearts, two pelvises, etc.) and/or the spatial/geometrical relationship between fetal parts, as described below.

[0095] The uterus can be considered maternal anatomy. Depending on the context, the placenta may be considered fetal anatomy, maternal anatomy, or an interface between the two. [0096] Figure 3B is a set of schematic, diagrammatic, cross-sectional views of two fetuses 200, 201 arranged in different positions within a uterus of a patient as seen by an ultrasound imaging plane 360, according to aspects of the present disclosure. The imaging plane 360 is example of an imaging plane (e.g., for one frame during one sweep), and the same imaging plane shown in all fetal positions for twins, and indicates that different fetal parts will be visible in the same imaging plane depending on which fetal position twins are in. [0097] In a first example 370, both twins 200, 201 are in a vertex (head-down) position. Approximately 45% of twin pregnancies fall into this category. In example 370, the imaging plane 360 may not include any duplicate anatomy.

[0098] In a second example 375, one twin 200 is in a breech (head-up) position, and the other is in a vertex (head-down) position. Approximately 37% of twin pregnancies fall into this category. Breech births are associated with a higher rate of complications. In example 375, the imaging plane 360 includes the heads of both fetuses 200, 201.

[0099] In a third example 380, both twins 200, 201 are in a breech (head-up) position. Approximately 10% of twin pregnancies fall into this category. In example 380, the imaging plane 360 may not contain any duplicate anatomy. [00100] In a fourth example 385, one twin 200 is in a transverse (sideways) position, and the other twin 201 is in a vertex position. Approximately 5% of twin pregnancies fall into this category. Vertex births are associated with a higher rate of complications. In example 385, the imaging plane 360 may include the hearts of both fetuses 200, 201.

[00101] In a fifth example 390, one twin 200 is in the breech (head-up) position, and the other twin 201 is in the transverse (sideways) position. Approximately 2% of twin pregnancies fall into this category. In example 390, the imaging plane 360 may include the hearts of both fetuses 200, 201.

[00102] In a sixth example 395, both twins 200, 201 are in the vertex position.

Approximately 0.5% of twin pregnancies fall into this category. In example 395, the imaging plane 360 may include the hearts of both fetuses 200, 201.

[00103] Figure 4 is a schematic, diagrammatic representation of a patient 300 on whose abdomen the ultrasound blind sweep protocol will be followed, according to aspects of the present disclosure. Visible on the abdomen 310 of the patient 300 is a desired sweep pattern 320 intended to capture images of desired features of the patient’s anatomy. The sweep pattern 320 includes multiple vertical sweep lines 330 and multiple horizontal sweep lines 340. Each sweep line 330, 340 represents a desired path for one imaging sweep of the abdomen 310. In the example shown in Figure 4, the sweep pattern includes three vertical sweep lines 330 labeled L (patient’s left), M (patient’s middle), and R (patient’s right), all in an upward direction with respect to the patient, and three horizontal sweep lines 340 labeled Cl (bottom), C2 (middle), and C3 (top), all in a right-to-left direction with respect to the patient. However, it is understood that a sweep pattern 320 may include more or fewer sweep lines 330, including vertical sweep lines 330, horizontal sweep lines 330, or combinations thereof, in any combination of upward, downward, left, or right directions based on the patient’s fundal height. For example, if the patient’s belly is bigger in size, more sweeps may be needed. Furthermore, a sweep pattern 320 may cover other portions of the patient’s body, including but not limited to the head, neck, spine, limbs, etc. Types of blind sweep protocol include but are not limited to obstetric sweep imaging (OSI), volume sweep imaging (VSI), 6-Stage, Fetal Age Machine Learning Initiative (FAMLI), and Philips.

[00104] These sweep patterns represent desired probe motion information, including desired positions, a desired velocity or velocities, and/or a desired orientation of the ultrasound probe while the ultrasound probe is obtaining a plurality of ultrasound image frames during the sweep. It is noted that the desired sweep patterns or blind sweep protocols stored in a memory of the processor may include only vertical sweeps, only horizontal sweeps, may include a grid (e.g., 3x3, 5x5, etc.) of vertical and horizontal sweeps, and may also include associated parameters such as desired probe motion (e.g., positions, velocities, and/or orientations) stored in the memory(e.g., blind sweep protocol 440 in Fig. 5). Depending on the implementation, sweeps may include curved, diagonal, and other types of sweeps. The protocols and their associated parameters can for example be based on standards established by authorities in the field (physician organizations, sonographer organizations, etc.), published in scholarly journals/textbooks, etc.

[00105] Figure 5 is a schematic, diagrammatic representation, in hybrid block diagram/flow diagram form, of an example ultrasound blind sweep multiple pregnancy detection system 400, according to aspects of the present disclosure. An ultrasound probe 110 operated by a novice user 410 performs a blind sweep protocol 440 on the body of a patient 300 and sends ultrasound imaging data to a host 130 such as a tablet, smartphone, ultrasound cart, etc. In step 415, the host 130 generates ultrasound images using the ultrasound image data obtained by the ultrasound probe 110. The host 130 can control the ultrasound probe 110 to obtain the ultrasound image data (e.g., the host 130 establishes communication with the ultrasound probe 110, the host 130 sends control signals to start and/or stop acquisition of ultrasound image data, the host 130 sends power signals to power the ultrasound probe 110, etc.).

[00106] In step 420, the host 130 uses the ultrasound images generated in step 415 and/or the ultrasound image data used to generate the ultrasound images to determine if the pregnant mother has multiple pregnancy or multiple gestation, as described in more detail below. In step 430, the host generates a visual representation on a display, showing an indication of whether there is a multiple pregnancy and/or a graphic associated with the determination that there is a multiple pregnancy.

[00107] In step 450, the host 130 determines whether the planned path of the blind sweep protocol 440 has been followed adequately. If no, execution proceeds to steps 430 and 460, wherein the host 130 provides feedback (e.g., audio feedback through a speaker and/or visual feedback from a display) to repeat one or multiple sweeps of the blind sweep protocol 440. In some aspects, the determination in step 450 can be based on the results of analysis in step 420 of whether the patient has multiple pregnancy. For example, if there is insufficient ultrasound image data to perform the determinations in step 420 or if the ultrasound image data that has been obtained is not suitable to perform the determination in step 420, then step 450 can determine that the one or multiple sweeps of the blind sweep protocol 440 has not been completed correctly. The one or multiple sweeps that need to be performed again can be communicated to the user through visual or audio feedback in steps 430 and 460.

[00108] In some aspects, the planned path determined in step 450 of the blind sweep protocol 440 can be used by the host 130 to determine whether the patient has multiple pregnancy in step 420. For example, the planned path determined in step 450 can provide absolute or relative spatial information (e.g., vertical location, horizontal location) about the ultrasound image data obtained by the ultrasound probe and/or the ultrasound images generated by the host. For example, referring again to Fig. 4, the planned path determined in step 450 can indicate that the M sweep is horizontally located (left-right direction in Fig. 4) between the R sweep and the L sweep. The host 130 can use this spatial information from step 450 to process the ultrasound image data and/or ultrasound images to determine whether there is multiple pregnancy in step 420.

[00109] Based on the visual representation 430 and/or the audio guidance 460, the novice user 410 makes a triage decision to either rule out or rule in multiple pregnancy / gestation. In step 470, the novice user 410 rules out multiple pregnancy, and in step 480, the novice user 410 performs or recommends normal periodic evaluations for the patient 300. In step 490, the novice user 410 rules in multiple pregnancy, and in step 495, the novice user 410 refers the patient to an expert user such as an experienced sonographer, radiologist, obstetrician, or other physician for follow-up imaging and diagnosis.

[00110] Flow diagrams and block diagrams are provided herein for exemplary purposes; a person of ordinary skill in the art will recognize myriad variations that nonetheless fall within the scope of the present disclosure. For example, any of the steps described herein may optionally include an output to a user of information relevant to the step, and may thus represent an improvement in the user interface over existing art by providing information not otherwise available. Similarly, block diagrams may show a particular arrangement of components, modules, services, steps, processes, or layers, resulting in a particular data flow. It is understood that some embodiments of the systems disclosed herein may include additional components, that some components shown may be absent from some embodiments, and that the arrangement of components may be different than shown, resulting in different data flows while still performing the methods described herein. The logic of flow diagrams may be shown as sequential. However, similar logic could be parallel, massively parallel, object oriented, real-time, event-driven, cellular automaton, or otherwise, while accomplishing the same or similar functions. In order to perform the methods described herein, a processor may divide each of the steps described herein into a plurality of machine instructions, and may execute these instructions at the rate of several hundred, several thousand, several million, or several billion per second, in a single processor or across a plurality of processors. Such rapid execution may be necessary in order to execute the method in real time or near-real time as described herein. For example, to provide an assessment of fetal anatomy locations and/or an indication of whether the pregnancy is multiple, the system may need to generate the visual representation 430 within one second of completing the blind sweep protocol.

[00111] Figure 6 is a schematic, diagrammatic representation, in block diagram form, of at least a portion of an example ultrasound blind sweep multiple pregnancy detection system 400, according to aspects of the present disclosure. A set of ultrasound image sequences 610, also known as cineloops or cine scans (e.g., one cine scan per sweep of the blind sweep protocol) are received (whether one at t time, simultaneously, or in groups) by an object detector 620. Each cine scan can include a plurality of ultrasound image frames (e.g., 100- 1000 ultrasound image frames, and/or other values both larger and smaller). The object detector 620 may for example be a machine learning (ML) neural network as described below, although other types of object detectors may be used instead or in addition, including classical image recognition algorithms. The object detector 620 can be a deep learning network (e.g., convolutional neural network or CNN) trained to detect maternal anatomy and/or fetal anatomy. An output of the object detector may for example include annotated versions of the cineloops 610 that include bounding boxes for each detected anatomy in each frame where anatomy was detected. The object detector 620 can generate, as its output, a detection of at an inter-twin membrane within ultrasound image frames of the cine scan(s) and/or one or multiple fetal anatomical parts within the ultrasound image frames of the cine scan(s). Detected anatomy may for example include the fetal head, abdomen, heart, or pelvis - unique features of which each fetus is expected to have only one - as well as the inter-twin membrane 350 and/or placenta 240 (see Fig. 3 A).

[00112] The outputs of the object detector 620 are then passed to a multiple pregnancy/gestation determination 630, which may for example be a software, hardware, firmware, analog/digital logic, analog/digital circuitry, or combinations thereof. The multiple pregnancy/gestation determination 630 determines, using the detection of the inter-twin membrane and/or the plurality of fetal anatomical parts (from object detector 620), whether the pregnancy is a multiple pregnancy/gestation. In the example shown in Figure 6, the multiple pregnancy/gestation determination 630 includes an inter-twin membrane detection 900, a detection 1100 for detecting co-existence of fetal parts within an image frame, a geometric mapping of fetal parts 1300, and a geometric sequence of fetal parts 1900. In an example, the determination 900, detection 1100, mapping 1300, and sequence 1900 may have a simple binary output indicating whether or not a multiple pregnancy is detected. This detection may occur individually in every frame of every cineloop 610, or may occur across all the frames of a cineloop 610, or may occur across all of the cineloops collectively.

[00113] Optionally, a combiner 640 may aggregate the outputs of the determinations 900, 1100, 1300, and 1900. The combiner 640 can be a software, hardware, firmware, analog/digital logic, analog/digital circuitry, or combinations thereof. For example, the combiner 640 may use a majority vote system to determine whether a multiple pregnancy has been detected, such that three out of the four determinations must vote yes for a determination of multiple pregnancy to be made. In other aspects, the combiner 640 may indicate a multiple pregnancy if any one of the determinations 900, 1100, 1300, or 1900 detects a multiple pregnancy. In still other aspects, the combiner 640 may require unanimity from the determinations 900, 1100, 1300, or 1900. In still other aspects, the combiner may indicate a multiple pregnancy if any two of the determinations 900, 1100, 1300, or 1900 detect a multiple pregnancy. In some aspects, the combiner 640 can weight the outputs of the determinations 900, 1100, 1300, or 1900 differently. For example, if a yes output by the determinations 900, 1100, 1300, or 1900 is assigned a value of 1 and a no output by the determinations 900, 1100, 1300, or 1900 is assigned a value of 0. The combiner 640 can require a total value of 2 to output a detection of multiple pregnancy. If the output of the determination 900, for example, is weighted by 1.5 weighting factor, then the yes value of the output by the determination 900 provides 1.5 value (1 x 1.5) out of the total value of 2 that is required. In still other aspects, the combiner may be a trained machine learning network that accepts the outputs of the determinations 900, 100, 1300, and 1900 as inputs, and generates a yes/no output indicating whether a multiple gestation is suspected. In some aspects, the combiner 640 is omitted.

[00114] The processor (e.g., processor 138 of Fig. 1, processor 116 of Fig. 1, and/or other processors) provide, to a display (e.g., display 132 of Fig. 1 and/or other displays) in communication therewith, an output representative of the determination of whether the pregnancy is a multiple gestation/pregnancy. For example, the output of the combiner 640, the output of the multiple pregnancy/gestation determination 630, and/or the outputs of the individual determinations 900, 1100, 1300, or 1900, are presented as a visual representation 430 (e.g., a screen display on the display 132 of Fig. 1). The visual representation 430 can be a binary indicator of whether or not a multiple pregnancy is detected. The visual representation 430 can include an indication 650 (e.g., text or symbols) of whether a multiple pregnancy/gestation is suspected or not suspected. The screen display 430 may also include a graphical representation 660 associated with the determination of whether a multiple pregnancy/gestation is suspected. The graphical representation 660 may for example include drawings, ultrasound image frames, cineloops, or generated graphics, either with or without text or symbols as annotations, including graphics, images, text, and/or other visual representations described herein (e.g., the output of the combiner 640, the output of the multiple pregnancy/gestation determination 630, and/or the outputs of the individual determinations 900, 1100, 1300, or 1900).

[00115] Figure 7A is a schematic, diagrammatic overview, in block diagram form, of a training mode 700 for an untrained neural network 710a, according to aspects of the present disclosure. In the example shown in Figure 7A, a set of training data 705a includes ultrasound cineloops of probe sweeps annotated with the corresponding anatomy (head, heart, abdomen, pelvis, membrane, placenta, etc.). The training data 705a is fed into an untrained neural network 710a in an iterative training process that will be familiar to a person of ordinary skill in the art.

[00116] The parameters of a network model (e.g., the weights at each artificial neuron) are initialized with initial values A that may be random values or with results from training on prior datasets. In an iterative process, the network is used to make detection inferences on the training images, the results are compared with the ground truth annotations, and an optimizer is used to adjust the network parameters B until a metric of accuracy is maximized.

[00117] Thus, an output of this training process 700 is a trained neural network 710b, wherein the parameters B (e.g., weights) are optimized for generating accurate bounding boxes for the anatomy imaged in the training data 705a.

[00118] Figure 7B is a schematic, diagrammatic overview, in block diagram form, of an inference mode or clinical usage mode 704 for the trained neural network 710b, according to aspects of the present disclosure. In clinical usage, an ultrasound video, cineloop, or cine sweep 720 of the blind sweep is fed to the trained and validated neural network 710b for analysis. The trained and validated neural network 710b then produces, as an output, anatomy detection bounding boxes 740 for each image (or the entire sweep). In some aspects, a confidence value can be determined as a normalized value in the range [0-1], where 0 indicates lowest confidence, and 1 indicates highest confidence that the detection is correct. [00119] Figure 8 is a schematic, diagrammatic illustration, in block diagram form, of the detection of anatomy (e.g., the head, heart, abdomen, pelvis, placenta, membrane, etc.), according to aspects of the present disclosure. A cineloop 810 comprising multiple frames 820 is fed into a trained object detector 830.

[00120] The object detector 830 may implement or include any suitable type of learning network. For example, in some aspects, the object detector 830 could include a neural network, such as a convolutional neural network (CNN). In addition, the convolutional neural network may additionally or alternatively be an encoder-decoder type network, or may utilize a backbone architecture based on other types of neural networks, such as an object detection network, classification network, etc. One example backbone network is the Darknet YOLO backbone, (e.g., Yolov3) which can be used for object detection. The CNN may for example include a set of N convolutional layers, where N may be any positive integer. Fully connected layers can be omitted when the CNN is a backbone. The CNN may also include max pooling layers and/or activation layers. Each convolutional layer may include a set of filters configured to extract features from an input (e.g., from a frame of the ultrasound video). The value N and the size of the filters may vary depending on the aspects. In some instances, the convolutional layers may utilize any non-linear activation function, such as for example a leaky rectified non-linear (ReLU) activation function and/or batch normalization. The max pooling layers gradually shrink the high-dimensional output to a dimension of the desired result (e.g., bounding boxes of a detected feature). Outputs of detection network may include numerous bounding boxes, with most having very low confidence scores and thus being filtered out or ignored. Fully connected layers may be referred to as perception or perceptive layers. In some aspects, perception/perceptive and/or fully connected layers may be found in object detector 830 (e.g., a multi-layer perceptron).

[00121] These descriptions are included for exemplary purposes; a person of ordinary skill in the art will appreciate that other types of learning models, with features similar to or dissimilar to those described above, may be used instead or in addition, without departing from the spirit of the present disclosure.

[00122] Outputs of the object detector 830 may include an annotated cineloop 840 made up of a plurality of annotated image frames 842, possibly including per-frame metrics 845 such as the confidence level of the detections.

[00123] The systems and methods disclosed herein are broadly applicable to different types of features, and can for example draw boxes around the head, heart, placenta, or other anatomical features depending on the implementation. The object detector can be one class or multi-class, depending how the model is built. If another detector is trained separately, then both models can be run separately (e.g., one model for each feature type). Otherwise, multiple feature classes can be identified, and enclosed in detection boxes, at the same time. In an example, the ML model for placenta detection can use exactly the same structure as a model for heart detection. One can either train/run a single detector that detects multiple feature types (a multi-class detector) and provides their locations as an output, along with the confidence score and feature type (class) of each detection. Alternatively, one could run several single-class detectors, each trained to detect a single feature type/class. These separate single-class detectors may have the same architecture (e.g., layers and connections), but would have been trained with different data (e.g., different images and/or annotations) and thus have different weights.

[00124] Figure 9 is a schematic, diagrammatic representation, in flow diagram form, of an example inter-twin membrane detection method 900, according to aspects of the present disclosure. It is understood that the steps of method 900 may be performed in a different order than shown in Figure 9, additional steps can be provided before, during, and after the steps, and/or some of the steps described can be replaced or eliminated in other aspects. One or more of steps of the method 900 can be carried by one or more devices and/or systems described herein, such as components of the ultrasound imaging system 100, ultrasound blind sweep multiple pregnancy detection system 400, and/or processor circuit 250.

[00125] In step 920, the method 900 includes receiving an ultrasound image frame 910 and detecting an inter-twin membrane using an object detector (e.g., a machine-learning-based object detector as described above). Execution then proceeds to steps 930 and 940.

[00126] In step 930, the method 900 includes determining that no inter-twin membrane was detected in the image frame. Execution then returns to step 910 until all image frames have been analyzed, and then proceeds to step 950.

[00127] In step 940, the method 900 includes drawing an inter-twin bounding membrane bounding box on the image frame, indicating the most probable boundaries of where the inter-twin membrane is detected. Execution then returns to step 910 until all image frames have been analyzed, and then proceeds to step 950

[00128] In step 950, the method 900 includes determining, from multiple sweeps of the blind sweep protocol, the number or percentage of frames that include an inter-twin membrane detection. Execution then proceeds to step 960.

[00129] In step 960, the method 900 includes determining whether the number or percentage from step 950 exceeds a threshold value 970. If yes, execution then proceeds to step 980. If no, execution proceeds to step 990. [00130] In step 980, the method 900 includes determining that a multiple pregnancy/gestation is suspected. The method 900 is now complete.

[00131] In step 990, the method 900 includes determining that a multiple pregnancy/gestation is ruled out or not suspected. The method 900 is now complete.

[00132] Figure 10 is an ultrasound image frame 1000 that includes an inter-twin membrane bounding box 1010, indicating that an inter-twin membrane 350 has been detected, according to aspects of the present disclosure. Also visible are two fetal heads 1030 and two placentas 240 on either side of the membrane 350. The detection of an inter-twin membrane 350 (e.g., in a number of frames that exceeds a threshold number) is indicative of a twin pregnancy, as described above in Figure 9. The presence of two fetal heads is also indicative of a twin pregnancy, as described below.

[00133] Figure 11 is a schematic, diagrammatic representation, in hybrid flow diagram and block diagram form, of an example method 1100 for detection of multiple co-existing fetal parts within an image frame, within a sweep, or within all of the sweeps of a blind sweep protocol, according to aspects of the present disclosure.

[00134] In step 1110, the method 1100 includes receiving an ultrasound image frame 1105, and detecting fetal body parts using an object detector (e.g., a trained neural network as described above, although other types of detector could be used instead or in addition). Execution then proceeds to steps 1115 and 1120.

[00135] In step 1115, the method 1100 includes placing bounding boxes around all detected fetal body parts. Execution then proceeds to steps 1125 and/or 1130.

[00136] In step 1120, the method 1100 includes determining that no fetal body parts were detected in the image. Execution then proceeds to step 1140.

[00137] In step 1125, the method 1100 includes using a statistical model (e.g., naive Bayes, one-class SVM, Gaussian Mixture Model (GMM), etc.) to analyze the image frame and the bounding boxes, and determine the statistical likelihood of the detected anatomy being found in a single frame, for a singleton pregnancy.

[00138] An example implementation of the statistical model is as follows: A numeric vector V of size IxN [N=Number of classes] is constructed for each frame with the value of the vector V at index “idx” being the number of non-overlapping boxes of class corresponding to “idx” detected by 1110 for that frame. A GMM is trained to learn the distribution of these vectors from singleton data. During inference/runtime, a similar vector is constructed for every frame and is fed to the trained GMM model which would return the likelihood of that frame coming from a singleton exam which is then used to determine if frame is indicative of MG or not. This is then repeated for all frames.

[00139] In an example, if both coexisting structures are unique enough to be suspected as MG, then via 1130 step (rule-based) this can be separable. However, a classifier can be used instead of or in addition to a rule-based approach. This classifier can self-learn these defined coexistences of structures for multiple gestation/pregnancy (as in the rule-based approach). Coexistence of structures coming from the same entity (e.g., fetus) should be classified as singleton.

[00140] If the analysis indicates a probable multiple gestation, execution proceeds to step

1135. If the analysis does not indicate multiple gestation, execution proceeds to step 1140. [00141] In step 1130, the method 1100 includes using a rule-based expert system to analyze the image frame and the bounding boxes. The rule-based expert system may for example include classical image detection and/or optical character recognition algorithms to identify the bounding boxes and/or their labels, coupled with one or more rules. The rules may for example be Boolean expressions indicating the presence, absence, or multiple occurrence of certain features, and/or the implications thereof.

[00142] For example, if two non-overlapping heart detection boxes exist in the same frame, that frame may be identified as indicating multiple gestation. Another example would be the presence of one axial head and another axial abdomen in same frame. Another example would be the presence of two non-overlapping heads. Another example would be the presence of a head with a pelvic bone/fetal urinary bladder. Other examples include 1) two heads , hearts, breech areas (same anatomies) together in a frame, 2) one axial head and one axial abdomen in a frame, 3) one axial head and one axial thorax in a frame, 4) one axial head, one axial abdomen in a frame, 5) one axial head one breech region identified by pelvic bones in a frame, or 6) any two axial same or different (non-limb) anatomies appearing in the same frame irrespective of location/direction. The above provided examples are not exhaustive of the set of possible rules for distinguishing multiple gestation.

[00143] If the analysis of the image frame indicates multiple gestation, execution proceeds to step 1135. If the analysis does not indicate multiple gestation, execution proceeds to step 1140.

[00144] In step 1135, the method 1100 includes determining that the image frame contains features indicative of multiple gestation. Execution then returns to step 1110 until all image frames have been analyzed, and then proceeds to step 1150. [00145] In step 1110, the method 1100 includes determining that the image frame does not contain features indicative of multiple gestation. Execution then returns to step 1110 until all image frames have been analyzed, and then proceeds to step 1150.

[00146] In step 1150, the method 1100 includes determining a number or percentage of image frames from the blind sweep protocol that are indicative of multiple gestation. Execution then proceeds to step 1160.

[00147] In step 1110, the method 1100 includes determining whether the number or percentage from step 1150 exceeds a threshold 1170. If yes, execution then proceeds to step 1180. If no, execution then proceeds to step 1190.

[00148] In step 1180, the method 1100 includes determining that a multiple gestation is suspected. The method 1100 is now complete.

[00149] In step 1190, the method 1100 includes determining that a multiple gestation is not suspected. The method 1100 is now complete.

[00150] Figure 12 is an ultrasound image frame 1200 that includes a single placenta bounding box 1210 and two separate, non-overlapping head bounding boxes 1220, indicating that two fetal heads have been detected, according to aspects of the present disclosure. The presence of two fetal heads is indicative of a twin pregnancy, as described above in Figure 11. Seeing two of a unique feature (e.g., one fetus only has one such anatomy, such as one fetus has only one head, one heart, one spine, one urinary bladder, one abdomen, one stomach, etc.) in the same image (e.g., two heads, two hearts, two spines, two stomachs, two abdomens, etc.) may be indicative of a multiple gestation. However, it is noted that other combinations of features may also indicate multiple gestation, if it would be generally impossible for a single fetus to show the combination in a single image. For example, a foot located immediately adjacent to a head may be indicative of multiple gestation. Another example would be the presence of one axial head and another axial abdomen in same frame. Another example would be the presence of one axial head and another axial thorax in same frame. Another example would be the presence of one axial head and a fetal urinary bladder/pelvic bone in same frame. Other examples include 1) breech area and thorax (heart), 2) one axial head and one axial abdomen in a frame 3) one axial head and one axial thorax in a frame, 4) one axial head, one axial abdomen in a frame, 5) one axial head and one breech region identified by pelvic bones in a frame. The above examples are not exhaustive. Depending on the implementation, image frames such as the image frame 1200, including a plurality of bounding boxes identifying a plurality of detected fetal anatomical parts, may be output to the user. The presence of a single placenta bounding box 1210 in this image may neither indicate nor exclude a multiple gestation. In some instances, the placenta is maternal fetal part (e.g., not a fetal body part).

[00151] Figure 13 is a schematic, diagrammatic representation, in hybrid block diagram / flow diagram form, of an example fetal parts geometric mapping method 1300, according to aspects of the present disclosure.

[00152] In step 1315, the method 1300 includes receiving image frames 1305 from one sweep (e.g., an intra sweep cineloop) and/or image frames 1310 from multiple sweeps (e.g., a plurality of inter-sweep cineloops), and detecting fetal parts using an object detector (e.g., a trained machine learning network). Steps of the method 1300 can be repeated for different sweeps, if the images frame 1305 (which are from only one sweep) are used. Outputs of step 1315 include image frames 1320 with fetal parts bounding boxes and image frames 1325 without fetal parts bounding boxes (e.g., image frames where no fetal parts were detected). Execution then proceeds to step 1330.

[00153] In step 1330, the method 1300 includes spatial mapping of the fetal parts bounding boxes. Outputs of step 1330 include fetal parts clusters 1335 in the spatial mapping, as described below, and the distances 1340 between the fetal parts clusters, as described below. Execution then proceeds to step 1345.

[00154] In step 1345, the outputs 1335, 1340 of the spatial mapping step 1330 are received by an artificial intelligence such as a rule-based expert system as described above. An example rule would be the presence of two clusters of head separated spatially (some near the beginning and another group near the end of the sweep - this is an intra sweep impossibility in a singleton). Another example would be the presence of two clusters of head separated spatially across sweeps (one cluster near Cl sweep and the other near C5 sweep - this is an inter sweep impossibility in a singleton). Similar examples can be constructed for heart, abdomen, fetal urinary bladder etc. which shouldn’t occur in spatially separate regions of the womb in case of a singleton.

[00155] In some aspects, instead or in addition, the artificial intelligence may be or include a trained neural network as described above. Based on the fetal part clusters 1335 and distances 1340, the artificial intelligence outputs either an indication of multiple gestation 1350 or an indication of not multiple gestation 1355. The method 1300 is now complete.

[00156] Figure 14 is a schematic, diagrammatic view of a scanning process 1400, according to aspects of the present disclosure. The scanning process 1400 includes horizontal scans or sweeps 1410 labeled Cl, C2, and C3, as well as vertical sweeps 1420, labeled R, M, and L. The horizontal sweeps 1410 and vertical sweeps 1420 occur within a sweep area 1430 on the mother’s abdomen 1440, bounded by the upper abdomen 1450 and the cervix 1460, thus forming a sweep grid 1470. The spatial mapping is such that, for example, of a fetal part such as a head is detected at the lower portion of the L vertical sweep, it may also be expected to occur at the left portion of the Cl horizontal sweep.

[00157] Figure 15A is a graphical representation of a horizontal scan or horizontal sweep 1500, according to aspects of the present disclosure. The horizontal sweep includes multiple ultrasound images 1510, some of which are images 1520 marked with a first color or pattern indicating no detections, and some of which are images 1530 marked with a second color or pattern indicating that a fetal part such as the head is detected. Detection of two separate clusters of images 1550 containing a fetal body part, separated by a sequence of images 1520 where the fetal body part is not detected, may be indicative of multiple gestation.

[00158] Figure 15B is a graphical representation 1520 of a distance calculation for a horizontal sweep, according to aspects of the present disclosure. The graphical representation 1520 includes a Y-axis 1530 indicating the field of view of the ultrasound probe (e.g., measured in degrees, centimeters, or otherwise) and an X-axis 1540 indicating the length of the sweep (e.g., in frames, centimeters, or otherwise). Visible in the graphical representation is a first detection cluster 1550 (e.g., a cluster of frames where a fetal body part such as a head is detected), a second detection cluster 1560 (e.g., a second a cluster of frames where a fetal body part such as a head is detected), and a distance 1570 between the clusters (e.g., measured in the units of the X-axis). Also visible is a color key 1580 indicating the probability level of the detections.

[00159] Figure 16A is a graphical representation of a vertical scan or vertical sweep 1600, according to aspects of the present disclosure. The vertical sweep includes multiple ultrasound images 1610, some of which are images 1620 marked with a first color or pattern indicating no detections, some of which are images 1630 marked with a second color or pattern indicating that a fetal part such as an abdomen is detected, and some of which are images 1640 marked with a third color or pattern indicating that two of the same fetal part, such as two abdomens, are detected. Detection of two of the same body part in the same images 1640 may be indicative of multiple gestation.

[00160] Figure 16B is a graphical representation 1650 of a distance calculation for a vertical sweep, according to aspects of the present disclosure. The graphical representation 1650 includes an X-axis 1660 indicating the field of view of the ultrasound probe (e.g., measured in degrees, centimeters, or otherwise) and a y-axis 1670 indicating the length of the sweep (e.g., in frames, centimeters, or otherwise). Visible in the graphical representation is a first detection cluster 1680 (e.g., a cluster of frames where a fetal body part such as a head is detected), a second detection cluster 1690 (e.g., a second a cluster of frames where a fetal body part such as a head is detected), and an overlap region 1685 between the clusters (e.g., measured in the units of the X-axis). The existence of an overlap region 1685 indicates that the distance between the two clusters 1680 and 1690 is zero. Also visible is a color key 1692 indicating the detection probability level for a single fetal abdomen detection, and a color key 1694 indicating the detection probability for a dual fetal abdomen detection.

[00161] Figure 17 is a screen display 1700 of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure. The screen display includes a representation 1710 of the horizontal sweeps, a representation 1720 of the vertical sweeps, a combined horizontal and vertical representation 1730, and a combined representation 1740 overlaid on a graphical representation 1750 of the two fetuses. In the example shown in Figure 17, the horizontal sweeps representation 1710 includes two detection clusters for a fetal head 1030 and two detection clusters for a fetal heart 1760. Similarly, the vertical sweeps representation 1720 includes two detection clusters for a fetal head 1030 and two detection clusters for a fetal heart 1070.

[00162] The combined representation 1730 includes the same detection clusters as both the vertical sweeps representation 1710 and the horizontal sweeps representation 1720, resulting in two head detection overlap regions 1760 and two heart detection overlap regions 1770. Also visible are color keys 1580 and 1780 indicating the detection probabilities for the fetal head and heart, respectively. As can be seen in the combined representations 1730 and 1740, the overlapping detection regions 1760 and 1770 are indicated at a higher probability than the single detection regions 1030 and 1760. In other words, the head is very likely to be found in the overlap region 1760, and the heart is very likely to be found in the overlap region 1770.

[00163] In the combined representation 1740 with graphical representation 1750 of the two fetuses, the overlap regions 1760 each coincide with the head of one fetus, and the overlap regions 1770 each coincide with the heart of one fetus. Thus, the graphical display 1700, or portions thereof such as the combined representation 1730 or combined representation 1740, can thus be displayed to a user to indicate the presence of two fetuses within the uterus. Depending on the implementation, alternative indications may be provided instead or in addition, including but not limited to text, symbols, colors, audible tones, and/or spoken words.

[00164] Figure 18 is a spatial mapping screen display 1800 of an example ultrasound blind sweep multiple pregnancy detection system, according to aspects of the present disclosure. Visible are head detection overlap regions 1760A and 1760B, and heart detection overlap regions 1770A and 1770B. Also visible are distances 1810, including for example the distance 1820 between head cluster 1760A and heart cluster 1770A, the distance 1825 between head cluster 1760A and heart cluster 1770B, the distance 1830 between head cluster 1760B and heart cluster 1770B, the distance 1835 between heart cluster 1770B and head cluster 1760B, the distance 1840 between heart cluster 1770A and heart cluster 1770B, and the distance 1850 between head cluster 1760A and head cluster 1760B.

[00165] The distance between two detected features can be used by an expert system or trained machine learning model to determine the likelihood that the two features belong to a single fetus vs. two different fetuses. For example, two heart clusters that are located very close together may represent two detections of the same heart, and may therefore belong to a single fetus (or at least not be determinative of the presence of two fetuses). However, two heart clusters that are greater than a threshold distance apart may be considered unlikely to belong to the same fetus. Similarly, a heart cluster and a head cluster from a single fetus may be expected to fall within a particular range of distances from one another. A heart and head outside of this distance range (e.g., too close or too far apart), may be unlikely to be part of the same fetus, and may thus be indicative of two or more fetuses located within the uterus. [00166] Depending on the implementation, distances may be calculated between clusters, between clusters and overlap regions, or between overlap regions only.

[00167] Figure 19 is a schematic, diagrammatic representation, in hybrid block diagram / flow diagram form, of an example fetal parts geometric sequencing method 1900, according to aspects of the present disclosure.

[00168] In step 1930, the method 1900 includes receiving image frames 1910 from one sweep (e.g., an intra sweep cineloop) and/or image frames 1920 from multiple sweeps (e.g., a plurality of inter-sweep cineloops), and detecting fetal parts using an object detector (e.g., a trained machine learning network). Steps of the method 1900 can be repeated for different sweeps, if the images frame 1910 (which are from only one sweep) are used. Outputs of step 1930 include image frames 1940 with fetal parts bounding boxes and tags, and image frames 1950 without fetal parts bounding boxes and tags (e.g., image frames where no fetal parts were detected). Execution then proceeds to step 1960.

[00169] In step 1960, the method 1900 includes, based text recognition of the tags, assembling a text sequence of the anatomy tags from the image frames 1940, in the order of the tags’ occurrence in the sweep. The text sequence may for example be null for frames where no anatomy was detected, and may include one or multiple tags for frames where anatomy was detected. The frames may be separated in the text sequence by a separator/delimiting character such as a comma, semicolon, colon, etc. The text sequence may for example take a form such as “[frame 1 anatomy tags]; [frame 2 anatomy tags]; . . . .”. One example of this is “heart, head; heart, head; head; head; ; ; ; ; ; ; abdomen; abdomen; . . . ”, or “head/head; head/head; head; ; ; heart; heart/heart; heart/heart; heart; ; ;. although other forms and other sequences are possible and fall within the scope of the present disclosure. The blank spaces between the separating/delimiting character can be image frames in which there is no anatomy detected by the anatomy detector Al (e.g., CNN or other deep learning network). Sequences from multiple sweeps can be appended by using other delimiting characters to create a sequence for the entire exam. Another example for combining sequences from multiple sweeps would be to create a new sequence whose 1^st entry is the 1^st entry for Cl, 2^nd entry would be the 1^st entry for C2 etc. although other approaches to combining sequences across sweeps are possible and fall within the scope of the present disclosure.

[00170] The text sequences are then received by a sequence model 1970, which may for example be an expert system or trained machine learning model (e.g., hidden Markov, Recurrent Neural Network, etc.) that detects (a) whether multiple occurrences of the same anatomy occur within a single frame, and/or (b) whether the text sequence represents an improbable arrangement of anatomical features for a single fetus (e.g., two detections of a heart, separated by several null frames), and/or (c) whether the text sequences of multiple sweeps collectively represent an improbable arrangement of anatomical features for a single fetus (e.g., a heart detected in the R sweep and the L sweep, but not in the M sweep). An example of how the sequence model can help in detecting multiple gestation is by assigning a low score to sequences such as “head;head;head;head;;;;;;;;;;head;head;head;head” which implicitly indicates spatially separate head clusters occurring in the exam which is impossible for a normal singleton patient. Similar examples can be constructed for other unique fetal anatomies. The sequence model can learn other such patterns empirically that may not be obvious to humans. In such cases, the text sequence may be deemed to indicate the presence of multiple fetuses within the uterus, whereas if neither condition is true, the sequence may be deemed not to indicate the presence of multiple fetuses. Thus, the output of the sequence model 1970 is either an indication 1980 of multiple gestation or an indication 1990 of not multiple gestation.

[00171] Figure 20 is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method 2000, according to aspects of the present disclosure. In the example shown in Figure 20, each sweep 2005, 2015, 2025, 2035, 2045, and 2055 includes bounding boxes and their associated tags that have been added by the object detector. The tags are then stripped and delimited per frame (e.g., delimited by a separator character such as a semicolon after each frame), with frames having no detections being indicated by a null frame, a space, the word ’’null”, or otherwise, to assemble a corresponding text sequence of anatomy tags 2010, 2020, 2030, 2040, 2050, 2060 representing the sweeps.

[00172] Figure 21A is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method 2100, according to aspects of the present disclosure. In the example shown in Figure 21A, each horizontal sweep 2005, 2015, 2025, includes bounding boxes and their associated tags that have been added by the object detector. The tags are then stripped and delimited per frame as described above, to assemble a corresponding single text sequence of anatomy tags 2110 that encodes the object detections of all three horizontal sweeps in the order in which they were captured.

[00173] Figure 21B is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method 2120, according to aspects of the present disclosure. In the example shown in Figure 2 IB, each vertical sweep 2035, 2045, 2055, includes bounding boxes and their associated tags that have been added by the object detector. The tags are then stripped and delimited per frame as described above, to assemble a corresponding single text sequence of anatomy tags 2110 that encodes the object detections of all three vertical sweeps, e.g., in the order in which they were captured.

[00174] Figure 22 is a is a schematic, diagrammatic representation, in block diagram form, of an example fetal parts geometric sequencing method 2120, according to aspects of the present disclosure. In the example shown in Figure 22, each sweep 2005, 2015, 2025, 2035, 2045, and 2055 includes bounding boxes and their associated tags that have been added by the object detector. The tags are then stripped and delimited per frame as described above, to assemble a corresponding single text sequence of anatomy tags 2110 that encodes the object detections of all three horizontal sweeps and all three vertical sweeps, e.g., in the order in which they were captured.

[00175] Other orders for the anatomy tags in the text sequences associated with Figs. 20, 21 A, 21B, and/or 22 may be used instead or in addition.

[00176] In some instances, when multiple sweeps are part of a given text sequence (as in Figs. 21A, 21B, and 22), the anatomy tags from the multiple sweeps can ordered differently than the order in which the sweeps were physically performed by the ultrasound probe. For example, the Cl sweep may be physically performed before the C2 sweep in the blind sweep protocol (e.g., Fig. 4). The portion of the anatomy tags from the Cl sweep can be put before the portion of the anatomy tags from the C2 sweep in the text sequence (matching the order in which the sweeps were physically performed) or after the portion of the anatomy tags from the C2 sweep in the text sequence (different than the order in which the sweeps were physically performed).

[00177] For example, the order of the anatomy tags in the text sequence may or may not match the order/direction of the movement of the ultrasound probe for a given ultrasound blind sweep (as shown in, e.g., Fig. 4). For example, the Cl sweep may include the ultrasound probe starting from the patient’s right and moving to the patient’s left (in other instances, it is vice versa). The anatomy tags may be ordered/arranged in the text sequence matching the probe movement direction (tags from the patient’s right are relatively earlier in the text sequence and tags from the patient’s left are relatively later in the text sequence) or opposite to this movement direction (tags from the patient’s left are relatively earlier in the text sequence and tags from the patient’s right are relatively later in the text sequence). [00178] The same or different orders can be used for anatomy tags from multiple sweeps when multiple sweeps are part of a given text sequence (as in Figs. 21 A, 21B, and 22). For example, in Fig. 21 A, the anatomy tags for each of the Cl, C2, and C3 sweeps can match the probe movement direction. Then, the text sequence takes a form similar to “[Cl anatomy tags - patient right to patient left], [C2 anatomy tags - patient right to patient left], [C3 anatomy tags - patient right to patient left]”. This can similarly be done for the text sequence associated with Fig. 21B (e.g., “[R anatomy tags - patient bottom to patient top], [M anatomy tags - patient bottom to patient top], [L anatomy tags - patient bottom to patient top]”). This can similarly be done for the text sequence associated with Fig. 22 (“[R anatomy tags - patient bottom to patient top], [M anatomy tags - patient bottom to patient top], [L anatomy tags - patient bottom to patient top], “[Cl anatomy tags - patient right to patient left], [C2 anatomy tags - patient right to patient left], [C3 anatomy tags - patient right to patient left]”). [00179] In other instances, the order of the anatomy tags for, e.g., the C2 sweep can be switched relative to the order of the anatomy tags in the Cl and C3 sweeps. Then, the text sequence takes a form similar to “[Cl anatomy tags - patient right to patient left], [C2 anatomy tags - patient left to patient right], [C3 anatomy tags - patient right to patient left]”. This can be done to try to correlate the physical location associated with the anatomy tags at the beginning and/or ending of the Cl, C2, and/or C3 portions of the text sequence. For example, when the order of the anatomy tags for the C2 sweep is switched, the physical location of the anatomy tags from the ending of the Cl sweep’s portion of the text sequence (at the patient’s left) is closer to and/or otherwise proximate to the location of anatomy tags from the beginning of the C2 sweep’s portion of the text sequence (on same side at the patient’s left, higher on the patient’s abdomen). In comparison, when the order of the anatomy tags for the C2 sweeps matches the order of the anatomy tags for the Cl and C3 sweeps, the physical location of the anatomy tags from the ending of the Cl sweep’s portion of the text sequence (at the patient’s left) is farther from the location of anatomy tags from the beginning of the C2 sweep’s portion of the text sequence (both at different side at the patient’s right and higher on the patient’s abdomen). Similarly, when the order of the anatomy tags for the C2 sweep is switched, the physical location of the anatomy tags from the ending of the C2 sweep’s portion of the text sequence (at the patient’s right) is closer to and/or otherwise proximate to the location of anatomy tags from the beginning of the C3 sweep’s portion of the text sequence (on same side at the patient’s right, higher on the patient’s abdomen). This can similarly be done for the text sequence associated with Fig. 21B which can take a form similar to “[R anatomy tags - patient bottom to patient top], [M anatomy tags - patient top to patient bottom], [L anatomy tags - patient bottom to patient top]”. This can similarly be done for the text sequence associated with Fig. 22. For example, a meandering/ snaking path can be taken for the order of the anatomy tags in Fig. 22 such that the text sequence has a form similar to “[R anatomy tags - patient bottom to patient top], [M anatomy tags - patient top to patient bottom], [L anatomy tags - patient bottom to patient top], [C3 anatomy tags - patient left to patient right], [C2 anatomy tags - patient right to patient left], [C3 anatomy tags - patient left to patient right]”.

[00180] As will be readily appreciated by those having ordinary skill in the art after becoming familiar with the teachings herein, the ultrasound blind sweep multiple pregnancy detection system advantageously permits untrained and minimally trained users to perform an ultrasound blind sweep protocol to gather anatomical images of high quality, including automated detection of potential health conditions such as multiple gestation (MG). This may result in higher accuracy and higher clinician trust in the results, while potentially improving health outcomes and/or decreasing the total cost of care. Potential benefits include detection of multiple pregnancies via blind sweeps performed by novice ultrasound users. The solution can be a quick initial check scan for a center with high volume ultrasound turnover to triage patients for a more detailed obstetric scan, and can provide or support referral of the subject diagnosed with multiple gestation for further diagnosis and management to a tertiary care center. Early detection of multiple gestation may be extremely helpful for follow-up and monitoring of the pregnancy.

[00181] The systems, methods, and devices described herein may be applicable in point of care and handheld ultrasound use cases such as with the Philips Lumify system. The ultrasound blind sweep multiple pregnancy detection system can be used for any handheld imaging applications, including but not limited to obstetrics and echocardiography. The ultrasound blind sweep multiple pregnancy detection system could be deployed on handheld mobile ultrasound devices, and on portable or cart-based ultrasound systems. The ultrasound blind sweep multiple pregnancy detection system can be used in a variety of settings including emergency departments, ambulances, accident sites, and homes. The applications could also be expanded to other settings.

[00182] The system is detectable from its functionality and output such as reporting of detected health conditions such as multiple pregnancy. This invention increases the value proposition of ultrasound applications in the obstetrics context, especially for use by minimally trained users.

[00183] Accordingly, the logical operations making up the aspects of the technology described herein are referred to variously as operations, steps, objects, layers, elements, components, algorithms, or modules. Furthermore, it should be understood that these may occur or be performed or arranged in any order, unless explicitly claimed otherwise or a specific order is inherently necessitated by the claim language.

[00184] All directional references e.g., upper, lower, inner, outer, upward, downward, left, right, lateral, front, back, top, bottom, above, below, vertical, horizontal, clockwise, counterclockwise, proximal, and distal are only used for identification purposes to aid the reader’s understanding of the claimed subject matter, and do not create limitations, particularly as to the position, orientation, or use of the ultrasound blind sweep multiple pregnancy detection system. Connection references, e.g., attached, coupled, connected, joined, or “in communication with” are to be construed broadly and may include intermediate members between a collection of elements and relative movement between elements unless otherwise indicated. As such, connection references do not necessarily imply that two elements are directly connected and in fixed relation to each other. The term “or” shall be interpreted to mean “and/or” rather than “exclusive or.” The word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. Unless otherwise noted in the claims, stated values shall be interpreted as illustrative only and shall not be taken to be limiting. [00185] The above specification, examples and data provide a complete description of the structure and use of exemplary aspects of the ultrasound blind sweep multiple pregnancy detection system as defined in the claims. Although various aspects of the claimed subject matter have been described above with a certain degree of particularity, or with reference to one or more individual aspects, those skilled in the art could make numerous alterations to the disclosed aspects without departing from the spirit or scope of the claimed subject matter. [00186] Still other aspects are contemplated. It is intended that all matter contained in the above description and shown in the accompanying drawings shall be interpreted as illustrative only of particular aspects and not limiting. Changes in detail or structure may be made without departing from the basic elements of the subject matter as defined in the following claims.

Claims

CLAIMS What is claimed is:

1. A system, comprising: a processor configured for communication with an ultrasound probe, wherein the processor is configured to: control the ultrasound probe to obtain a plurality of ultrasound image frames during a blind sweep protocol on a patient with a pregnancy; provide the plurality of ultrasound image frames as an input to a deep learning network trained to detect at least one of maternal anatomy or fetal anatomy; generate, as an output of the deep learning network, a detection of at least one of: an inter-twin membrane within the plurality of ultrasound image frames; or a plurality of fetal anatomical parts within the plurality of ultrasound image frames; determine, using the detection of at least one of the inter-twin membrane or the plurality of fetal anatomical parts, whether the pregnancy comprises a multiple gestation; and provide, to a display in communication with the processor, an output representative of the determination of whether the pregnancy comprises the multiple gestation.

2. The system of claim 1, wherein the processor is configured to perform the determination of whether the pregnancy comprises the multiple gestation based on at least one of: the detection of the inter-twin membrane; a co-existence of the plurality of fetal anatomical parts within a single ultrasound image frame; a geometric mapping of the plurality of fetal anatomical parts; or a geometric sequence of the plurality of fetal anatomical parts.

3. The system of claim 2, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation, the processor is configured to combine results of at least two of: the detection of the inter-twin membrane; the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame; the geometric mapping of the plurality of fetal anatomical parts; or the geometric sequence of the plurality of fetal anatomical parts.

4. The system of claim 2, wherein the output of the deep learning network comprises a plurality of detections of the inter-twin membrane in the plurality of ultrasound image frames, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the detection of the inter-twin membrane in the plurality of ultrasound image frames, the processor is configured to: determine a quantity of the plurality of ultrasound image frames with the detection of the inter-twin membrane; compare the quantity to a threshold quantity; and determine that the pregnancy comprises the multiple gestation when the quantity exceeds the threshold quantity.

5. The system of claim 4, wherein the output provided to the display comprises: at least one ultrasound image frame with the detection of the inter-twin membrane; and a bounding box identifying the inter-twin membrane overlaid on the at least one ultrasound image frame.

6. The system of claim 2, wherein the output of the deep learning network for the single ultrasound image frame comprises a plurality of detections of the plurality of fetal anatomical parts, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame, the processor is configured to: provide the plurality of detections as input to at least one of a statistical model or a rule-based expert system; and generate, as an output of at least one of the statistical model or the rule-based expert system, a determination that the single ultrasound image frame is indicative of the multiple gestation.

7. The system of claim 6, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame, the processor is configured to: repeat the determination that the single ultrasound image frame is indicative of the multiple gestation for the plurality of ultrasound image frames; determine a quantity of the plurality of ultrasound image frames indicative of the multiple gestation; compare the quantity to a threshold quantity; and determine that the pregnancy comprises the multiple gestation when the quantity exceeds the threshold quantity.

8. The system of claim 7, wherein the output provided to the display comprises: the single ultrasound image frame; and a plurality of bounding boxes identifying the plurality of fetal anatomical parts overlaid on the single ultrasound image frame.

9. The system of claim 2, wherein the output of the deep learning network comprises a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts, the processor is configured to: generate a spatial mapping of the plurality of fetal anatomical parts; and perform the determination that the pregnancy comprises the multiple gestation using a rule-based expert system and the spatial mapping.

10. The system of claim 9, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts, the processor is configured to: detect a plurality of regions with the plurality of fetal anatomical parts in the spatial mapping; determine at least one distance between the plurality of regions; provide, as input to the rule-based expert system, the plurality of regions and the at least one distance; and generate, as an output of the rule-based expert system, the determination that the pregnancy comprises the multiple gestation.

11. The system of claim 9, wherein the output provided to the display comprises: the spatial mapping.

12. The system of claim 2, wherein the output of the deep learning network comprises a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, wherein, to perform the determination of whether the pregnancy comprises the multiple gestation based on the geometric sequence of the plurality of fetal anatomical parts, the processor is configured to: generate a text sequence of anatomy tags representative of the plurality of fetal anatomical parts and the plurality of ultrasound image frames; provide the text sequence as an input to a sequence model; and generate, as an output of the sequence model, the determination that the pregnancy comprises the multiple gestation.

13. The system of claim 12, wherein the plurality of ultrasound image frames comprises a first sweep and a second sweep of the blind sweep protocol, wherein a physical location associated with the anatomy tags from an ending of a first portion of the first sweep in the text sequence proximate to the physical location associated with the anatomy tags from the beginning of a second portion of the second sweep in the text sequence.

14. The system of claim 1, further comprising the ultrasound probe.

15. A method, comprising: controlling, with a processor, an ultrasound probe in communication with the processor to obtain a plurality of ultrasound image frames during a blind sweep protocol on a patient with a pregnancy; providing, with the processor, the plurality of ultrasound image frames as an input to a deep learning network trained to detect anatomy of the patient; generating, with the processor and as an output of the deep learning network, a detection of at least one of: an inter-twin membrane within the plurality of ultrasound image frames; or a plurality of fetal anatomical parts within the plurality of ultrasound image frames; determining, with the processor, whether the pregnancy comprises a multiple gestation, using the detection of at least one of the inter-twin membrane or the plurality of fetal anatomical parts; and providing, with the processor, an output representative of the determination of whether the pregnancy comprises the multiple gestation to a display in communication with the processor.

16. The method of claim 15, wherein determining whether the pregnancy comprises the multiple gestation is based on at least one of: the detection of the inter-twin membrane; a co-existence of the plurality of fetal anatomical parts within a single ultrasound image frame; a geometric mapping of the plurality of fetal anatomical parts; or a geometric sequence of the plurality of fetal anatomical parts.

17. The method of claim 16, wherein the output of the deep learning network comprises a plurality of detections of the inter-twin membrane in the plurality of ultrasound image frames, wherein determining whether the pregnancy comprises the multiple gestation based on the detection of the inter-twin membrane in the plurality of ultrasound image frames comprises: determining a quantity of the plurality of ultrasound image frames with the detection of the inter-twin membrane; comparing the quantity to a threshold quantity; and determining that the pregnancy comprises the multiple gestation when the quantity exceeds the threshold quantity.

18. The method of claim 16, wherein the output of the deep learning network for the single ultrasound image frame comprises a plurality of detections of the plurality of fetal anatomical parts, wherein determining whether the pregnancy comprises the multiple gestation based on the co-existence of the plurality of fetal anatomical parts within the single ultrasound image frame comprises: providing the plurality of detections as input to at least one of a statistical model or a rule-based expert system; and generating, as an output of at least one of the statistical model or the rulebased expert system, a determination that the single ultrasound image frame is indicative of the multiple gestation.

19. The method of claim 16, wherein the output of the deep learning network comprises a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, wherein determining whether the pregnancy comprises the multiple gestation based on the geometric mapping of the plurality of fetal anatomical parts comprises: generating a spatial mapping of the plurality of fetal anatomical parts; and performing the determination that the pregnancy comprises the multiple gestation using a rule-based expert system and the spatial mapping.

20. The method of claim 16, wherein the output of the deep learning network comprises a plurality of detections of the plurality of fetal anatomical parts in the plurality of ultrasound image frames, wherein determining whether the pregnancy comprises the multiple gestation based on the geometric sequence of the plurality of fetal anatomical parts comprises: generating a text sequence of anatomy tags representative of the plurality of fetal anatomical parts and the plurality of ultrasound image frames; providing the text sequence as an input to a sequence model; and generating, as an output of the sequence model, the determination that the pregnancy comprises the multiple gestation.