Method and apparatus for detecting simultaneously brain vascular activity and physiological parameters during ultrasound imaging

Abstract

The present disclosure proposes a method for monitoring simultaneously vascular activity in at least one region of a brain, of a human or animal, and the temporal evolution of physiological parameters, the method including generating a sequence of Doppler or Ultrasound Localization Microscopy "ULM" images of said brain with an ultrasound probe and simultaneously computing physiological information based on a temporal series of brain tissue motion data in a selected region of interest of the brain.

Description

DescriptionMethod and apparatus for detecting simultaneously brain vascular activity and physiological parameters during ultrasound imagingTechnical Field

[0001] This present disclosure concerns methods and apparatuses for detecting simultaneously brain vascular activity and physiological parameters during ultrasound imaging in a human or animal.Background Art

[0002] Imaging the brain activity in vivo provides useful information on how and where the brain is activated in response to a certain function or stimulus. More specifically, in case of traumatic brain injuries and aneurysm-related hemorrhages, it is crucial to monitor brain vascular activity for appropriate neurotically care.

[0003] Most brain-wide functional imaging modalities exploit neurovascular coupling to map brain activity in millimetric resolutions. More specifically, activated zones of the nervous system need more oxygen, thus locally increasing the flow of blood in the vascular network of said nervous system, in particular in the capillaries, venules and arterioles of the vascular network. Hence, the nervous activity in the nervous system may be estimated at a millimetric scale, based on the activity of the vascular network in said nervous system.

[0004] One type of functional imaging is Functional Ultrasound (fUS), which has been described in particular by Mace et al. [Mace, E. et al. Functional ultrasound for the brain. Nat. Methods, vol. 8, pp. 662-664, Aug. 2011], This method allows in-depth imaging of functional hemodynamic changes in the cerebral microvacularization. Applied to the brain, fUS allows to map brain activation at high spatial resolution (100 pm) and temporal resolution (200 ps). fUS imaging can be performed non-invasively in mice and through chronic transparent windows in rats or larger animal models. Functional Ultrasound can image both task-evoked and spontaneous activity of the brain. During spontaneous brain activity, the study of similarities in the slow changes of cerebral blood volume (CBV) signals in different voxels or regions of the brain permits to build dynamic functional connectivity matrices providing unique information on the functional interactions between different brain regions [Osmanksi B. et al, Functional Ultrasound imaging of intrinisic connectivity in the living rat brain with high spatiotemporal resolution, Nature Comms, 5, 5023, 2014]]

[0005] Another type of imaging of high interest, particularly regarding spatial resolution and penetration depth, is Ultrasound Localization Microscopy (ULM), which has been described in particular by Errico et al. [Errico, C. et al. Ultrafast ultrasound localization microscopy fordeep super-resolution vascular imaging. Nature 527, 499-+ (2015)] and Demene et al. [Demene, C. et al. Transcranial ultrafast ultrasound localization microscopy of brain vasculature in patients. Nat. Biomed. Eng 5, 219-228 (2021)]. Ultrasound Localization Microscopy achieves transcranial imaging of cerebrovascular flow, up to micron scale, by localizing millions of intravenously injected microbubbles (MB). Thus, fUS combined with intravenous microbubbles injections allows to reveal the vascular network and neurovascular response down to the microscopic range.

[0006] Recently, the large increase in acquisition frame rate from 50 Hz in conventional imaging to 1000 Hz in ultrafast Ultrasound, by emitting ultrasonic plane waves, has enabled tracking of tiny tissue motions induced by fast transient mechanical vibrations, such as shear waves generated via the radiation force of ultrasonic focused beams. The ultrafast ultrasound imaging can be also used to image the propagation of natural mechanical waves, such as the arterial pulse wave or muscular contractions. Specifically, natural shear waves induced by cardiac pulsatility generate a physiological mechanical noise, typically ranging between 1 and 50 pm local tissue displacement, that can be exploited to map the mechanical properties of organs. In Ultrasound localization microscopy, this estimation of pixel-sized drift from ultrasonic speckle tracking is also applied to improve spatial resolution and correct motion artifacts. When the heart of a patient is examined, the local tissue displacement or ultrasonic speckle tracking can be used to extract the cardiac pulsatility on short time scales to synchronize ultrasound frames with the cardiac cycle and extract systolic and diastolic vascular properties.

[0007] Recent studies have shown that neuronal activity is highly synchronized with physiological parameters such as breathing rate and heart rate. Indeed, the cardiac rate of animals is correlated to a wide range of global biological processes ranging from stress, pain to sleep. On the other hand, breathing has been shown to be a fundamental rhythm of brain function with growing evidence of strong synchronization between brain activity and respiratory rate. Breathing is also implicated in sensorial rhythms and attention modulation. Therefore, these rhythms have been monitored and registered during in vivo experiments to obtain quantitative information of the global health status of the animal as well as to gain a deeper insight into the connection between neuronal processes and cardiac and respiratory rhythms.

[0008] To date, unfortunately, in pre-clinal functional ultrasound experimental setups, the recording of physiological parameter in subjects, for example in small animals, requires the use of implanted electrodes for assessing breathing and heart rates. It can be used to monitor the depth of anesthesia and study the influence of drugs. The physiological signal can be used as an external trigger signal to synchronize the ultrasound imaging sequencesto the moment of the patient’s heart rates. However, these invasive sensors yield strong limitations; they are only compatible with anaesthetized rodents, they increase the number of probes and cables used for the experiment, and in some cases, the electrode needles can disturb the rodent and affect the functional response. In some situations, such as awake and / or behaving preclinical experiments, it is not possible to detect these parameters using electrodes.

[0009] Therefore, there exists a need for reliably imaging brain vascular activity in a human or animal, dynamically at a microscopic scale, and simultaneously detecting and quantifying physiological parameters.

[0010] One aim of the present disclosure is to provide a method to image the brain vascular activity without the use of an additional setup while having the possibility to monitor the heart and breathing rates.

[0011] Another aim of the present disclosure is to provide a method allowing to identify automatically optimal brain regions where average tissue motion reflects underlying physiological parameters.

[0012] Another aim of the present disclosure is to provide a method allowing the extraction of the physiological rhythms and neuroimaging data simultaneously even though animals are awake.Summary

[0013] The present disclosure improves the situation.

[0014] To this end, the present disclosure proposes a method for imaging vascular activity in a brain of a living being, the method including:(a) transmitting unfocused ultrasound waves in the brain by an array ultrasound probe and acquiring raw data from backscattered ultrasonic waves by said array ultrasound probe;(b) generating a sequence of compound ultrasound images of the brain from said raw data during a recording period, said sequence of compound ultrasound images forming a sequence of successive raw-IQ (in-phase / quadrature) frames;(c) computing a temporal series of brain tissue motion based on said sequence of successive raw-IQ frames;(d) selecting at least one region of interest “ROI” of the brain in said temporal series of brain tissue motion for computing at least one physiological information;(e) computing said at least one physiological information based on the temporal series of brain tissue motion in the selected region of interest;(f) generating a sequence of Doppler or Ultrasound Localization Microscopy “ULM” images of said brain based on said sequence of successive raw-IQ frames.

[0015] In the prior art, tissue motion is regarded as an undesirable artifact: it is systematically removed from the raw ultrasound data by applying a clutter filter, typically based on singular value decomposition (SVD), prior to any further analysis. This step is intended to eliminate low-frequency components associated with global tissue motion (such as respiration or cardiac pulsatility of the tissue), so that only signals of interest from blood flow or microbubbles remain. Thus, tissue motion is considered as background noise to be eliminated.

[0016] In contrast, the present disclosure uses an approach in which tissue motion is not removed, but rather preserved and specifically analyzed. The temporal series of brain tissue motion, computed from the raw IQ frames, is used to optimally select regions of interest (ROIs) in which pulsatility (cardiac or respiratory) is most representative and reliable.

[0017] In other words, the computation of tissue motion is performed on the unfiltered, original raw-IQ data, before any clutter filter is applied. By analyzing the data prior to clutter filtering, the method preserves all components of tissue motion, including those due to physiological pulsatility, which are essential for accurate extraction of heart rate and breathing rate.

[0018] In contrast to the functional ultrasound (fUS) imaging, extraction of physiological parameters requires the determination of a localized region of the brain for the methods described herein.

[0019] Indeed pulsatility — arising from cardiac and respiratory cycles — does not affect all brain regions uniformly. The amplitude, periodicity, and signal-to-noise ratio of tissue motion signals can vary significantly depending on the anatomical location within the brain. As a result, an arbitrary or suboptimal selection of ROIs may lead to signals that are weak, noisy, or not representative of the underlying physiological rhythms, making it impossible or unreliable to accurately extract heart rate or breathing rate. By contrast, the ROI selection, based on spatiotemporal or frequency analysis of the tissue motion signal, enables an identification of those specific regions where the tissue motion signal exhibits clear, periodic patterns corresponding to physiological pulsatility, allowing an accurate extraction of physiological parameters.

[0020] In addition, whole-image based analysis of cardiac and respiratory pulsatility requires multiple repetitions of the physiological rhythms in order to differentiate between them, whereas analysis based on a region of interest enables extraction of the signal periodicity from a single cycle. Thus ROI based approach proposed in the present invention enables higher temporal resolution measurement of pulsatility rythms up to the physiological periodicity itself.

[0021] Furthermore, if the average variance of one rhythm across the entire image is much higher than that of the other, ROI selection enables the focus on areas where the weaker rhythm is more pronounced, thus preventing it from being masked by the dominant signal. For instance, in situations where breathing-induced motion exhibits a substantially higher amplitude than the cardiac signal, the displacement averaged over the whole image is predominantly determined by the breathing component.

[0022] By performing the measurement within a localized region, sensitivity to the subject motion artifacts is reduced, and computational demands are lowered, which enables realtime monitoring of heart rate and breathing rates.

[0023] The approach enables real-time extraction of physiological parameters directly from the same ultrasound data used for neuroimaging, without the need for additional sensors or invasive procedures.

[0024] Furthermore, since the physiological parameters are extracted from the same data stream as the vascular activity, the method allows for precise temporal correlation between neurovascular events and physiological rhythms, facilitating advanced studies of neurovascular coupling and brain state dynamics.

[0025] The present ROI-based analysis can be used as a monitoring tool to detect abnormal physiological parameters. This is particularly important for identifying patients or subjects in physiological condition outside physiological limits for example in the case of large variations of rhythm from baseline. Having two different ROIs for cardiac pulsatility and respiratory pulsatility measurement also enables to correctly measure and differentiate cardiac and respiratory pulsatility when both rhythms become close.

[0026] Therefore, the use of ROI and tissue motion that is considered as an artifact in the prior art provides robust tool to extract physiological information, enabling simultaneous, accurate, and non-invasive monitoring of both brain vascular activity and physiological parameters, which is not possible with the approaches disclosed in the prior art, even in challenging scenarios such as overlapping rhythms, high variance disparity, abnormal physiological conditions.

[0027] The following features, can be optionally implemented, separately or in combination one with the others:

[0028] The region of interest may be spatially defined on the initial image of the brain, that is used for functional ultrasound (fUS). For breathing-related regions of interest, the selected area may occupy less than 50% of the image, typically around 20% of the initial image. For heart rate analysis, the region of interest may be limited to a smaller area, oftenbetween 1 and 20 pixels of the initial image, corresponding to less than 1% of the initial image.

[0029] The step of selection of said at least one region of interest “ROI” may be performed automatically using the temporal periodicity of a signal representative of brain tissue motion.

[0030] Automated ROI selection reduces inter-operator variability and the need for manual trial-and-error, leading to more consistent and reproducible results across different subjects and imaging sessions.

[0031] The at least one physiological information may be chosen in a group comprising: breathing rate, heart rate.

[0032] The at least one region of interest may comprise a region of interest specific for computing the breathing rate or the heart rate or both.

[0033] In some embodiments, the step of selection of at least one region of interest “ROI” may comprise:- computing the spatio temporal decomposition of the temporal series of brain tissue motion into a base of spatial vectors and a base of temporal vectors;- identifying a temporal vector having a signal period corresponding to an expected physiological information;- selecting a region of interest based on the spatial vector corresponding to the identified temporal vector.

[0034] In one example, the step of identifying a temporal vector having a signal period corresponding to an expected physiological information may comprise:- computing the autocorrelation of the temporal vectors;- detecting temporal peaks in the autocorrelation by applying a peak detection algorithm with a periodicity range corresponding to a signal period representative of the physiological information.

[0035] Alternatively, the step of identifying a temporal vector having a signal period corresponding to an expected physiological information may comprise :- computing the periodicity of temporal vectors using Fourier transform;- detecting temporal peaks in the Fourier spectrum in a periodicity range representative of the physiological information.

[0036] In some embodiments, the computation of the decomposition of the temporal series of brain tissue motion may comprise a Singular Value Decomposition “SVD” or a Principal Component Analysis “PCA” or an Independent component analysis “ICA” or any other spatiotemporal decomposition method.

[0037] In another embodiment, the step of selection of at least one region of interest “ROI” may comprise:- selecting a periodicity range corresponding to the periodicity of at least one expected physiological information;- generating a map in which each pixel's intensity corresponds to the amplitude of the Fourier transform of the temporal series of brain tissue motion in the preselected periodicity range;- selecting a region of interest based on the pixel's intensity of the previous map.

[0038] In some embodiment, the periodicity range for the breathing rate is between 10 bpm and 80 bpm in clinical imaging, and between 50 and 200 bpm in preclinal imaging.

[0039] In some embodiments, the periodicity range for the heart rate is between 30 bpm and 180 bpm in clinical imaging, and between 150 and 650 bpm in preclinical imaging.

[0040] In some embodiments, the temporal series of brain tissue motion is computed by means of tissue speckle tracking or any other motion tracking technique applied to the Ultrasound data.

[0041] In some embodiments, the method may further comprise:(g) computing a temporal evolution of at least one vascular dynamic parameter in a region of the brain based on said Doppler images.

[0042] In one or more embodiments, said recording period of the sequence of ultrasound compound images including at least one external dynamic event, due to a cause other than cardiac pulsatility and / or breathing pulsality, the dynamic event activating a change in a local hemodynamics and / or a change in a structural conformation in a region of the brain, the method may further comprise:(g) computing a temporal evolution of at least one vascular dynamic parameter in a region of the brain in response to the dynamic event.

[0043] In one or more embodiments, the method may further comprise:(g) computing at least one vascular or neurovascular dynamic parameter in a region of the brain based on said ULM images, said step (g) of computing at least one dynamic parameter comprising a localization step and a tracking step of microbubbles or ultrasound contrast agents, in a patient that has been previously administrated with microbubbles or ultrasound contrast agents.

[0044] In some embodiments, the at least one dynamic parameter may be chosen in the group comprising: blood flow, blood velocity, blood volume, vessel diameter, brain region size and any combination thereof.

[0045] In one or more embodiments, at least one Doppler image may be registered on a brain atlas, a computed tomography scan of said brain, a magnetic resonance image scan of said brain and any combination thereof.

[0046] The present disclosure also concerns an apparatus for imaging vascular activity in a brain of a living being, said apparatus including at least an array ultrasonic probe and a control system configured to:(a) transmit unfocused ultrasound waves in the brain by an array ultrasound probe and acquiring raw data from backscattered ultrasonic waves by said array ultrasound probe;(b) generate a sequence of compound ultrasound images of the brain from said raw data during a recording period, said sequence of compound ultrasound images forming a sequence of successive raw-IQ (in-phase / quadrature) frames;(c) compute a temporal series of brain tissue motion based on said sequence of successive raw-IQ frames;(d) select at least one region of interest “ROI” in said temporal series of brain tissue motion for computing at least one physiological information;(e) compute the at least one physiological information based on the temporal series of brain tissue motion in the selected at least one region of interest of the brain;(f) generate a sequence of Doppler or Ultrasound Localization Microscopy “ULM” images of said brain based on said sequence of successive raw-IQ frames.

[0047] In embodiments of the apparatus, the control system may be configured to display Doppler or ULM image showing the brain activity and at the same time the value or temporal evolution of the breathing rate or / and the value or temporal evolution of the heart rate.

[0048] In another aspect, it is proposed a computer software comprising instructions to implement the method as defined here when the software is executed by a processor. In another aspect, it is proposed a computer-readable non-transient recording medium on which a software is registered to implement the method as defined here when the software is executed by a processor.Brief Description of Drawings

[0049] Other features, details and advantages will be shown in the following detailed description and on the figures, on which:Fig. 1

[0050] [Fig. 1] Figure 1 is a block diagram illustrating an embodiment of an apparatus according to the present disclosure.Fig. 2

[0051] [Fig. 2] Figure 2 is a flowchart of a method for monitoring simultaneously the temporal evolution of the brain activity and the temporal evolution of physiological parameters according to an embodiment of the present disclosure with the apparatus of Figure 1.Fig. 3

[0052] [Fig. 3] Figure 3 illustrates schematically in a brain large-scale periodic tissue motions created by the breathing in a first region of interest “ROI1” and localized periodic tissue motions created by the heartbeat in a second region of interest “ROI2”.Fig. 4

[0053] [Fig. 4A] Figure 4A shows an example of B-Mode images obtained in a coronal acquisition in mice.

[0054] [Fig. 4B] Figure 4B shows an example of Power-Doppler images obtained from the B-Mode images of Figure 4A overlaid with an atlas.

[0055] [Fig. 4C] Figure 4C shows an example of three cartographies of tissue displacement computed based on the B-Mode images of Figure 4A, respectively at time to, time to + 80 ms, to + 160 ms.Fig. 5

[0056] [Fig. 5A] Figure 5A shows spatial mean of tissue velocity computed in a first region of interest ROI1 that is manually selected in the cartographies of tissue displacement of figure 4C.

[0057] [Fig. 5B] Figure 5B shows spatial mean of tissue velocity computed in a second region of interest ROI2 that is manually selected in the cartographies of tissue displacement of figure 4C.

[0058] [Fig. 5C] Figure 5C shows average signal period of the tissue velocity in the first region of interest ROI1 .

[0059] [Fig. 5D] Figure 5D shows average signal period of the tissue velocity in the second region of interest ROI2.Fig. 6

[0060] [Fig. 6] Figure 6 shows a B-Mode image obtained by averaging over 200 IQ frames at 500 Hz on which 4 regions of interest ROI1 , ROI2, ROI3 and ROI4 have been manually selected.Fig. 7

[0061] [Fig. 7A] Figure 7A shows spatial average of tissue velocity computed in the first ROI1 and the periodicity of the signal by vertical dashed lines.

[0062] [Fig. 7B] Figure 7B shows spatial average of tissue velocity computed in the second ROI2 and the periodicity of the signal by vertical dashed lines.

[0063] [Fig. 7C] Figure 7C shows spatial average of tissue velocity computed in the third ROI3 and the periodicity of the signal by vertical dashed lines.

[0064] [Fig. 7D] Figure 7D shows spatial average of tissue velocity computed in the fourth ROI4 and the periodicity of the signal by vertical dashed lines.Fig. 8

[0065] [Fig. 8] Figure 8 illustrates an embodiment of the step of selecting of the region of interest for computing physiological parameters.Fig. 9

[0066] [Fig. 9] Figure 9 illustrates the Singular Value Decomposition of tissue motion to identify regions of the brain corresponding to heart and breathing rates: (A) a temporal series of ultrasound images; (B) six first more energetic spatial singular vectors; (C) six corresponding temporal singular vectors; (D) normalized autocorrelation of the six temporal singular vectors, two selected temporal vectors designed by a star (1stand 4th) to compute the breathing and heart rates; (E) first mask or first ROI constructed from first selected spatial singular vector for breathing rate extraction corresponding to the first temporal vector; (F) second mask or second ROI constructed from second spatial singular vector for heart rate extraction corresponding to the fourth temporal vector.Fig. 10

[0067] [Fig. 10] Figure 10 illustrates in detail the sub-step of selecting of the two temporal singular vectors for selecting the first region ROI corresponding to the breathing rate region and the second ROI corresponding to the heart rate region illustrated in Figure 9.Fig. 11

[0068] [Fig. 11] Figure 11 illustrates another embodiment of the step of selecting of the region of interest for computing physiological parameters.Fig. 12

[0069] [Fig. 12A] Figure 12A illustrates the comparison of heart and breathing rates computed based on a temporal series of the brain tissue motion according to an embodiment of the method with heart and breathing rates measured by an electrode setup in a mouse.

[0070] [Fig. 12B] Figure 12B illustrates the comparison of heart and breathing rates computed based on a temporal series of the brain tissue motion according to an embodiment of the method with heart and breathing rates measured by an electrode setup in a rat.Fig. 13

[0071] [Fig. 13] Figure 13 illustrates an example of screen of an ultrasound imaging apparatus of the present disclosure in which it is possible to display at same time the Doppler or ULM image showing the brain activity, the breathing rate ROI and heart rate ROI overlaid on the patient’s Atlas, the temporal evolution of the breathing rate and the temporal evolution of the heart rate.Description of Embodiments

[0072] In the Figures, the same references denote identical or similar elements.

[0073] The present disclosure proposes a method and apparatus for imaging vascular activity in at least one zone of a brain of a patient, human or animal, by obtaining a temporal series of two-dimensional 2D, or three-dimensional, 3D, Doppler images or a 2D or 3D ULM image of said brain and simultaneously a temporal evolution of physiological parameters of the patient based on a same sequence of ultrasound images. In the present disclosure, the physiological parameters may include the heart rate and the breathing rate. Therefore, it is possible to monitor simultaneously the neurofunctional activity and heart and breathing rates solely from ultrasound imaging of brain without using an additional device to measure the physiological parameters of the patient.

[0074] During brain ultrasound imaging, the heart motion and the breathing of the patient can induce brain tissue displacements. It is known that the cardiac pulse wave propagating through arteries induces localized brain tissue displacements whereas the breathing induces large-scale tissue displacement. It is customary to eliminate these motion artifacts caused by the motion of the heart and the breathing to improve blood flow image quality. In the present disclosure, the inventors propose to exploit brain tissue displacement signals from a temporal series of brain ultrasound images to extract the breathing and heart rates. Therefore, the heart and breathing rates can be simultaneously evaluated and monitored alongside neurofunctional ultrasound imaging of brain activity.

[0075] The present disclosure uses fUS or ULM imaging and their acquisition of ultrasonic data at ultrafast frame rates by emitting of ultrasonic plane waves, greater than 250 Hz, during periods larger than the typical observation time required in brain hemodynamics, for example 200ms (in rodents) to several seconds in human to track brain tissue motion related to physiological parameters. In the present disclosure, they demonstrate that localultrasonic estimation of brain tissue motion allows reliable detection and quantification of both heart and breathing motions during in-vivo functional ultrasound imaging directly on the ultrasonic data, without requiring any additional device for physiological monitoring. Thus, instead of removing these motions, the present disclosure provides a method based on the computation of brain tissue motion obtained from speckle tracking of raw-IQ data acquired at ultrafast frame rates during fllS or ULM imaging to detect and quantify physiological parameters. Thanks to pre-identification of specific regions of the imaging slice, the tissue motion can be averaged to obtain periodic tissue signal generated by heartbeat and respiratory motion in real time. It is also possible to extract cardiac pulsatility or heart rate and breathing rate continuously simultaneously with ultrasound neuroimaging data.

[0076] The temporal sequence of 2D or 3D Doppler images enables also to obtain values of at least one vascular parameter in at least one chose area of interest of a zone of the brain. It is possible to measure a temporal evolution of the at least one vascular dynamic parameter.

[0077] In the present disclosure, the imaged zone of the brain may cover between 3% and 100 % of the brain.

[0078] Figure 1 is a block diagram that schematically illustrates an ultrasound imaging apparatus 1 (VA APP) for imaging the brain and brain tissue motions, in accordance with an embodiment that is describes herein.

[0079] The apparatus 1 may comprise a control system 2 (SYSTEM), for instance a specialized signal processing device controlled by a computer or a group of computers, possibly a group of computers including servers.

[0080] The control system 2 may include a computing module 3 (COMP MOD), the operation of which will be explained later.

[0081] The control system 2 may control an ultrasound contrast agent injection device 4 (I NJ).

[0082] The ultrasound contrast agent injection device 4 may comprise ultrasound contrast agents. The ultrasound contrast agents may be microbubbles, nanobubbles, gas vesicles, or equivalent ultrasound contrast agents. The ultrasound contrast agents typically reflect ultrasound echoes much stronger than red blood cells and produce nonlinear signal. Therefore, the ultrasound contrast agents can be used for high-contrast imaging of blood vessel structure, as well as blood perfusion in organs. For example, prior to imaging, the ultrasound contrast agents are administrated intravenously by the ultrasound contrast agent injection device 4.

[0083] The imaging apparatus 1 comprises an ultrasound probe 5 (PRB) and the control system 2 may be configured to control the ultrasound probe 5.

[0084] The ultrasound probe 5 may include an array 7 (ARR) of ultrasonic transducers. The array may be a linear array adapted to generate a 2D image of a slice of the zone of the brain to be imaged, or a 2D array adapted to generate a 3D image of the zone. When the array is a 2D array, it may be a sparse matrix of transducers, as known in the art.

[0085] During imaging, the array of ultrasonic transducers is coupled to a patient body. The array of ultrasonic transducers transmits an ultrasonic signal into the tissue and receives backscattered signals or echoes from the zone of the brain to be imagined. In the present disclosure, the array 7 is controlled by the control system 2 to transmit the ultrasonic signals according to specific pulse sequences and to receive the backscattered signal from the zone of the brain.

[0086] The received signals are then processed by the computing module 3 using the imaging method that is described herein in order to generate a sequence of ultrasound images.

[0087] Typical arrays of transducers may include a few tens to a few thousand of transducers. The array may also in some examples, be limited to one single transducer adapted to image only one line of the zone, in the direction of the depth from the transducer or a few transducers adapted to image respectively lines of the zone, in the direction of the depth from the transducer.

[0088] The following detailed description is done for the case of a linear or 2D array, so that the apparatus generates Doppler images (more generally: ultrasound measurements) having pixels (more generally: Doppler samples). In the case where the array would include just one transducer or a few transducers, the apparatus would generate an image limited to one line (ultrasound measurement) or a few lines in the direction of depth, the line(s) having pixels (the Doppler samples) and the process would be similar except for the generation of the ultrasound measurements which would not require inclined planar waves of different angles of inclination and would not require translated diverging waves from different virtual sources.

[0089] During imaging, the transducer array 7 may be adapted to transmit and receive ultrasound waves having a central frequency comprised for instance between 0.5 and 100 MHz, for instance between 1 and 20 MHz.

[0090] In certain embodiments, the probe 4 may further include a motorization 6 (MOT) adapted to position the array probe 7.

[0091] An example of method of Doppler image acquisition and physiological parameters extraction will now be explained with regards to Figure 2.

[0092] Ultrasound imaging phase

[0093] The array 7 of transducers may be controlled by the control system 2 to acquire raw ultrasound data S1 of the zone of the brain.

[0094] The array 7 of transducers may be controlled by the control system 2 to transmit unfocused ultrasonic waves in the zone of the brain to be imaged and to receive the resulting backscattered ultrasonic waves, at a predefined Pulse Repetition Frequency (PRF), for instance 6 kHz, i.e. around every 0.2 ms. In a typical brain imaging apparatus, the Pulse Repetition Frequency PRF may be over 500Hz. The received signals are registered as a set of raw data for each transmitted diverging ultrasonic wave. The successive transmitted diverging waves are transmitted by N virtual sources which are regularly spaced with regards to the direction of the width in the zone to be imaged, i.e with regards to the direction lateral of the array 7 as illustrated in Figure 3. Unfocused ultrasound waves can be either plane waves, diverging waves, multi-focused waves (corresponding to several focused beams at different locations in the same transmission) or any ultrasound wave insonifying several regions of interest in a single transmission.

[0095] The computing module 3 may process the acquired raw data to acquire a sequence of IQ compound images of the zone S2.

[0096] The array 7 of transducers may be controlled by the control system 2 to receives backscattered signals from the zone of the brain. The backscattered signals contain linear and nonlinear signals of the ultrasound waves reflected by the ultrasound contrast agents, the blood and the surrounding tissue. The control system 2 comprises a sampler and a DAC for converting analog signals received from the probe 7 into a digital form. For example, the received signals are demodulated and sampled into In-Phase and Quadrature (IQ) signals of the respective transducers to produce an IQ image of the region. Thus, each IQ image comprises pixels representing the received backscattered signals from the region.

[0097] Thus, for each image of the zone, a number N of diverging ultrasonic waves are successively transmitted by the N virtual sources and the N sets of raw data are processed to synthesize an In phase Quadrature (IQ) image reconstructed on a polar grid. An example of such IQ image is shown in Figure 4A.

[0098] For example, in the case of N = 12 and PRF = 6 kHz, the rate of the synthesized images of the region (framerate) is thus around 500 Hz. N may be different than 12, in which case the framerate of compound images is different. For instance, N=5 may be used.

[0099] Alternatively, the array 7 of transducers may be controlled by control system 2 to transmit planar ultrasonic waves in the zone to be imaged and to receive the resulting backscattered ultrasonic waves, at a rate of for instance 5.5 kHz (Pulse Repetition Frequency PRF), i.e. around every 0,18 ms. More generally, the Pulse Repetition Frequency PRF may be over 500Hz.The received signals are registered as a set of raw data for each transmitted planar ultrasonic wave. The successive transmitted planar waves have propagation direction which are inclined of varying successive angles with regards to the direction of the depth in the region to be imaged, i.e. with regards to the direction normal to the array 7. For each image of the zone, a number N of planar ultrasonic waves are successively transmitted with different angles and the N sets of raw data are coherently added to synthesize said image of the region, which is thus a compound IQ image. For instance, N may be 11 with angles varying between -10 deg and +10 deg by steps of 2 deg. In the case of N=11 and PRF=5.5kHz, the rate of the compound IQ images of the region (framerate) is thus 500 Hz. N may be different than 11 , in which case the framerate of compound images is different. For instance, N=5 may be used.

[0100] Neurofunctional ultrasound monitoring

[0101] In both cases, based on the sequence of IQ compounds images, Doppler images of the zone are then computed by computing module 3 at step S8. For instance, 200 successive synthesized images are used for each Doppler image. In this case, the rate of Doppler images is thus of 2,5 Hz.

[0102] The computing module 3 may compute vascular dynamic parameters in the zone of the brain based on the Doppler images S9.

[0103] The vascular parameter may be computed for instance from successive filtered images using spatiotemporal filtering based on single value decomposition (SVD). The filtered images may be computed for instance by a Singular Value Decomposition (SVD). More specifically, a SVD spatiotemporal clutter filter, as for instance described by Demene et al. [Demene, C. et al. Spatiotemporal Clutter Filtering of Ultrafast Ultrasound Data Highly Increases Doppler and fUltrasound Sensitivity. IEEE Transactions on Medical Imaging 34, 2271 - 2285 (2015)], may be applied. By discarding some singular values of the SVD, the tissue signal may be filtered. For instance, the first 10 singular values of the SVD, mostly reflecting the tissue signal of the region, may be discarded.

[0104] The successive filtered mages may be squared and averaged into a final ultrasensitive Doppler image.

[0105] More generally, the Doppler images may be computed by any Doppler technique, including ultra-sensitive Doppler, power Doppler, micro-Doppler, spectral Doppler. TheDoppler signal constituting said Doppler images may be for instance power Doppler, color Doppler, spectral Doppler, index of vascular resistivity, or any combination thereof. Said Doppler signal may be filtered on different Doppler frequency bandwidths so as to assess sensitivity of the Doppler signal on blood velocity.

[0106] In some embodiments, the control system 2 may rotate the array 7 to obtain a three- dimensional, 3D, Doppler scan of the brain by acquiring Doppler images at different rotation angles. For instance, the array 6 is rotated over a 180° range by a 2° rotational step. The 3D Doppler scans may be reconstructed from the Doppler images at different rotation angles by interpolating the Doppler images on a cartesian grid.

[0107] Figure 4B illustrates a registration of a 2D Doppler image obtained with the method describes in reference to Figure 2 on a known brain atlas, which is a preexisting cartography of the brain. Such maps are available for the human brain and some animal brains, e.g the swine. In some variants, the Doppler image may be alternatively or in combination registered to a CT scan of the subject’s brain and / or a MR scan of the subject’s brain.

[0108] Computing module 3 is adapted to select an area of interest of the Doppler image. The area of interest may be equal to the imaged region. Alternatively, the area of interest may be smaller than the imaged region. The areas of interest may be chosen based on vascular territories. For instance, the areas of interest may be chosen based on a vascular brain atlas. Alternatively, or in combination, the areas of interest may be chosen based on a CT scan of the subject’s brain and / or a MR scan of the subject’s brain. Depending on the embodiments, the areas of interest may or may not cover the entire imaged region.

[0109] In another embodiment, the area of interest selected to compute the vascular parameter may be the region of interest selected to extract the physiological parameters, allowing to establish a link between the temporal evolution of the vascular parameter and the temporal evolution of the physiological parameters.

[0110] Once the area of interest is determined, the computing module 3 may average the Doppler signal on the area of interest through the Doppler scans, thus obtaining the temporal evolution of the vascular parameter in the area of interest.

[0111] To this end, computing module 3 may be adapted to compute at least one vascular dynamic parameter from the temporal series of 2D or 3D Doppler scans, said at least one parameter being chosen in the group comprising blood flow, blood velocity, in particular systolic velocity and I or diastolic velocity, blood volume and arterial resistivity index.

[0112] For determining whether the temporal evolution of the vascular parameter is normal, computing module 3 may compare said evolution to a threshold. The threshold may be set by an expert, such as a medical practitioner.

[0113] In another embodiment, the ultrasound contrast agent injection device 4 may be controlled by the control system 2 to inject intravenously ultrasound contrast agents to a zone of the vascular network prior the step S1 . The ultrasound contrast agent injection may be a bolus injection. The ultrasound contrast agent injection may be a continuous injection. In this embodiment, the array 7 of transducers may be controlled by the control system 2 to acquire S1 the raw data of the zone containing the ultrasound contrast agents.

[0114] The computing module 3 may process S2 the acquired raw data of the zone to acquire a sequence of compound IQ frames of the zone containing the ultrasound contrast agents at step 2.

[0115] The computing module 3 may compute Ultrasound Localization Microscopy (ULM) images at step S8 of the zone based on the acquired compound IQ frames at step 2. ULM images may be based on methods already known in the art and explained in the articles of Errico et al. [Errico, C. et al. Ultrafast ultrasound localization microscopy for deep superresolution vascular imaging. Nature 527, 499-+ (2015)] and Demene et al., [Demene, C. et al. Transcranial ultrafast ultrasound localization microscopy of brain vasculature in patients. Nat. Biomed. Eng 5, 219-228 (2021)].

[0116] Based on the successive compound IQ frames images of the zone of the brain, filtered images of the vascular network in said region may then be computed by the computing module 3. The filtered images may be computed for instance by a Singular Value Decomposition (SVD). More specifically, a SVD spatiotemporal clutter filter, as for instance described by Demene et al. [Demene, C. et al. Spatiotemporal Clutter Filtering of Ultrafast Ultrasound Data Highly Increases Doppler and fUltrasound Sensitivity. IEEE Transactions on Medical Imaging 34, 2271 - 2285 (2015)], may be applied. Alternatively, owing to the SVD of the compound images, for each filtered image, the ultrasonic signature of the individual ultrasound contrast agents may be discriminated from the tissue signal of the region. By discarding some singular values of the SVD, the tissue signal may be filtered. For instance, the first 10 singular values of the SVD, mostly reflecting the tissue signal of the region, may be discarded.

[0117] Ultrasound contrast agents may be detected in the filtered images. For instance, ultrasound contrast agents may be detected as the brightest local maxima with high correlation with a point spread function. High correlation may be defined as a correlation superior to a threshold. The threshold may be a value comprised between 0.5 and 1. For instance, the threshold may be equal to 0.7.

[0118] Ultrasound contrast agents may be tracked in the filtered images. A tracking of the ultrasound contrast agents may be performed using a particle tracking algorithm known inthe art. Tracks may be computed based on the tracking of the ultrasound contrast agents in the filtered images. A track may correspond to positions of a tracked ultrasound contrast agent in the filtered images. Each position of a tracked ultrasound contrast agent in the filtered images may be associated with a time position corresponding to the time position of the filtered image in which the position is located. Tracks with ultrasound contrast agents detected in a predefined number of successive filtered images may be computed. Successive positions of ultrasound contrast agents in a track may be used to compute velocity parameters S9.

[0119] At step S8, ULM images of ultrasound contrast agent count may be computed based on the detected ultrasound contrast agents. For instance, ULM count images may be computed by counting, for each pixel, the number of ultrasound contrast agents detected in the corresponding pixel of the filtered images.

[0120] At step S9, ULM velocity images of the ultrasound contrast agent velocity may be computed based on the detected ultrasound contrast agents. For instance, velocity maps may be computed based on the interframe ultrasound contrast agent velocity vector components and / or the absolute velocity magnitude.

[0121] ULM images may be used to precisely quantify the vascular parameter. In particular, the blood flow velocity and direction may be precisely quantified.

[0122] In one embodiment, during the recording period of Doppler images or ULM images, a dynamic event object is applied. This dynamic event is due to an external cause other than cardiac pulsatility and the breathing of the subject. The dynamic event may activate the vascular network in the at least one zone of the brain imaged. Based on the temporal series of ULM images or Doppler images, a measure of an evolution of the values of the at least one vascular dynamic parameter in response to the dynamical event may be computed.

[0123] The apparatus 1 as described above may be used to monitor brain functional activity. In particular, the apparatus 1 may be used to provide a reliable measure of cerebral blood flow across the whole brain.

[0124] Physiological parameters monitoring

[0125] Figure 3 illustrates schematically large-scale movements due to the breathing in a first region of interest ROI1 of the brain and more localized higher frequency movements due to heartbeat in a second region of interest ROI2 of the brain. The cardiac pulse wave propagating through arteries induces small motions in tissues

[0126] As indicated above, instead of eliminating the artifacts causes by these local brain tissue motion in the ultrasound images, the method of the present disclosure propose to compute a temporal series of brain tissue motion in these regions to quantify the physiological information, e.g. the breathing rate and the heart rate from the same sequence of IQ compound frames obtained at step S2. Indeed, breathing rate and heart rate create periodic tissue motion in the brain detectable through fllS images or ULM images. In some cases, physiological information may be extracted from the signal amplitude representative of brain tissue displacement instead of the signal period.

[0127] In other words, before to apply the SVD filter to remove the signal linked to tissue, the method of the present disclosure proposes to compute a temporal series of brain tissue motion from this sequence of IQ compound frames. For example, an autocorrelation algorithm between IQ frames can be used in this step S3.

[0128] Brain tissue motion computation may be performed by the module 3 using known methods. In one example, tissue motions may be computed by performing 1 D crosscorrelation to obtain IQ frame-to-frame tissue displacement and is based on 1 D speckle tracking. The signal is spatially averaged using a 2D spatial filter. A third-order Butterworth temporal filter is then applied with a centre frequency value of 20 Hz before being smoothed again using a 2D spatial average filter. An averaged signal over different spatial regions is computed. This averaged signal will be called spatial mean of tissue velocity (mm / s) as illustrated in Figures 5(A) and 5(B).

[0129] Thus, the same sequence of IQ frames obtained after step S2 are used to generate Doppler or ULM images S7 and to compute a temporal series of tissue motions S3.

[0130] The method proposes to select at least one region of interest in tissue motion image to extract the heart rate or the breathing rate or both S4. For example, in some cases, only one region of interest is selected in the tissue motion image to extract the two physiological parameters. In some cases, the method proposes to select a first region of interest in tissue motion images to compute the breathing rate and / or a second region of interest to compute the heart rate.

[0131] In one example, these two regions ROI1 and ROI2 may be selected manually by the user till the breathing and heart rates computed from the signal period related to tissue motion according to the method of the present disclosure correspond to the expected values.

[0132] In another embodiment, the step of selecting of the ROI 1 and ROI2 may be performed automatically by applying a filter algorithm to tissue motion data in order to reveal their different spatial and temporal characteristic.

[0133] In one embodiment, the step of automatic selection of at least one region of interest “ROI” comprises the following steps :- computing the spatio-temporal decomposition of the temporal series of brain tissue motion into a base of spatial vectors and a base of temporal vectors;- identifying a temporal vector having a signal period corresponding to an expected physiological information;- selecting a region of interest based one the spatial vector corresponding to the identified temporal vector.

[0134] In one example, the step of identifying a temporal vector having a signal period corresponding to an expected physiological information may comprise the following substeps:- computing the autocorrelation of the temporal vectors;- detecting temporal peaks in the autocorrelation by applying a peak detection algorithm with a periodicity range corresponding to a signal period representative of the expected physiological information.

[0135] Alternatively, the step of identifying a temporal vector having a signal period corresponding to an expected physiological information may comprise the following substeps:- computing the periodicity of temporal vectors using Fourier transform- detecting temporal peaks in the Fourier spectrum in a period range representative of the expected physiological information.

[0136] In one example, the method proposes to apply a Singular Value Decomposition “SVD” to the temporal series of motion data to select the two regions of interest where tissue motions signals reflecting breathing motion and heartbeat as illustrated in Figure 8.

[0137] In another example, the method proposes to apply a frequency filter to the temporal series of motion data to select the two regions of interest as illustrated in Figure 11.

[0138] In both cases, the tissue velocity in the selected ROI obtained after step S4 are used to compute spatial averaged of tissue velocity S5 by the computing module 3 of the apparatus.

[0139] The periodicity of the tissue velocity in the selected ROI obtained after step S5 are computed S6. This periodicity can be obtained for example by Fourier transformation, autocorrelation, peak detection or Lomb scargle periodograms.

[0140] Finally, the heart and breathing rates are extracted from the average estimates of the signal periods S7. In addition, it is possible to measure a temporal evolution of the heartand breathing rates from the temporal series of brain tissue motion either in real time during live imaging or in postprocessing of previously acquired data.

[0141] Figure 4(C) illustrates an example of three brain tissue motion images computed from the sequence of IQ frames obtained at step S2 by autocorrelation at time to, time to + 80 ms and at time to + 160 ms. In the example of Figure 4(c), the tissue motion images are obtained from a sequence of IQ frames acquired for a mouse, in a coronal plane. The three tissue motion images are registered on an atlas of brain. The two regions of interest ROI1 and ROI2 specific respectively to the breathing rate and the heart rate are selected and presented in the three images. The ROI1 is presented by a 2.0 mm x 2.0 mm square and the ROI2 is presented by a 0.8 mmx 0.8 mm square.

[0142] Figure 5(A) and Figure 5(B) illustrate an example of the spatial mean of tissue velocity computed respectively in the first manually defined region of interest ROI1 and second manually defined region ROI2 presented in Figure 4(C). The range of tissue velocity amplitude is 1 mm / s.

[0143] Figure 5(C) illustrates the average period between peaks of Figure 5(A) using a sliding window of 1 .2 seconds for RO11 .

[0144] Figure 5(D) illustrates the average period between peaks of Figure 5(B) using a sliding window of 2 seconds for ROI2.

[0145] The two periodic signals measured in RO11 and ROI2 shows two distinct frequencies or periodicities. In RO11 , the average signal period is 195 bpm (± 2 bpm), while in ROI2, the average period is 345 bpm (± 5 bpm). These two periods match the expected breathing rate and heart rate of anaesthetised mice. As explained with reference to Figure 3, large-scale movements of the brain tissue are induced by the breathing and more localised higher frequency movements of the brain tissue are induced by the heartbeat.

[0146] Figure 6 and Figures 7 show that brain tissue pulsatility periodicity depends strongly on the region in which it is averaged.

[0147] Figure 6 illustrates an example of ultrafast B-Mode image obtained from 200 IQ frames. Four regions of interest RO11 , ROI2, ROI3, ROI4 are selected manually in the image.

[0148] Figures 7(A)-7(D) show spatial averaging of tissue velocity in the four regions of interest over a time period of 4 seconds. The four measurements show that the amplitude and periodicity vary with the location of the region of interest selected. In other words, an arbitrary selection of regions of interest leads to spatially averaged signals of complex shapes with varying amplitude and periodicity. Thus, the detection of the physiological brainrhythms depends strongly on the selection of well-defined, plane-specific, and possibly animal-specific regions of interest.

[0149] Thus, in case of manual selection of the region of interest, it is necessary to repeat the steps S3 to S7 till the breathing rate and the heart rate extracted are comprised in a range of values expected. This approach may be time consuming and might lead to interoperator variability.

[0150] The method of the present disclosure proposes to decompose tissue motion data caused by heartbeat and breathing motion, obtained after step 3, by leveraging their distinct spatial location and temporal characteristics.

[0151] Figure 8 illustrates in detail an example of the processing step S4 of tissue motion images to select automatically the ROI1 and the ROI2.

[0152] The step of selection of region of interest RO11 and ROI2 may comprise the following steps:- computing the decomposition of the brain tissue motion into a base of spatial vectors and a base of temporal vectors by using a Singular Vector decomposition;- computing the autocorrelation of the temporal vectors;- detecting temporal peaks in the autocorrelation by applying a peak detection algorithm with a periodicity range corresponding to a signal period representative of the expected physiological information.- identifying a temporal vector having a signal period corresponding to a physiological information ;- selecting a region of interest based one the spatial singular vector corresponding to the identified temporal singular vector.

[0153] The 3D tissue motion data (2D + time) are shaped into a 2D space-time Casorati matrix form.

[0154] A space-time decomposition is applied on the 2D Casorati matrix into a spatial base comprising spatial singular vectors and a temporal base comprising temporal singular vectors. This step of singular vector decomposition decomposes the data into an ordered sum of separable spatio-temporal vectors. Each mode has individual ranks. It is possible to select the number of singular vectors to be removed in each dimension separately, i.e. respectively in spatial and temporal dimensions. Thus, thanks to the SVD processing, the Casorati matrix is decomposed on a time basis and a spatial basis.

[0155] Other decompositions into separable (or uncorrelated) spatio-temporal bases than SVD are possible. For example, it is possible to use a Principal Component Analysis “PCA”to decompose the motion data into separable spatio-temporal bases or an Independent component analysis “ICA”.

[0156] The processing S4 of selecting ROI comprises a step of finding the threshold to select singular values and vectors. Based on the assumption that tissue signal is described in the six first most energetic singular vectors, a rejection is performed using a threshold selection on the number of singular vectors to be removed. For instance, the six first singular values of the SVD, mostly reflecting the tissue signal of the region, may be retained. Figure 9(A) illustrates six first spatial singular vectors and Figure 9(B) illustrates six first temporal singular vectors. While some temporal vectors (such as the 2nd and 6th) do not exhibit any well-defined periodicity, others (such as the 1st and 4th) clearly matched the signals presented in Figure 5. The first spatial singular vector, corresponding to a 3.2 Hz periodic signal, encompasses a large region of the left brain, while the fourth spatial vector, associated with a 6 Hz periodic signal, is linked to a small region in the centre of the brain.

[0157] A selection process is applied to extract the singular vectors corresponding to heart and breathing rates based on the assumption that two of the first six singular vectors pulse at a frequency or periodicity in an expected periodicity range for the animal’s heart rate and respiration rate, respectively. This expected frequency range or periodicity range for the physiological parameters depends on the animal model.

[0158] The frequency range for the breathing rate may be between 10 bpm and 80 bpm in case of clinical imaging, and between 50 and 200 bpm in case of preclinical imaging. The frequency range for the heart rate may be between 30 bpm and 180 bpm in case of clinical imaging, and between 150 and 800 bpm in case of preclinical imaging.

[0159] The processing S4 comprises a step wherein the autocorrelation of each temporal singular vector is computed. The results are shown in Figure 9(D).

[0160] The processing S4 comprises a step wherein a peak detection algorithm is applied with a frequency threshold, for example with a frequency threshold of 500 bpm, a minimum peak threshold of 0.3 and a minimum distance between peaks of 0.12 seconds. These parameters may varying with the experimental setup. The temporal positions of these peaks, reflecting the periodicity of the singular vector, are sorted.

[0161] The processing S4 comprises a step wherein a first temporal singular vector having a signal period corresponding to a breathing rate and a second temporal singular vector having a signal period corresponding to a heart rate are identified. This step of detecting two eigenvectors corresponding to the heart signal and to the breathing signal among the six temporal vectors is performed by assuming that the first peak position of all six temporal vectors corresponded to the highest frequency detected in tissue pulsatility, presumably theheart rate. The relative height of the six time-vectors autocorrelations are compared in a temporal window, for example of 0.01 second around this peak to determine which of the vectors has a pulsatility whose period is closest to that of the heart rate. The first temporal singular vector is selected from this maximum, and its corresponding spatial singular vector is used to extract the region of interest in which the cardiac pulsatility signal is calculated. The second temporal vector is selected among other vectors using the second-lowest temporal peak positions, presumably corresponding to the breathing rate.

[0162] From the previous step, the first singular eigenvector and the fourth singular eigenvector are identified as corresponding to the breathing signal and the heartbeat signal. In Figure 9(D), the results are marked by a star.

[0163] Once temporal singular vectors corresponding to the breathing rate and the heart rate are selected among the six temporal singular vectors, a region of interest based on the spatial singular vector associate with the identified temporal vectors is extracted.

[0164] The processing step S4 comprises a step wherein two spatial masks are constructed, indicating the two ROI as shown in Figure 9(E) and Figure 9(F).

[0165] To extract the breathing rate, 300 pixels may be retained with the highest intensity for mice, or 1000 pixels in the case of rats in the ROI to compute the spatial average of tissue velocity. The pixels whose intensity is greater than 80% of the maximum value are retained. These two spatial masks for respiratory and cardiac estimation are then used to estimate average tissue pulsatility. The heart rate and the breathing rate are extracted from this estimated average tissue pulsatility.

[0166] Figure 10 illustrated in detail the step of identifying the two eigenvectors corresponding to the breathing rate and the heart rate among the six vectors according an embodiment :- SVD decomposition is performed to extract temporal singular vectors. Figure 10(A) shows six more energetic temporal singular vectors;- autocorrelation is applied on the temporal singular vectors (Figure 10(B);- a peak detection algorithm is applied in the computed autocorrelation (lag greater than 0.1 seconds and 500 Hz) (Figure 10(C). The intensity threshold, represented by a dashed line is fixed to 0.3. Detect peaks are marked with an arrow;- the first time position of the peaks is selected and the relative heights of the autocorrelation are compared in a temporal window around the first peak indicated by a grey patch as shown in Figure 10(D);- the second time position of the peak is selected and their relative heights of the autocorrelation are compared in a time window around the second peak indicated by a greypatch as shown in Figure 10(E);- the two temporal singular vectors for heart rate and breathing rate are framed respectively in the first vector (Vector 1) and in the fourth vector (Vector 4) as shown in Figure 10(F).

[0167] With reference to Figure 11 , the step of selection of at least one region of interest “ROI” according to another embodiment is detailed below. This embodiment uses the spatial approach and comprises the following steps:- defining a periodicity range corresponding to the expected signal period of a physiological information;-generating a map in which each pixel's intensity corresponds to the amplitude of the Fourier transform of the temporal series of brain tissue motion in the predefined periodicity range;- selecting a region of interest based on the pixel's intensity of the previous map. In other words, the ROI may contain pixels having a pixel intensity higher than a threshold chosen.

[0168] In a similar manner, the periodicity range for the breathing rate may be between 10 bpm and 80 bpm in case of clinical imaging, and between 50 and 200 bpm in case of preclinical imaging. The frequency range for the heart rate may be between 30 bpm and 180 bpm in case of clinical imaging, and between 150 and 650 bpm in case of preclinical imaging.

[0169] The present method may be applied in experimental setups in which the ultrasound acquisition is not continuous. The physiological parameters can still be extracted from brain tissue motion using specific periodicity detector tools such as the Lomb-Scargle periodogram, to overcome missing data points.

[0170] The efficiency of the method of the present disclosure will be demonstrated below.

[0171] Experimental results

[0172] Ultrasound imaging and sequences

[0173] For image acquisition, functional ultrasound imaging acquisitions are performed with a 1 D linear transducer (15 MHz central frequency, 128 elements, 110 spatial pitch) connected to a functional ultrasound scanner. A sequence of 11 plane waves tilted from - 10° to +10° are emitted at a 5.5 kHz pulse repetition frequency and the backscattered echoes of each group are coherently compounded in 2D images at either 500 Hz and 1000 Hz.

[0174] As indicated above, Doppler images are computed by applying a clutter filter based on Singular Value decomposition (SVD) of each block of 200 consecutives frames. The SVD threshold is fixed to 30. Each pixel of the final power Doppler image was reconstructed on a 1 lOjjmxIOOjjm grid in the plane and the slice thickness was approximately 400 m.Functional ultrasound acquisition lasted between 2min to 10min for anaesthetised experiments and 60 min for sleep sessions.

[0175] Three parameters used in the evaluation of the image quality are contrast-to-tissue ratio (CTR), contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR), respectively, as follow:

[0176] Assessment of physiological parameters from implanted electrodes

[0177] Breathing rate is monitored by a pillow placed under the animal's body and connected to the Power Lab system. Heart rate is measured using 3 needles placed under the skin and connected to the same device. These two signals, acquired at 1000 Hz were recorded and post-processed rather than directly with built-in labchart tools to overcome user dependency thresholds and to increase the final resolution of breathing and heart rate. A 3Hz Butterworth low-pass filter is applied to the respiratory motion signal and finally the peak detection function is used to compute the average period of the signals over time. This extraction of physiological parameters gave a final frequency of measure of 1 Hz and 0 5Hz for HR and BR respectively.

[0178] Figure 12 illustrates the comparison between the method of the present disclosure and the electrode method to extract the breathing and heart rates.

[0179] The experimental setup comprises the device as illustrated in Figure 1 and an electrode measurement device. Ultrasound acquisitions and electrode signal acquisitions are synchronized in time. A one-second delay was applied to the electrode signal to account for the time required by the ultrasound scanner to launch the acquisition.

[0180] The comparison of the two measures is performed in a mouse as illustrated in Figure 12A, covering coronal plane during a 10 min acquisition. The average temporal values respectively for heart rate and breathing rate are 363.3 bpm and 208.7 bpm for functional ultrasound imaging assessment, and 364.1 and 208.7 bpm for electrodes measurements.

[0181] Superposition of the heart rate (top) and breathing rate (bottom) assessed with electrode and function ultrasound imaging as described above with reference to Figure 2 shows a low deviation between fUS and electrode assessments during 10 min acquisitions.

[0182] In a similar manner, the comparison of the two measures is performed in a rat as illustrated in Figure 12B, covering sagittal planes during a 10 min acquisition. Superposition of the heart rate (top) and breathing rate (bottom) assessed with electrode and function ultrasound imaging shows a low deviation between fUS and electrode assessments during 10 min acquisitions.

[0183] The method of the present disclosure allows to evaluate and monitor simultaneously the heart and breathing rates alongside neurofunctional ultrasound imaging of brain activity. Both heart and breathing rates can be extracted directly from brain tissue motion signals detected in the ultrasound images instead of using an external device to monitor these physiological parameters.

[0184] The simultaneous extraction of the heart and breathing rates during any fllS or ULM acquisition, without the need for a specific ultrasonic acquisition sequence, offers the systematic integration of these physiological parameters into fllS neuroimaging analysis, helping the understanding of the connections between physiological rhythms and brain connectomics, neurovascular coupling, brain states or circadian rhythms.

[0185] In other words, the method of the present disclosure may be easily integrated in functional ultrasound technology for simultaneous neurofunctional ultrasound imaging and physiological monitoring.

[0186] This integration would facilitate live monitoring of physiological parameters during fllS or ULM experiments, reducing the need for additional sensors and simplifying the setup while ensuring precise synchronisation.

[0187] The present method can be used for a better understanding of the effects of drug treatments on both neuronal activity and physiological parameters in preclinical models.

[0188] The method of the present disclosure can be also used for simultaneous assessments of physiological information and fUS or ULM data in clinal settings.

[0189] The present method can be used to perform behavioral experiments by implanting a portable ultrasonic probe on the animal head, enabling observation of animals' interactions with their environment, other animals, and their internal rhythms.

[0190] The present disclosure provides an apparatus allowing the implementation of realtime assessment of heart rate and respiratory rate in an ultrafast ultrasound imaging system.

[0191] Figure 13 illustrates an example of screen of such ultrasound imaging apparatus in which it may possible to display at same time the Doppler or ULM image showing the brain activity, the breathing rate ROI and heart rate ROI overlaid on the patient’s Atlas, the temporal evolution of physiological parameters, e.g. breathing rate and heart rate.

Claims

-28-Claims

1. Method for imaging vascular activity in a brain of a living being, the method including:(a) transmitting unfocused ultrasound waves in the brain by an array ultrasound probe and acquiring raw data from backscattered ultrasonic waves by said array ultrasound probe;(b) generating a sequence of compound ultrasound images of the brain from said raw data during a recording period, said sequence of compound ultrasound images forming a sequence of successive raw-IQ (in-phase / quadrature) frames;(c) computing a temporal series of brain tissue motion based on said sequence of successive raw-IQ frames;(d) selecting at least one region of interest “ROI” of the brain in said temporal series of brain tissue motion for computing at least one physiological information;(e) computing said at least one physiological information based on the temporal series of brain tissue motion in the selected region of interest;(f) generating a sequence of Doppler or Ultrasound Localization Microscopy “ULM” images of said brain based on said sequence of successive raw-IQ frames.

2. Method according to claim 1 , wherein the step of selection of said at least one region of interest “ROI” is performed automatically using the temporal periodicity of a signal representative of brain tissue motion.

3. Method according to claim 1 or 2, wherein the at least one physiological information is chosen in a group comprising: breathing rate, heart rate.

4. Method according to claim 3, wherein the at least one region of interest comprises a region of interest specific for computing the breathing rate or the heart rate or both.

5. Method according to any of claims 1 to 4, wherein the step of selection of at least one region of interest “ROI” comprises:- computing the spatio temporal decomposition of the temporal series of brain tissue motion into a base of spatial vectors and a base of temporal vectors;- identifying a temporal vector having a signal period corresponding to an expected physiological information;- selecting a region of interest based on the spatial vector corresponding to the identified temporal vector.

6. Method according to claims 5, wherein the step of identifying a temporal vector having a signal period corresponding to an expected physiological information comprises:- computing the autocorrelation of the temporal vectors;- detecting temporal peaks in the autocorrelation by applying a peak detection algorithmwith a periodicity range corresponding to a signal period representative of the physiological information.

7. Method according to claim 5, wherein the step of identifying a temporal vector having a signal period corresponding to an expected physiological information comprises:- computing the periodicity of temporal vectors using Fourier transform;- detecting temporal peaks in the Fourier spectrum in a periodicity range representative of the physiological information.

8. Method according to any of claims 5 to 7, wherein the computation of the decomposition of the temporal series of brain tissue motion comprises a Singular Value Decomposition “SVD” or a Principal Component Analysis “PCA” or an Independent component analysis “ICA” or any other spatiotemporal decomposition method.

9. Method according to any of claims 1 to 4, wherein the step of selection of at least one region of interest “ROI” comprises:- selecting a periodicity range corresponding to the periodicity of at least one expected physiological information;- generating a map in which each pixel's intensity corresponds to the amplitude of the Fourier transform of the temporal series of brain tissue motion in the preselected periodicity range;- selecting a region of interest based on the pixel's intensity of the previous map.

10. Method according to any of claims 2 to 9, wherein the periodicity range for the breathing rate is between 10 bpm and 80 bpm in clinical imaging, and between 50 and 200 bpm in preclinal imaging.

11. Method according to any of claims 2 to 9, wherein the periodicity range for the heart rate is between 30 bpm and 180 bpm in clinical imaging, and between 150 and 650 bpm in preclinical imaging.

12. Method according to any of claims 1 to 11 , wherein the temporal series of brain tissue motion is computed by means of tissue speckle tracking or any other motion tracking technique applied to the Ultrasound data.

13. Method according to any of claims 1 to 12, further comprising:(g) computing a temporal evolution of at least one vascular dynamic parameter in a region of the brain based on said Doppler images.

14. Method according to any of claims 1 to 12, wherein said recording period of the sequence of ultrasound compound images including at least one external dynamic event, due to a cause other than cardiac pulsatility and / or breathing pulsality, the dynamic eventactivating a change in a local hemodynamics and / or a change in a structural conformation in a region of the brain, the method further comprises:(g) computing a temporal evolution of at least one vascular dynamic parameter in a region of the brain in response to the dynamic event.

15. Method according to any of claims 1 to 12, further comprising:(g) computing at least one vascular or neurovascular dynamic parameter in a region of the brain based on said ULM images, said step (g) of computing at least one dynamic parameter comprising a localization step and a tracking step of microbubbles or ultrasound contrast agents, in a patient that has been previously administrated with microbubbles or ultrasound contrast agents.

16. Method according to any of claims 13 to 15, wherein the at least one dynamic parameter is chosen in the group comprising: blood flow, blood velocity, blood volume, vessel diameter, brain region size and any combination thereof.

17. Method according to any of claims 1 to 16, wherein at least one Doppler image is registered on a brain atlas, a computed tomography scan of said brain, a magnetic resonance image scan of said brain and any combination thereof.

18. Apparatus for imaging vascular activity in a brain of a living being, said apparatus including at least an array ultrasonic probe (5) and a control system (2) configured to:(a) transmit unfocused ultrasound waves in the brain by an array ultrasound probe and acquiring raw data from backscattered ultrasonic waves by said array ultrasound probe;(b) generate a sequence of compound ultrasound images of the brain from said raw data during a recording period, said sequence of compound ultrasound images forming a sequence of successive raw-IQ (in-phase / quadrature) frames;(c) compute a temporal series of brain tissue motion based on said sequence of successive raw-IQ frames;(d) select at least one region of interest “ROI” in said temporal series of brain tissue motion for computing at least one physiological information;(e) compute the at least one physiological information based on the temporal series of brain tissue motion in the selected at least one region of interest of the brain;(f) generate a sequence of Doppler or Ultrasound Localization Microscopy “ULM” images of said brain based on said sequence of successive raw-IQ frames.

19. Apparatus according to claim 18, wherein the control system is configured to display Doppler or ULM image showing the brain activity and at the same time the value ortemporal evolution of the breathing rate or / and the value or temporal evolution of the heart rate.

Images