A system and method for measuring intracranial pressure using optical interferometry.

Abstract

A method and system for estimating intracranial pressure using a regression model. Optical interferometric measurements of cerebral blood flow are generated. Intracranial pressure is estimated using a regression model based on the optical interferometric measurements and one feature of extracerebral blood flow or extracerebral blood pressure.

Description

[Technical Field]

[0001] This disclosure relates to a system and method for measuring intracranial pressure using optical interferometry. [Background technology]

[0002] Conventional solutions for measuring intracranial pressure can have various problems. In this regard, conventional systems and methods for measuring intracranial pressure can be expensive, cumbersome, and / or inefficient. [Overview of the Initiative]

[0003] Systems and methods for measuring intracranial pressure using optical interferometry are shown and / or described in association with at least one of the figures and are more fully described in the claims.

[0004] These advantages, aspects and novel features of the present disclosure, as well as other advantages, aspects and novel features of the present disclosure and details of embodiments relating thereto, will be better understood from the following description and drawings. [Brief explanation of the drawing]

[0005] The various features and advantages of this disclosure may be more readily understood by referring to the following detailed description in conjunction with the attached drawings, where similar reference numbers indicate similar structural elements. [Figure 1] Figure 1 shows an interferometric measurement system for measuring intracranial pressure (ICP). [Figure 2] Figure 2 shows an example of cerebral blood flow as a function of time. [Figure 3a] Figure 3a shows three exemplary processes for estimating ICP. [Figure 3b] Figure 3b shows an exemplary CBFi vs. ABP plot. [Figure 4a] Figure 4a shows an exemplary measurement configuration for obtaining extracerebral blood flow measurements. [Figure 4b]Figure 4b shows the histogram of received photons as a function of time of flight. [Figure 4c] Figure 4c shows the time of flight of photons with respect to the separation between two different light sources and detectors. [Modes for carrying out the invention]

[0006] The following description provides various examples. These examples are non-limiting, and the attached claims should not be limited to the specific examples disclosed. In the following description, the terms “example” and “for example” are non-limiting.

[0007] Each figure illustrates a general form of the structure, and descriptions and details of well-known features and techniques may be omitted to avoid unnecessarily obscuring this disclosure. Furthermore, elements in each figure are not necessarily drawn to scale. For example, the dimensions of some elements in each figure may be exaggerated relative to others to enhance understanding of the examples described in this disclosure. The same reference numeral in different figures indicates the same element.

[0008] The term "or" means one or more of the items in the list being joined by "or". For example, "x or y" means any element of the 3-element set {(x), (y), (x, y)}. Another example is "x, y, or z" meaning any element of the 7-element set {(x), (y), (z), (x, y), (x, z), (y, z), (x, y, z)}.

[0009] The terms “equipped,” “containing,” “included,” and / or “contained” are “open-ended” terms that identify the presence of the described feature but do not exclude the presence or addition of one or more other features.

[0010] The terms “First,” “Second,” etc., may be used herein to describe various elements, but the elements should not be limited by these terms. The terms are used solely to distinguish one element from another. For example, the first element described herein may be referred to as the second element without departing from the teachings of this disclosure.

[0011] Unless otherwise specified, the term “connected” may be used to describe two elements that are in direct contact with each other, or two elements that are indirectly connected by one or more other elements. For example, when element A is connected to element B, element A may be in direct contact with element B, or it may be indirectly connected to element B by an intervening element C. Similarly, the terms “above” or “on top of” may be used to describe two elements that are in direct contact with each other, or two elements that are indirectly connected by one or more other elements.

[0012] Intracranial pressure (ICP) measurement is performed during the diagnosis and treatment of various medical conditions, including traumatic brain injury, stroke, and hydrocephalus. Current measurement solutions are invasive and can take several forms, including pressure gauges inserted directly into the brain or catheters inserted directly into the ventricles. Such invasive ICP measurement carries a considerable risk of infection, requires highly trained staff, and is therefore costly in terms of health, hospital resources, and finances. As a result of the risks and costs of such invasive measurement, ICP measurement is not applicable to many conditions where the risks of such invasive measurement do not outweigh the value of such measurement and cannot be used outside of the operating room or intensive care unit. For these and other reasons, non-invasive ICP measurement using near-infrared (NIR) light in the approximate wavelength range of 700 nm to 2500 nm may be desirable. For example, the use of NIR light may be advantageous because it is less absorbed by skin, bone, and / or other human tissues than visible light.

[0013] Broadly, the present disclosure relates to a system and method for interferometric detection of intracranial pressure (ICP). The system may comprise two or more optical paths, namely, a first reference optical path and a second optical path that may travel through an object, which is a subject of a study. Typically, the object for ICP measurement may be a human head. The light that may cross the second path may be referred to as sample light and is combined with the reference light. In many cases, the sample light may be combined with the reference light by interfering two optical signals. The combined light illumination field is then directed to one or more detectors that may appropriately process the received light and convert the received light into an electrical signal for further processing.

[0014] The signal thus obtained may indicate characteristics related to blood flow, such as the pulsatile waveform of blood flow, which may be related to ICP as will be described in more detail below.

[0015] To further improve the system, the acquired interferometric measurements may be combined with measurements of extracranial blood flow.

[0016] The system may be enhanced by using a machine learning-based data-driven model.

[0017] Referring now to FIG. 1, an interferometric measurement system 10 for measuring intracranial pressure (ICP) in a subject's head 2 is shown.

[0018] The iNIRS system 10 may include a light source 20, a light source modifier 22, and an optical splitter 24. Further, a light delivery probe 25’, a light reception probe 35’, a sample delivery channel 25, a reference channel 26, and a sample reception channel 35 are shown. Also shown are a photodetector 30 and a controller 40. Insert FIG. A shows a more detailed view of the photodetector 30 shown in the dashed box in FIG. 1. Insert FIG. B shows a more detailed view of an alternative camera-based photodetector 30.

[0019] Multiple sample delivery channels 25, probes 25', multiple sample receiving probes 35', and associated sample receiving channels 35 may be present. The sample receiving probes 35' may be operable to receive light. The sample receiving channels 35 may be single-mode fibers (typically associated with the mechanism of a photodiode detector 30 as shown in inset A) or multi-mode fibers (typically associated with the mechanism of a camera-based detector 30 as shown inset B). According to various embodiments of this patent, the system may comprise one or more detectors 30 and / or controllers 40.

[0020] The iNIRS system 10 is connected to the object / head 2 to be imaged / monitored, as shown. The light source 20 may comprise a laser capable of emitting light conforming to specific characteristics, some of which may be controlled by a light source modifier 22. According to various embodiments of this patent, the laser may also be a high-coherence laser. The laser may be, for example, a distributed feedback laser ("DFB") or a MEMS vertical cavity surface-emitting laser ("MEMS-VCSEL"). Other suitable types of lasers include distributed Bragg reflector lasers ("DBR"), Fourier domain mode-locked lasers ("FDML"), vertical cavity surface-emitting lasers ("VCSEL"), external cavity diode lasers (ECDL), feedback-stabilized lasers, or line-lock lasers. The main oscillator power amplifier (MOPA) configuration may be used according to various embodiments. As will be apparent to those skilled in the art, various laser configurations comprising, for example, a LiDAR laser, a laser for coherent telecommunications, or an optical coherence tomography (OCT) laser may be used for this patent. Furthermore, or alternatively, a pulsed supercontinuum laser may be used in conjunction with a pulse stretching mechanism such as a grating or GRISM pulse stretcher or a dispersed optical fiber of a certain length. The mechanism may be configured to temporally separate wavelengths in the pulse, for example, so that a frequency chirp pulse is created (for example, to provide an interferogram when the sample pulse and the reference pulse are ultimately compared).

[0021] The light source modifier 22 may include a source (e.g., a variable current or voltage provider) that provides a variable electrical control signal to the light source 20. The light source modifier 22 is connected to the light source 20. The light source modifier 22 may be electrically connected to the light source 20 to provide a variable current / voltage to the light source 20 and to be able to control the light source 20 so as to preferably adjust the characteristics of the light emitted by the light source 20. Specifically, the light source modifier 22 may interact with the light source 20 to change the wavelength of the light emitted from the light source 20. This may be referred to as wavelength-swept emission or frequency-swept of light.

[0022] The light source 20 is connected to an optical splitter 24. The optical splitter 24 may include a suitable circuit and / or hardware having an input for receiving light from the light source and capable of outputting light at two or more outputs. The optical splitter 24 of the present disclosure may include an optical fiber splitter or a free-space beam splitter. The outputs of the optical splitter 24 are connected to, for example, a reference delivery channel 26 and a sample delivery channel 25. The splitter may be configured such that most of the light is directed towards the subject's scalp via the sample delivery channel 25. The splitter may be, for example, a 90:10 splitter or a 99:1 splitter. Each photodetector 30 may be indirectly connected to the light source 20 to receive reference light from the reference delivery channel 26 and one or more sample delivery channels 35. The sample delivery channel 25, the reference delivery channel 26, the sample reception channel 35, the first sample reception channel 351, and the second sample reception channel 352 can be communicatively connected to an optical device by enabling optical transmission and are typically implemented as optical fibers. Some or all of the optical channels of the iNIRS system 10 may be provided by optical fibers. The iNIRS system 10 may include, as relevant, optical devices such as lenses for beam steering, reflection, and / or refraction devices.

[0023] The sample delivery channel 25 connects the optical splitter 24 to the sample delivery probe 25'. The sample delivery probe 25' is positioned in the scalp of the subject's head 2. The sample delivery probe 25' is capable of connecting the light received from the optical splitter 24 to the object / subject's head 2. Similarly, the sample receiving probe 35 is capable of connecting to the subject's head / object 2 to receive light. The sample delivery probe 25' may include, for example, one or more lenses that spatially distribute the sample light from the sample delivery channel 25 toward the subject's brain tissue. As another example, one or more of the sample receiving probes 35 may include a lens that focuses the received light toward a sample receiving channel 35 associated with the received light (such as one connected to that sample receiving probe). As yet another example, the probes 25', 35' may be bare optical fibers that have been split and / or polished.

[0024] Each sample receiving channel 35 may comprise a single-mode fiber (SMF), a minority-mode fiber (FMF), or a multimode fiber ("MMF"). In systems using a camera sensor instead of a photodiode, the sample delivery channel 25, the reference channel 26, and / or the sample receiving channel 35' may often also be multimode fibers (MMF). The reference channel 26 may be provided by an SMF, MMF, or FMF.

[0025] Object 2 may be, for example, the subject's head for general interferential neurobiological imaging. Generally, Object 2 is any biological object or biological tissue / substance such as skin, bone, muscle, fat, and / or brain, and may include bodily fluids such as blood. In the case of ICP measurement, Object 2 is typically the subject's head.

[0026] Each sample receiving probe 35' may be positioned on the scalp of the subject's head 2. Each sample receiving probe 35' may be connected to an associated sample receiving channel 35, which is connected to an associated photodetector 30. In other words, each photodetector 30 is connected to receive sample light indirectly from the light source 20, where the sample light travels from the sample delivery probe 25 through the subject's head 2 to one or more sample receiving probes 35.

[0027] The detector 30 is capable of receiving and processing optical input signals. Furthermore, the detector 30 may be equipped with appropriate logic, circuits, and / or code that can appropriately convert the received optical input signals into electrical signals and process them. The detector 30 may generate electrical output signals that can be connected to the input of the controller 40. The detector 30 may, for example, receive light that has traveled through the object 2 via the sample receiving probe 35' and the sample delivery channel 35. Alternatively, the detector 30 may receive light from the light source 20 that has not traveled through the object 2 via the reference delivery channel 26, the reference receiving channel 36, and the optical splitter 24. Further details of the detector 30 are described below with reference to insets A and B.

[0028] The controller 40 may include appropriate logic, circuitry, and / or code having data reception and data processing functions. The controller 40 may include, for example, one or more application-specific integrated circuits ("ASICs"). Other examples of the controller 40 may include field-programmable gate arrays ("FPGAs") and / or data acquisition modules ("DAQs"). The controller 40 may also include a microcontroller or microprocessor. According to various embodiments of this patent, the controller 40 may include machine learning logic, circuitry, and / or code.

[0029] The controller 40 is electrically connected to each of the detectors 30 (one detector 30 is shown in Figure 1), as indicated by the dashed lines. The controller 40 may be connected to each detector 30 via a wired connection (to receive electrical signals indicating detection from each detector 30), and / or the connection may be wireless (to receive transmitted data indicating detection from each detector 30). The controller 40 may also be connected to a light source modifier 22. This connection may be wired or wireless. Accordingly, the controller 40 may control the function of the light source modifier 22 based on the processing of input signals received from one or more detectors 30. Furthermore, according to various embodiments of this patent, the controller 40 may control the function of the light source modifier 22 independently of the signals received from the photodetectors 30.

[0030] An exemplary configuration of a detector 30 that can be used to convert and process optical signals into electrical signals according to various embodiments of the present invention is shown with reference to inset A of Figure 1. The exemplary detector 30 may be configured to convert three optical inputs into discrete-time / digitalized, i.e., sampled electrical signals. The same method may be used with only two optical inputs, one of which may be connected to a reference receiving channel and the other to a sample receiving channel.

[0031] Referring to inset A in Figure 1, the detector 30 may comprise an optical combiner / splitter 301, a balanced photodetector 303, and an analog-to-digital converter ("ADC") 306. An exemplary balanced photodetector 303 may comprise a detection photodiode 310, 312, a transimpedance amplifier (TIA) 304, and an amplifier 305. The ADC 306 is arranged to provide a digital signal output 307 that can correspond to the electrical output of the detector 30, as shown by the dashed line in Figure 1. The digital signal output 307 may also be referred to as an interferogram, as described below.

[0032] The optical input of the detector 30 to the optical combiner / splitter 301 is connected to a reference receiving channel 36 and exemplary sample receiving channels 351, 352 (which may be connected to, for example, one or more sample receiving channels 35, but only one sample receiving channel is shown).

[0033] An example of a mechanism for converting a received optical signal into digital data is shown in inset A of Figure 1. Inset A shows the mechanism of a component that may be used as part of the photodetector 30 of the present disclosure. As shown in the iNIRS system 10 of Figure 1, the detector 30 is configured to receive three inputs: (i) reference light traveling along a reference transmission channel 26 and a reference reception channel 36, (ii) first sample light being received through a first sample reception channel 351, and (iii) second sample light being received through a second sample reception channel 352.

[0034] The optical combiner / splitter 301 is positioned to additionally couple the electric fields of the optical signals received through channels 36, 351, and 352. The channels may be connected to, for example, beam coupling elements. This coupling may be achieved using, for example, fused fiber couplers, beam splitter cubes, diffraction gratings, or other separation / coupling optical elements such as multiplexing optics. The coupled optical signal, sometimes referred to as the mixed signal, is then optically separated in the optical combiner / splitter 301 and output to the first optical channel 302a and the second optical channel 302b. The mixed signal energy may be separated 50:50 between the first optical channel 302a and the second optical channel 302b, for example. According to various embodiments of the present invention, the ratio of how the mixed signal energy is separated between channels 302a and 302b may be adapted to the specific configuration of the detector 30.

[0035] The optical combiner / splitter 301 may receive, for example, optical signals at its input that can be represented as electric fields E36(t), E351(t), and E352(t) for channels 36, 351, and 352, respectively. The optical combiner / splitter 301 may also generate output signals proportional to the following:

number

[0036] Here, * indicates a complex conjugate operation, and Re[.] may represent the real part of a complex quantity. The detector 30 (as part of the interferometer) uses the reference light E36(t) as the sample light E s (t) = E 351 (t) + E 352 (t) is positioned to be coupled with (t). The iNIRS system 10 may be positioned to determine one or more properties of the subject's brain tissue based on this coupling of reference light and sample light (as described in more detail below).

[0037] The transimpedance amplifier TIA304 is configured with appropriate logic, circuitry, and / or code to convert the input current I(t) into a proportional output voltage; that is, the transimpedance amplifier is configured as a current-to-voltage converter. The voltage output of TIA304 may be connected to amplifier 305 for further amplification. In some embodiments, amplifier 305 may not be necessary depending on the specific configuration of the photodetector 303.

[0038] Amplifier 305 may be used to scale the output signal to the full range of ADC 306 and limit the electronic frequency of the circuit to further maximize the SNR. This amplified voltage is then provided to ADC 306 to be digitized. ADC 306 is equipped with a digitizer having sufficient bandwidth so that the entire signal bandwidth, including time-of-flight information, can be digitized without attenuation. ADC 306 may be configured to convert a continuous-time electrical input signal into a discrete-time electrical output signal sampled at a desired sampling rate. This discrete-time signal is output in the digital signal output channel 307 and connected to the controller 40 for appropriate processing.

[0039] Each detector 30 may provide part of an interferometer, such as a Mach-Zehnder interferometer, in the system 10 (when receiving sample light and reference light from the light source). Each different photodetector 30 may be connected to the same light source 20 (each via one or more reference channels 36). The photodetectors 30 may be spatially separated from the light source 20. The sample receiving probes 35' may also be spatially separated from each other, or they may be located in the same place in tissue regions that are similar enough that the received signals can be averaged together. The reference light travels along one or more reference channels 26 / 36 so that the reference light reaches the photodetector 30 from the light source 20. The sample light travels indirectly through the brain tissue of the subject's head 2 so that the sample light reaches the photodetector 30 from the light source 20. The sample light is directed to the scalp of the subject's head 2 via one or more sample delivery channels 25'. The sample light may then travel through the subject's brain tissue into the receiving channels and into the photodetector 30. The first optical path may be, for example, an optical path from the splitter 24 through the sample delivery probe 25' to the subject's head 2, then through the sample receiving probe 35' and the sample receiving channel 35 to the detector 30. When multiple sample receiving probes are present, the multiple sample receiving channels 35 may generally have different (optical path) lengths according to various embodiments of this patent. Thus, irradiation of the subject's brain tissue may occur using different optical channels for detecting light from the subject's brain tissue.

[0040] According to various embodiments of the present invention, the iNIRS system 10 may be housed entirely or partially within a head covering (not shown) of a subject's head 2. The iNIRS system 10 may be provided, for example, in a hat / cap that can be worn by the subject on the subject's head 2. Alternatively, the covering may be a headband with attachments, or any other suitable fixture that connects at least a portion of the iNIRS system 10 to the subject's head 2. The head covering may be positioned to hold the light source 20 and the detector 30 to the scalp of the subject's head 2 with a fixing mechanism. The head covering may include, for example, a plurality of sample receiving probes 35' and / or sample receiving channels 35 and / or detectors 30. The controller 40 may be separate from the head covering (for example, connected wirelessly or wired), or the controller 40 may be provided as part of the head covering (for example, by an ASIC in the head covering that can be wired to the detector 30 and / or light source corrector 22). The covering may be configured, for example, to include a sample delivery probe 25', a channel 25, a sample receiving probe 35', and a channel 35, with other components of the system 10 located elsewhere.

[0041] The sample delivery probe 25' is positioned to be located in the subject's head 2 in order to provide imaging of a selected area of ​​the brain in the head 2. At least a portion of the probe 25' may be positioned to be spatially separated from the light source 20.

[0042] A portion of the light delivered to the subject's head 2 via the sample delivery probe 25' is received by the sample receiving probe 35'. The light travels through the subject's head 2. The light is scattered by brain tissue, resulting in delay and / or attenuation. The scattered transmit channel in the subject's head 2 may be modeled as the sum of the delayed and attenuated signals that can be received by the sample receiving probe. For example, as follows:

number

[0043] Here, y(t) may be an exemplary received signal at the probe, x(t) may be a transmitted optical signal, i may represent the i-th delay path with a delay τi, and wi may be the i-th attenuation coefficient.

[0044] In some examples, the sample receiving probes may be spatially close to each other. In other examples, the sample probes may not be substantially juxtaposed and may be separated by distance.

[0045] The detector 30 may be connected to the light source 20 via a reference receiving channel 36 (and a reference delivery channel 26). Alternatively, the detector 30 may be connected to the subject's scalp via two separate optical channels, namely a first sample receiving channel 351 and a second sample receiving channel 352 (Inset A). Each of these sample receiving channels may be connected to the subject's head 2 via a sample receiving probe 35 (only one probe is shown in Figure 1, but there are both a first and second sample receiving probe). Generally, the first sample receiving channel 351 has a different optical length than the second sample receiving channel 352. Generally, the distances between multiple light sources and receiving probes may be separated by different distances.

[0046] Hereafter, the sample light that travels along the first sample receiving channel 351 and is received by the detector 30 from the object being imaged / monitored 2 (e.g., the subject's brain) will be referred to as the "first sample light." Similarly, the sample light that is received by the detector 30 but travels along the second sample receiving channel 352 will be referred to as the "second sample light." Due to a delay line (not shown) for the second sample receiving channel 352, the second sample light takes longer to travel from the subject's scalp to the detector 30 than the first sample light. The reference light is light that does not travel through the subject's head 2 but is connected to the light source 20 via the reference receiving channel 36.

[0047] The light source 20 may be configured, for example, to provide wavelength-swept emission of light. In this regard, the light source 20 may be configured to generate a series of emission pulses of light. Between each pulse, the wavelength of light may be "swept" through a range of wavelengths. The sweep may be, for example, in the form of chirp pulses. Light is emitted at multiple different wavelengths during one pulse. The wavelength may increase or decrease continuously during one pulse (the rate of change of the wavelength may be constant or variable). The series of chirp pulses may be consecutive (for example, with zero time intervals between pulses). The light source 20 may be configured to continuously emit a series of pulses, each pulse having a wavelength sweep. However, it will be understood that there may be cases where it is not necessary for the light source 20 to provide a continuous sweep, or where it is desirable that there be no sweep at all. The light source may be adjusted stepwise, rather than continuously, so that the light source 20 emits light at different wavelengths at different time intervals (for example, separate time intervals for emission at each of the multiple wavelengths). The light source 20 may be swept unidirectionally (for example, by only increasing or decreasing the wavelength during one wavelength sweep), or the light source 20 may be swept bidirectionally (for example, by both increasing and decreasing the wavelength during one wavelength sweep). Unidirectional sweeping may be advantageous because it increases the number of detected photons per sweep.

[0048] The controller 40 may be configured to selectively control the wavelength sweep of the light source 20 via the light source modifier 22. The wavelength sweep of the light source 20 may be controlled by using the light source modifier 22 to apply a corresponding electrical signal to the light source 20. The controller 40 may be configured to use the light source modifier 22 to control the application of current / voltage to the light source 20 in order to provide a selective pattern regarding the wavelength of light emitted by the light source 20.

[0049] The light source 20 may be controlled for wavelength sweeping according to a selection pattern regarding the sweeping. The light source 20 may sweep, for example, through a selected range of the wavelength of light, and / or the light source 20 may sweep through the wavelength of light according to a selected sweeping profile (for example, linear increase, sine curve, triangle, etc.). The light source 20 may sweep, for example, according to a selected sweeping speed or a selected total sweeping time. The light source 20 may be configured to perform a wavelength sweep of light such that, during one wavelength sweep, the light is directed at the subject's brain tissue through the sample delivery channel at each of a plurality of different wavelengths (to the detector via the reference channel). The wavelength of the light emitted by the light source 20 changes over time. Therefore, the display of the time when the light is emitted from the light source 20 may be determined based on the wavelength of that light.

[0050] If the light emitted by the sample delivery probe 25 changes the frequency of the same light linearly over a certain period as f(t)=f0+δt, the observation of the frequency f1 of the light received via the sample receiving probe (for example, 35') may be used, for example, for this specific frequency sweep pattern, to estimate the delay introduced by the channel through the subject's head 2 from the following.

Number

[0051] The same frequency difference may be obtained from the balanced photodetector 303. As described above, the balanced photodetector 303 generates a signal proportional to the following.

Number

[0052] At a specific time point, the reference signal E36(t) may be, for example, cos(2πf0t), and the sample receiving signal is frequency shifted by Δf<<f0 due to the delay introduced by the channel through the subject's head 2 in combination with the frequency sweep, that is, E s(t) = cos(2π(f0+Δf)t), and by the trigonometric identity, I(t) ∝ cos cos(2πΔft) + cos(2π(2f0+Δf)t). The measured intensity I(t) includes a high-frequency term that can be removed by a low-pass filter and a frequency component at the offset frequency Δf. This may be called the beat frequency.

[0053] The light source 20 may be configured to sweep across a selected wavelength range. The light source 20 may be configured to sweep in optical frequencies in the range of 50 GHz, for example. This may allow the light source 20 to emit modulated light at multiple different wavelengths, for example, between 829.94 nm and 830.06 nm when centered at 830 nm, or between 1309.857 nm and 1310.143 nm when centered at 1310 nm. The light source 20 may be configured to sweep across a wavelength range of at least 0.025 nm, for example, at least 0.05 nm, for example, at least 0.075 nm, for example, at least 0.1 nm, for example, at least 0.11 nm (for example, around the central wavelength with respect to the light source 20). The light source 20 may have high power output, long coherence time, and a wide range of mode-hop free wavelength tuning. Since the light source 20 does not sweep in particularly large bandwidths, the light source 20 may have, for example, a relatively narrow line width and a longer coherence length.

[0054] The light source 20 of the present disclosure may be configured to provide highly coherent light, for example, substantially coherent light emission. Since light at different wavelengths changes phase at different rates, it will be understood that the light source 20 may not emit, for example, completely coherent light, nor provide wavelength-swept emission of light. The light source of the present disclosure may be controlled to sweep through a relatively narrow wavelength range compared to the absolute wavelength of the same light source. Accordingly, typically, Δf << f0. In other words, the difference between the maximum wavelength and the minimum wavelength for one wavelength sweep is relatively small compared to the same absolute wavelength. Each light source 20 may be configured to emit light (i.e., an electric field) whose phase does not change significantly over time.

[0055] The iNIRS system 10 of the present disclosure receives, for example, a first sample light, a second sample light, and a reference light, all of which originate from the same light source 20. In other words, one or more sample lights and the reference light may be processed according to various embodiments of the present patent. The light source of the present disclosure is configured to provide wavelength-swept emission of sufficiently coherent light such that the sample light and the reference light received by the photodetector 30 are in a relatively constant relative phase with respect to each other. In other words, the coherence length of the light source may be lower than the noise floor for measurements of multiple scatterings in tissue and may not reduce coherence or interference fringe contrast.

[0056] The iNIRS system may be configured, for example, to have a coherence length or range of approximately 50 m in air, and a light source having a coherence length between 50 m and 100 m (a coherence period between 166 ns and 333 ns) may be selected. This particular range is not intended to be restrictive, but rather should be understood as an example of an approximate range for the light source. The light source may be selected so that it has a coherence length that is more than twice greater than the difference in the maximum expected optical path length, for example, the coherence length may be three or four times greater. Having a light source with a coherence length much greater than the optical path length may improve the accuracy of measuring sample photons that have undergone numerous scattering interactions within the subject's brain tissue.

[0057] The iNIRS system 10 may be configured such that the light source-detector path lengths for the reference light and the sample light are different. In other words, the iNIRS system 10 is configured such that the average or expected optical path length for light traveling from the light source 20 to each detector 30 through the subject's brain tissue is different from the optical path length for light traveling from the light source 20 to the detector 30 through the reference channel 36.

[0058] As understood in the context of this disclosure, the photons of sample light directed at the subject's brain tissue may travel from the light source 20 to the photodetector 30 via a substantially infinite number of different paths through object 2. The sample light photons may undergo numerous scattering events between the sample delivery probe 25' and the sample receiving probe 35'. The iNIRS system 10 may be configured to provide neuroimaging and analysis based at least partially on activity in the subject's brain tissue that may affect the received sample light. The time of flight of the sample light photons from the light source 20 to the photodetector 30 increases, of course, as the path length taken by the photons increases. Therefore, photons that travel a longer path and penetrate deeper into the subject's brain tissue take longer to reach the photodetector 30 than photons that travel a shorter path. A longer time of flight of a sample light photon indicates that the photon has penetrated deeper into the subject's brain tissue. Sample light photons received via the sample receiving probe 35' have a longer time of flight than reference light via reference channels 26 / 36.

[0059] As understood in the context of this disclosure, the path each individual sample photon takes through the imaged object 2 (between the sample delivery probe 25' and the sample receiving probe 35') is unpredictable. However, in the presence of a large number of identical sample photons, the overall time-of-flight distribution for each sample photon may be statistically modeled. Thus, one or more expected characteristics of the time-of-flight distribution for the sample light may be known for a given light source-detector probe pair. The expected time difference between the shortest and longest flight times of the photons may be known, for example, for each light source-detector probe pair. This difference may be referred to as the delayed diffusion of the channel through the object 2. This may be based, for example, on previously observable signals relating to the earliest arriving detectable photon and the latest arriving detectable photon. In other words, a known maximum expected delay for the resulting time-of-flight distribution relating to the sample photons may exist for any given light source-detector pair.

[0060] The iNIRS system 10 may be configured such that the interference pattern measured by the detector 30 can be sampled fast enough to allow measurement of the temporal interferogram generated at each sweep of the light source 20, so that the light source 20 can be wavelength swept.

[0061] The resulting temporal interferogram obtained by the detector 30 may include multiple different beat frequencies due to the differences in various wavelengths contained in signals 36, 351, and 352, the differences in time of flight through an arbitrary associated delay line, and the different times of flight of the detected photons passing through the sampled object 2. The frequency components of the interferogram may encode the time of flight of the detected light passing through the sampled object for each sample light signal 351 and 352. Higher beat frequencies may correspond to photons with longer times of flight.

[0062] The iNIRS system 10 may be configured to acquire a digital representation of each resulting interferogram at a digital signal output 307. For example, the detector 30 may include an analog-to-digital converter ("ADC") 306 configured to acquire discrete-time interferogram data from each interferogram provided by the detector 303. Each acquired interferogram may be Fourier analyzed (e.g., using FFT or iFFT) to obtain a representation of the time-of-flight distribution ("DTOF") for sample photons incident on the photodetector. For example, this operation may be performed in the controller 40. Each determined DTOF may provide a distribution showing the time of flight for sample photons that may be incident on the photodetector 30 at a given moment.

[0063] If the detector 30 processes multiple received sample signals having different delay lengths, a single interferogram may be processed to obtain multiple DTOFs.

[0064] When the detector 30 processes multiple received sample signals, multiple interferograms and multiple DTOFs may be acquired.

[0065] As is known to those skilled in the art, the DTOF obtained by a particular configuration of the iNIRS system 10 may provide time-of-flight data relating to the absolute optical scattering characteristics and absolute optical absorption characteristics of the object being sampled 2. The absolute optical scattering coefficient and absolute optical absorption coefficient may be calculated from the DTOF. This calculation may be performed in the controller 40.

[0066] As is known to those skilled in the art, the iNIRS system 10 may be configured to repeatedly provide DTOF measurements over time at a given rate. The change in DTOF over time may relate to changes in the absolute optical scattering coefficient and the absolute optical absorption coefficient, which can be calculated in the controller 40, and therefore to changes in these coefficients over time.

[0067] An index of blood flow may be obtained by measuring the temporal autocorrelation of signals derived from DTOFs at multiple DTOFs. When the autocorrelation of DTOFs decreases more rapidly in amplitude (i.e., when signals lose correlation more quickly), the observed blood flow is higher. The acquired data of autocorrelation decay may be called G1 and may be a function of the time of flight τs, the time between consecutive DTOF measurements td, and the autocorrelation lag τd.

[0068] Referring to inset B in Figure 1, the reference receiving channel 26, the sample receiving channel 35, the beam splitter 60, the optical assemblies 50 and 55, the camera sensor 70, and the signal output 307 are shown.

[0069] The detector 30 may include a camera sensor 70. A reference signal via the reference delivery channel 26 may be delivered to the beam splitter 60 through an optical assembly 50, which may include a collimating lens. A sample signal via the sample delivery channel 35 may be delivered to the beam splitter 60 through an optical assembly 55, which may include a collimating lens. In such a case, the sample signal received via the sample receiving channel 35 may include multiple modes and a larger beam width representing multiple speckles. Therefore, the image received by the sample receiving channel 35 may interfere with the reference signal via the reference delivery channel 26 in the beam splitter 60. The interference signal output by the beam splitter 60 may be received by the camera sensor 70. The camera sensor 70 may include multiple photosensitive locations, for example, pixels in an m × n array. Each pixel may be operable to record speckles received from the beam splitter 60. As will be apparent to those skilled in the art, the physical mechanism of the photosensitive area in the camera sensor 70 may also be square, circular, elliptical, or any other suitable mechanism for the photosensitive area.

[0070] In the camera-based method shown in inset B, multiple modes of the detected light (via 35) may interfere with the same reference light (via 26) for different pixels of the camera sensor 70. Accordingly, the signal from each sensor pixel may be processed in a manner similar to the signal received in the photodiode. Thus, for each pixel p, the light intensity may be obtained as follows:

number

[0071] The camera sensor-based system may be implemented with either an adjustable laser as described above for inset A, or an unmodulated laser; that is, the system may not use a specific sweep pattern and may emit at a constant wavelength.

[0072] If the iNIRS system 10 is equipped with a camera sensor 70, the temporal sampling rate of the camera sensor may not be sufficient to generate a temporal interferogram. The iNIRS system 10 may then be configured such that, without first measuring the temporal interferogram, the interference pattern may be repeatedly sampled by the camera sensor at a rate sufficient to enable the measurement of the autocorrelation decay of the electric field measured at each pixel due to the movement of scatterers in the object being sampled. The acquired autocorrelation decay data may be referred to as G1.

[0073] If the iNIRS system 10 is equipped with a camera sensor 70, the temporal sampling rate of the camera sensor may not be sufficient to measure temporal speckle intensity fluctuations and thus generate autocorrelation decay measurements. The iNIRS system 10 may then be configured such that an index of blood flow can be determined from the spatial speckle statistics of the captured camera frame, so that interference patterns can be repeatedly sampled by the camera sensor. This technique may be referred to as speckle contrast spectroscopy.

[0074] The speckle contrast may be proportional to the lag-time integral autocorrelation, which can provide an indicator of blood flow; that is, the speckle contrast is as follows:

number

[0075] As mentioned above, the camera sensor-based system may be implemented without using a modulated / tuned laser, in which case the system may not provide time-of-flight resolution. Then, G1(p, τ) may be acquired for each pixel p of the camera sensor with a delay τ due to the autocorrelation of I(p, τ, t), and then G1(p, τ) may be averaged across all pixels p to acquire G1(τ). The uncorrelated velocity of G1(τ) may be directly related to the blood flow velocity (volume per unit time).

[0076] As described above, the controller 40 may be configured to determine a cerebral blood flow index ("CBFi") relating to the subject's brain tissue. CBFi may provide a representation of movement occurring within the subject's brain. CBFi may provide, for example, a relative index describing movement occurring within the subject's brain. Movement may be attributable to blood moving (e.g., flowing) through the subject's brain. CBFi may provide a relative index. CBFi may include a representation of the mean square displacement of scatterers in bulk tissue, measured in units of cm^2 / s. CBFi relating to a given volume within the subject's brain may provide, for example, a representation of movement occurring within that volume. Since this movement may be attributable to blood movement, CBFi may provide a representation of average blood flow in terms of cerebral blood flow velocity (CBFV, cm / s) and blood tracer clearance (ml / 100g / min). An increase in the value of CBFi may indicate, for example, an increase in velocity occurring within the brain volume.

[0077] The controller 40 may be configured to acquire cerebral blood flow data (i.e., data that includes a representation of blood flow within the subject's brain tissue) which may include a representation of one or more pulses of blood flow through the subject's brain tissue. The controller 40 may also determine a CBFi for the subject's brain tissue, and pulses of blood flow through the subject's brain tissue may be observable in the determined CBFi for the subject's brain tissue. An example of such a determined CBFi may be shown in Figure 2.

[0078] Referring now to Figure 2, which shows an example of a determined CBFi unfolding in one cardiac cycle in the tissue of a subject's brain 2. One pulse is shown in Figure 2. The pulses may occur in a periodic and repeating manner (each individual pulse may vary in size and shape, and the frequency at which the pulse occurs may also vary). Typically, each pulse may have a diastolic (lower) CBFi value and a systolic (higher) CBFi value, as shown. During one pulse, the CBFi may start at its diastolic value (optionally chosen) and then rapidly increase to its systolic value (P1). The CBFi then decreases back down from the systolic value to its diastolic value, but as the CBFi decreases from the systolic value to the diastolic value, the CBFi may temporarily increase at least once (or at least begin to decrease more slowly). In other words, local peaks may exist, as shown in Figure 2 by P2 and P3. Once the heart rate returns to its diastolic value, the process may be repeated again based on the subject's heart rate.

[0079] The iNIRS system 10 may be configured to use optical interferometry to acquire measurement signals, and the controller 40 of the iNIRS system 10 may acquire cerebral blood flow data from the measurement signals that show this behavior regarding blood flowing through the subject's brain tissue (for example, the controller 40 may acquire the temporal evolution of CBFi values ​​as shown in Figure 2).

[0080] The controller 40 may be configured to process cerebral blood flow data to determine the ICP representation for the subject's brain tissue. The controller 40 may be configured to determine the ICP based on one or more characteristics of a pulsatile waveform relating to blood flow. In this regard, the controller 40 may use any of a number of suitable features. The features may describe, for example, the shape of the pulsatile waveform or other indicators relating to the pulsatile waveform.

[0081] The volume of a subject's cranial tissue may be constrained by the subject's skull. Here, cranial tissue may refer to tissue within the skull and subdural space, and may include the brain, cerebrospinal fluid, and blood. Typically, since the skull may be very rigid, the subject's cranial tissue may be constrained by the fixed internal volume of the skull. Since this internal volume does not change, if the pressure within the skull increases, blood flow through the material in that internal volume may become restricted. For blood to flow through the subject's cerebral vascular structure, either blood, cerebrospinal fluid, or brain tissue must be moved from the fixed volume of the skull, or the pressure within the fixed volume of the skull must increase. On a short time scale, this may result in intracranial flow that fluctuates with the heartbeat. The shape of the waveform of this flow may depend on the intracranial counterpressure, which is the ICP.

[0082] Therefore, an increase in ICP may restrict blood flow within the skull. In situations where ICP can remain excessively high for too long, blood flow may be severely restricted, potentially leading to insufficient oxygenation of brain tissue, which can result in brain damage.

[0083] The controller 40 may be configured to use cerebral blood flow data to determine how to display the effect of ICP on pulsatile waveforms relating to blood flow through the subject's brain tissue. The controller 40 may achieve this by observing changes in specific characteristics of the pulsatile waveforms of CBFi and / or arterial blood pressure (ABP).

[0084] Specifically, arterial blood pressure (ABP) may refer to blood pressure and blood flow data that can be obtained from locations other than within the skull. Because ABP can be measured outside the skull, it is not under the control of ICP and may, but may not be, reflect the characteristics of blood flow and blood pressure within a subject. Therefore, ABP data may function as additional and / or reference data related to ICP and / or blood flow.

[0085] The controller 40 may be configured to identify features as changes in a pulsatile waveform, such as changes in the shape of the waveform and / or changes in the maximum / minimum values of the waveform (e.g., P1, P2, P3), and this information may be used to determine the display of the ICP of the subject. The pulsatile flow waveform has a characteristic peak (systole, F SYS , P1) representing the maximum cerebral flow during the cardiac cycle, and a trough (diastole, F DIA ) representing the minimum value. During the cardiac cycle, the pulsatile flow waveform may include several distinguishable cardiac features that most prominently have three peaks P1, P2, P3 corresponding to the "shock wave", "tidal wave", and "dicrotic wave" of the cardiac cycle. In some states where the brain may be unable to effectively regulate its own volume (i.e., cerebral autoregulation is disrupted), the ICP may be very high (e.g., >20 mmHg), and the characteristic shape of the pulsatile flow waveform may change as a function of the ICP level. One particular change in the CBFi flow waveform is the normalized difference between the diastolic minimum and the systolic maximum (F SYS -F DIA / F BAR) may also be used. This relationship may be called the pulsation index (PI), and in many cases, it may show a linear relationship with ICP. Other specific features that may be expected to change with ICP are the slope and integral of the CBFi waveform features, as shown in Figure 2, or the amplitude of individual waveform samples at any particular phase of the pulse period. The controller may also be configured to extract features related to ICP by comparing ABP measured using an external device with CBFi obtained through interference measurements. These features may include the absolute values ​​of CBFi and ABP, any combination of these two waveforms including the phase relationship between CBFi and ABP, or the amplitude, protrusions, widths, and statistical features of peaks within the pulse period, including P1, P2, P3. Features may be shown over multiple pulse periods; for example, P1 may be modulated at the breathing frequency, and the intensity of this modulation may be used as a feature. Depending on the specific application, it may be desirable to synchronize the ABP and CBF measurements for further processing.

[0086] The controller 40 may also be configured to extract features from the CBFi waveform alone or in combination with the ABP waveform using an "unsupervised" method that does not require a human to explicitly identify features.

[0087] Referencing Figure 3a, exemplary types of ICP estimation, specifically types 1–3, are shown. All exemplary processes may include CBFi, ABP, and extracerebral blood flow input. Type 1 may represent a more traditional process, as previously described. The Type 2 process may be based on unsupervised feature selection from input data. The selected features may then be mapped to ICP estimation using a regression model. The Type 3 process may allow for skipping explicit feature extraction, as will be discussed further later.

[0088] Unsupervised methods for feature extraction may include any form of matrix decomposition or any neural network engaged in so-called representation learning, and may therefore be designed to learn common structures within these time-series inputs. An example of how matrix decomposition can be used for feature extraction is by applying principal component analysis (PCA) to a matrix (m, n), where "m" may be a dimension of length equal to the number of pulse waveforms (samples) and "n" may be a dimension of length equal to the number of waveform features. PCA may also result in lower-dimensional embeddings of new features "k", where each "k" may be an affine transformation of the original features "n", where k <= n. Each new feature "k" may be ranked by the variance that it describes in the original dataset of "m" samples, and the new features "k" may be used in a regression model that maps ICP to the same features. In other words, PCA may provide lower-dimensional approximations to the input data. Since the feature "k" may exist in a dimensional space that can be rotated from the original space which may have dimensions labeled by each feature, the feature "k" found using matrix decomposition such as PCA does not necessarily have to be readily described in relation to any physical properties of the CBFi pulse waveform. Therefore, the PCA feature vector may be dimensionless and abstract, but may be remapped to the original feature space. Thus, the most significant PCA feature may take the form of a vector of weights applied to each input feature, such as a1* projection (P1), a2* time (P2), a3* (width (P1) - time (P3)), where the set {a i} may also be a set of weight coefficients. An example of a neural network architecture that can implement representation learning capable of learning the same features may be an autoencoder network. The network may transform the input features into a typically lower-dimensional embedding space through one or more network layers, while minimizing the error in predicting the input of the network itself. For example, if a pulse waveform is represented by 100 time points, the autoencoder network may embed the 100 features into a space having, for example, 8 dimensions, where each 8-dimensional vector may contain enough information about the network to reproduce the original input with acceptable accuracy. The 8-dimensional representation may then be used as feature input to a regression model that maps the ICP to the features. As will be apparent to those skilled in the art, any number of features and any number of dimensions may be used in practical applications, and this disclosure should not be construed as limiting. According to various embodiments of this patent, the iNIRS system 10, specifically the controller 40, may be implemented using any one of a number of machine learning techniques. As an alternative to using simple linear regression to associate ICPs with CBFi features, machine learning-based regression models may enable more complex mappings between ICPs and CBFi, providing better estimation accuracy.

[0089] The desired machine learning method may predict ICP based on various predictor features, such as those described above. The predictive model may aim to minimize the error between the predicted ICP and the feature combination between training and validation, so that the trained model can then predict ICP based solely on the data input. The model may take the form of a regression model, which associates a continuous set of features with a continuous target ICP value, or classifies the input into a discrete but numerical set of target ICP values, and thus performs regression using multi-class classification. The regression model may vary in complexity and may also take the form of an ensemble model that accumulates predictions from a number of individual regressors or classifiers. An example of an ensemble model is a random forest regressor, which may be an ensemble of many decision trees, each of which may associate one or more data features with a target variable, in this application, ICP. In this example, the random forest regressor is constructed to include, for example, 100 decision trees, each of which may be, in this example, a random subset of, for example, 153 input features, each containing, for example, 50% random selections, such as 153 input features, each containing, for example, 100 decision trees, each containing, for example, 100 decision trees containing, each

[0090] Furthermore, the controller 40 may be configured to infer ICP through a pre-trained model that does not require separate feature extraction steps, as shown in the Type 3 process in Figure 3a. This process may be referred to as a sequence-sequence regression model method. This class of models may directly learn the correspondence between pulse waveforms (CFBi alone, or CBFi and ABP together) and the ICP target. This form of model may be called “supervised representation learning,” and may directly map CBFi values ​​from time-series input pulse waveforms to ICP values ​​without explicit intermediate representations in the feature space. An example of this type of model is, for example, a long-short-term memory (LSTM) neural network that can accept a continuous sequence of 10 pulse waveforms and predict ICP values ​​from these pulse waveforms. In this example, the model input may be two vectors, one containing, for example, 1500 CBFi values ​​(e.g., 150 values ​​for each of 10 pulse waveforms) and the other containing, for example, 1500 ABP values, and the target may be a single ICP value, which may be, for example, the average ICP measured over the same period. An LSTM network is a class of recurrent neural networks that can process long sequences of data and maintain and update a historical representation of that sequence, which may be called the “cell state” of the LSTM. The LSTM network may achieve this by updating “hidden states” that can influence the “cell state” through a set of filters, which can be read at each individual point in time and implemented as a recurrent neural network. In this way, the LSTM network may be able to simultaneously learn the statistical structure of its sequence input (e.g., the periodicity of the statistical structure) and map the input in a way that predicts the target output with (in a sense) minimum error.

[0091] Another example of determining ICP based on cerebral blood flow data and extracerebral blood flow data is shown in Figure 3b. In Figure 3b, cerebral blood flow data may be plotted relative to extracerebral blood flow data, and these cerebral and extracerebral blood flow data may be shown as CBFi on the y-axis and ABP on the x-axis. The two data plots (i.e., data for cerebral blood flow and extracerebral blood flow) may be aligned in the graph, for example, so that the linear regions of the data plots substantially overlap each other (or, if the data plots do not completely overlap each other, at least come close to each other). As can be seen in Figure 3b, the linear region extends from where the line intersects the x-axis (labeled "CrCP" for critical occlusive pressure) to where the data branches and a loop region exists (labeled "heavy beat notch"). As can be seen in Figure 3c, the curve intersects the x-axis at a positive ABP value (i.e., the CBFi value is 0). The controller 40 of the iNIRS system 10 may be configured to process brain data and extra-brain data in order to identify the value at this point, i.e., CrCP.

[0092] CrCP may represent critical occlusion pressures related to blood vessels in the subject's brain tissue. As mentioned above, CBFi may provide information about the movement of blood flow through blood vessels in the subject's brain tissue. When the CBFi value decreases, this may indicate that there is little blood movement in the subject's brain tissue (for example, a measurement of zero for CBFi may indicate no blood flow at all). The intersection on the x-axis in the graph of Figure 3d represents the value for arterial blood pressure where CBFi is zero (i.e., there is no blood flowing through the blood vessels in the subject's brain tissue). This is called critical occlusion pressure because it is the pressure at which small blood vessels in the brain "occlude". In other words, the value for CrCP represents a state in which the pressure in the blood vessel is low enough to be occluded (by external pressure applied to the blood vessel, i.e., ICP).

[0093] Controller 40 does not need to generate graphs such as Figure 3b, and it will be understood that such graphs are shown to illustrate the method. Controller 40 may be configured to compare extracerebral blood flow data (e.g., blood pressure data for one or more of the subject's veins / arteries outside the subject's brain tissue) with cerebral blood flow data (e.g., CBFi values ​​from the subject's brain tissue). Based on this comparison, Controller 40 may be configured to identify the corresponding pressure at which a blood vessel in the subject's brain tissue is occluded (e.g., within the subject's skull). Controller 40 may be configured to determine the display of ICP based on this CrCP value (e.g., by determining the required pressure in the subject's brain to occlude a cerebral blood vessel). This CrCP value may also be used as a feature in either a Type 1 ICP estimation method or a Type 2 ICP estimation method, or the CrCP value may be used as a continuous input to any Type 3 ICP estimation method.

[0094] Where extracerebral blood flow data is available, it will be understood that, in addition to deriving the data from interference data, any other suitable device may be used to acquire the data. For example, the iNIRS system 10 may be provided in combination with separate sensing elements that can be configured to acquire the data. The separate sensing elements may be configured, for example, to measure blood pressure in a separate area of ​​the subject's body, such as in one of the subject's limbs, such as in the arm or the subject's hand / fingers. The sensing elements may be configured to acquire data that includes a representation of one or more characteristics of the pulsatile waveform in the same area of ​​the subject's body (including, for example, diastolic and / or systolic values ​​and / or mean values). The sensing elements may be configured, for example, to acquire a representation of how blood pressure changes over time during the pulsatile waveform of blood flowing through the subject's veins / arteries. This data may be referred to as ABP data throughout this disclosure.

[0095] Referring here to Figure 4a, a system similar to that shown in Figure 1 is shown. The same reference numerals shall refer to similar elements as in Figure 1. Figure 4a may be used to illustrate one method of obtaining ABP data from an interfering system such as iNIRS system 10.

[0096] As described above, the iNIRS system 10 may be configured such that light from a light source 20 (not shown) is separated (by an optical splitter 24) into a sample delivery channel 25 and a reference delivery channel 26. The photodetector 30 may then receive the reference light (from the reference delivery channel 26) and the sample light (from the sample receiving channel 35). The sample delivery probe 25' and the reference delivery probe 35' are shown.

[0097] Figure 4a may show different layers of the subject's head. The upper layer may be the scalp surface 101. The second layer below the scalp surface may be scalp tissue 102. The scalp tissue may comprise multiple arteries and veins (not shown). The subject's skull 103 may be beneath the scalp tissue, and the brain tissue 104 may be located within the skull. The veins and arteries in the subject's scalp tissue are relatively unconstrained by the skull and ICP. That is, in this region of the subject's body, the veins / arteries are not hindered in expanding in volume by the skull (whereas the blood vessels in the subject's brain tissue are hindered).

[0098] For the photons of the light to reach the subject's brain tissue from the light source 20 (through the sample delivery channel 25 and sample delivery probe 25'), the photons must penetrate the subject's scalp surface, scalp tissue, and skull. For the photons of the light to reach the veins / arteries in the subject's scalp tissue, the photons must penetrate only the subject's scalp surface (and the relevant parts of the scalp tissue). Two exemplary photon pathways, namely a shallow photon pathway 202 and a deep photon pathway 204, are shown for the sample light reaching the detector 30 from the light source 20. The deep photon pathway between the light source 20 and the photodetector 30 is much longer than the shallow photon pathway (because the sample light travels deeper into the subject's brain tissue). Numerous scattering events occurring within the subject's head are present for both photon pathways. In this example, photons traveling along the superficial photon pathway interact with at least one blood-carrying region (e.g., a vein or artery) in the subject's scalp tissue, while photons traveling along the deep photon pathway interact with at least one blood vessel in the subject's brain tissue.

[0099] The light source 20 emits a much larger number of photons, and the path these photons take through the subject's head from the light source 20 to the detector 30 is altered. The sample delivery probe 25' and the sample receiving probe 35' are spatially positioned on the subject's scalp so that at least some of the sample photons arriving at the detector 30 from the light source 20 travel along a shallow photon path and some travel along a deep photon path. The probes are, for example, sufficiently spatially separated so that some deep photon paths occur.

[0100] Figure 4b shows the time-of-flight distribution for sample photons received by detector 30. As shown, the peak number of sample photons may occur at relatively short time-of-flight intervals, and then the number of arriving photons decreases with longer time-of-flight intervals. As mentioned above, the time-of-flight for sample photons roughly corresponds to the penetration depth of the sample photons. That is, photons entering from deeper depths have longer time-of-flight intervals than photons entering from shallower depths.

[0101] Accordingly, the controller 40 may measure surface blood flow data by, for example, separating shallow-entering photons corresponding to extracerebral blood flow from deeper-entering photons corresponding to intracranial blood flow based on time of flight.

[0102] The controller 40 may be configured to process cerebral blood flow data and extracerebral blood flow data separately. For example, the controller 40 may be configured to acquire time-of-flight data (including, for example, the measured time-of-flight distribution) and separate this time-of-flight data into different data streams, namely, (i) an extracerebral data stream (regarding shorter-time-of-flight photons thought to be associated with scalp tissue) and (ii) a brain data stream (regarding longer-time-of-flight photons thought to be associated with brain tissue). Each data stream may be processed independently in the manner described above to acquire blood flow index data including a representation of one or more pulses of blood passing through a relevant region of the subject's body (e.g., through veins / arteries in the subject's scalp or blood vessels in the subject's brain tissue). In this manner, the extracerebral data stream may be used in conjunction with, or as a substitute for, the arterial blood pressure data stream used in the type 1, 2, and 3 ICP estimation methods.

[0103] In the example described above in relation to the figure, the iNIRS system 10 uses one light source and one detector (i.e., only one light source-detector channel exists). However, this should not be considered limiting. The iNIRS system 10 may include multiple light sources and / or multiple photodetectors. Each light source may be connected to multiple other detectors (e.g., via multiple reference channels). Multiple photodetectors may be juxtaposed (e.g., the same multiple photodetectors may be provided to the same area on the subject's scalp and / or provided in close proximity to each other on the subject's scalp). Different photodetectors may then be arranged to detect photons of sample light from the same light source that has traveled through similar areas of the subject's brain tissue. The controller 40 may be configured to combine data from different photodetectors (e.g., average the data to provide a single combined reading for the subject's brain tissue). For example, the controller 40 may be configured to determine cerebral blood flow index data (optionally, extracerebral data as well) for the subject's brain tissue based on data acquired using multiple photodetectors. This mechanism may result in improved signal-to-noise ratio for measurements acquired using the iNIRS system 10.

[0104] The iNIRS system 10 may include multiple light source-detector channels associated with different regions of the subject's brain tissue. For example, the multiple photodetectors may be located in different places on the subject's scalp (these multiple photodetectors may be connected to the same or different light sources). The controller 40 may be configured to use the multiple detectors of the iNIRS system 10 to acquire multiple ICP measurements (as described above, for each of the multiple detectors). The controller 40 may then determine an ICP value for the subject based on the multiple ICP measurements from the different detectors. This mechanism may increase the reliability of the measurements, for example, because ICP may be relatively constant in different regions of the subject's brain tissue.

[0105] In the example described above, the iNIRS system 10 may use one light source-detector channel to acquire both cerebral blood flow data and extracerebral blood flow data. However, the iNIRS system 10 does not need to acquire extracerebral blood flow data, and additional sensors may be provided, for example, to acquire extracerebral blood flow or blood pressure values. Furthermore or alternatively, different iNIRS light source-detector channels may be used to acquire extracerebral data, for example, here, the light source and detector are positioned closer to each other on the subject's scalp (for example, to increase the selection of sample photons that travel from the light source to the detector in a shorter time). The iNIRS system 10 may, for example, comprise an extracerebral light source-detector channel, i.e., one light source, one detector positioned to measure extracerebral blood flow, and one or more (e.g., multiple) photodetectors positioned to measure cerebral blood flow.

[0106] In the example described above, extracerebral blood flow data may be acquired using the iNIRS system 10 to measure the characteristics of blood flow in the subject's scalp tissue. However, this should not be considered limiting, as other areas may be used for extracerebral blood data. For example, the iNIRS system 10 may be configured to measure characteristics of the subject's neck, ears, frontal lobe, and even areas further away from the subject's brain tissue (e.g., if multiple light source-detector channels are used). The iNIRS system 10 may be configured to acquire blood flow data for areas of the subject's body (not within the subject's skull) where the monitored blood flow does not have significant external constraints (i.e., is traveling through areas where the blood is not sufficiently compressed). Extracerebral blood may, for example, flow in the veins / arteries beneath the subject's skin in areas not under significant compression by surrounding material.

[0107] Figure 4c shows two different photon time-of-flight distributions for short-time source-detector separation 405 and long-time source-detector separation 410.

[0108] As shown in inset B, in a system configuration 10 having a camera sensor and, for example, not modulating the light source 20, the received sample light may not be easily separable by time of flight. However, by using multiple light source-detector pairs and / or adjusting the separation distance between the light source-detectors, it may be possible to select a longer or shorter photon path to dominate the received signal measurement.

[0109] Referring to Figure 4C, the time-of-flight distribution of photons for short light source-detector separation 405 is shown. In this case, the separation between light source 25' and detector 35' may be, for example, 5 mm. Also shown is the time-of-flight distribution of photons for long light source-detector separation 410, where the separation between light source 25' and detector 35' may be, for example, 20 mm. The plots of 405 and 410, such as the separation distances of light source-detector for 405 and 410, are not to scale but are merely illustrative.

[0110] For short light source-detector separation 405, most photons travel along a short, shallow path between the light source 25' and the detector 35'. Accordingly, most photons travel through extrabrain tissue, as shown by the shallow photon path 202 in Figure 4A. For long light source-detector separation 410, more photons travel along deeper photon paths, as shown by the deep photon path 204 in Figure 4A. As shown in Figure 4C, the average travel time of photons, and therefore the average path depth, may increase with light source-detector separation. Accordingly, by appropriately selecting short light source-detector separation, the measurements may be dominated by extrabrain photon paths, even when direct time-of-flight selection is not available.

[0111] A configuration of one system 10 may, for example, use a single light source probe 25' and two sample light receiving probes 35', where the first sample light receiving probe 35' is at a short light source-detector separation distance where the measurement can be dominated by an extrabrain photon pathway. Another detector with a longer light source-detector separation distance is used for measurements dominated by longer photon pathways, and therefore by intrabrain blood flow.

[0112] The iNIRS system 10 may include multiple light sources. The light sources may emit light in different wavelength ranges. The iNIRS system 10 may be configured to determine ICP based on measurements obtained using light in each of the multiple different wavelength ranges. The iNIRS system 10 may include a first light source configured to emit light in a wavelength range above the isosbestic wavelength of oxygen measurement, and a second light source configured to emit light in a wavelength range below the isosbestic wavelength of oxygen measurement. The controller 40 may be configured to determine ICP based on the received optical signals from both the first and second light sources. For example, one of the light sources may emit light in a wavelength range more suitable for absorption by oxygenated hemoglobin, and the other light source may emit light in a wavelength range more suitable for absorption by deoxygenated hemoglobin, thereby enabling greater reliability for measurements where blood oxygenation values ​​change.

[0113] In the context of this disclosure, it will be understood that the examples described herein are not intended to be limiting. In fact, the examples describe specific potential ways of implementing the claimed technology. For example, iNIRS system 10 is described together with a set of optical cables providing a channel and a probe connecting the channel to the scalp of a subject. However, it will be understood that the probe itself may be part of the optical channel, or the probe may not be provided at all. Similarly, the reference channel mechanism is intended only to indicate that reference light is delivered from the light source to the photodetector via the optical channel (rather than through the subject's brain tissue). Each light source may include, for example, one reference channel for each photodetector, where the reference channel directly connects the light source to the photodetector. In this case, there may be no reference connection in the system at all. Alternatively, the reference light may be transmitted in a common reference light channel, where a portion of the reference light is received by each of the photodetectors from the common reference light channel. The light source may also be positioned to deliver the light to one of several different locations on the subject's scalp. The light source may be connected to, for example, multiple different sample delivery channels, each extending toward the subject's scalp (e.g., from a light splitter).

[0114] From the above description, it will be understood that the examples shown in the figures are merely illustrative and may include features that may be generalized, omitted, or replaced as described herein and in the claims. Generally, referring to the drawings, it will be understood that schematic functional block diagrams are used to illustrate the functions of the systems and apparatus described herein. In addition, processing functions may also be provided by devices supported by electronic devices. However, it will be understood that functions do not need to be divided in this manner and should not be interpreted as suggesting any specific hardware structure other than those described below and in the claims. One or more functions of the elements shown in the drawings may be further subdivided and / or distributed throughout the apparatus of this disclosure. In some examples, the functions of one or more elements shown in the drawings may be integrated into a single functional unit.

[0115] As will be understood by those skilled in the art in the context of this disclosure, each of the examples described herein may be implemented in a variety of different ways. Any feature of any aspect of this disclosure may be combined with any of the other aspects of this disclosure. For example, a method aspect may be combined with an apparatus aspect, and a feature described in reference to the operation of a particular element of an apparatus may be provided in a way that does not use that particular type of apparatus. In addition, unless it is expressly stated that some other feature is essential for the operation of that feature, each feature relating to each example is intended to be separable from the feature that each of those features is combined with and described. Each of those separable features may, of course, be combined with any of the other features of the example in which each of those separable features is described, or with any of the other features or combinations of features relating to any of the other examples described herein. Furthermore, equivalents and modifications not described herein may be adopted without departing from the present invention.

[0116] Certain features of the methods described herein may be implemented in hardware, and one or more functions of the apparatus may be implemented in method steps. Furthermore, it will be understood in the context of this disclosure that the methods described herein do not necessarily have to be performed in the order in which they are described, nor do they necessarily have to be performed in the order shown in the drawings. Thus, aspects of this disclosure described with reference to a product or apparatus are also intended to be implemented as methods, and vice versa. The methods described herein may be implemented as computer programs or hardware, or any combination thereof. Computer programs include software, middleware, firmware, and any combination thereof. The programs may be provided as signals or network messages, or recorded on computer-readable media such as tangible computer-readable media capable of storing computer programs in a non-temporary form. Hardware includes computers, handheld devices, programmable processors, general-purpose processors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), and arrays of logic gates.

[0117] Any controller in this disclosure may be implemented using fixed logic, such as an assembly of logic gates, or programmable logic, such as software and / or computer program instructions executed by a processor. The controller may comprise a central processing unit (CPU) and associated memory connected to an image processing unit (GPU) and associated memory with the GPU. Other types of programmable logic include programmable processors, programmable digital logic, such as field-programmable gate arrays (FPGAs), tensor processing units (TPUs), erasable programmable read-only memory (EPROMs), electrically erasable programmable read-only memory (EEPROMs), application-specific integrated circuits (ASICs), or any other types of digital logic, software, code, electronic instructions, flash memory, optical discs, CD-ROMs, DVD-ROMs, magnetic or optical cards, other types of machine-readable media suitable for storing electronic instructions, or any suitable combination of the above. In particular, any controller in this disclosure may be provided by an ASIC.

[0118] Other examples and modifications of this disclosure will be obvious to those skilled in the art in the context of this disclosure.

Claims

1. A method for estimating intracranial pressure using a regression model, The aforementioned method, To generate optical interferometric measurements related to cerebral blood flow, Estimating intracranial pressure using the regression model based on the optical interference measurements of cerebral blood flow and one or more features of extracerebral blood flow or extracerebral blood pressure, including, A method characterized by the following:

2. The estimation of intracranial pressure is as follows: Unsupervised feature selection, including, The method according to claim 1.

3. The aforementioned unsupervised feature selection is, Low-dimensional matrix decomposition of the input data matrix, including, The method according to claim 2.

4. The aforementioned low-dimensional matrix decomposition is principal component analysis. The method according to claim 3.

5. The estimation of intracranial pressure is as follows: Representation learning neural networks, including, The method according to claim 1.

6. The aforementioned representation learning neural network is an autoencoder network. The method according to claim 5.

7. The aforementioned representation learning neural network is Random forest regressor, Equipped with, The method according to claim 5.

8. The estimation of intracranial pressure is as follows: Sequence-sequence regression model method, including, The method according to claim 1.

9. The aforementioned sequence-sequence regression model method is implemented as a long-short-term memory (LSTM) neural network. The method according to claim 8.

10. The generation of the aforementioned optical interference measurement values ​​uses multiple sensors and / or camera sensors. The method according to claim 1.

11. One or more of the features extracted from the measurement of extracerebral blood flow are estimates of arterial blood pressure (ABP) data. The method according to claim 1.

12. To generate the aforementioned optical interference measurements, near-infrared wavelengths are used. including, The method according to claim 1.

13. To generate the aforementioned optical interference measurement values, the near-infrared wavelength is changed in the sweep pattern. including, The method according to claim 1.

14. To estimate one or more characteristics of the extracerebral blood flow, select optical interferometric measurements with a short flight time. including, The method according to claim 1.

15. To estimate the aforementioned multiple characteristics of cerebral blood flow, long-flight optical interferometric measurements are selected. including, The method according to claim 1.

16. To estimate the intracranial pressure using supervised learning, including, The method according to claim 1.

17. For the aforementioned supervised learning, a random forest regressor is used. including, The method according to claim 16.

18. A system that estimates intracranial pressure using a regression model, The aforementioned system, One or more circuits that generate optical interferometry measurements relating to cerebral blood flow, and estimate intracranial pressure using the regression model based on the optical interferometry measurements relating to cerebral blood flow and one or more features of extracerebral blood flow or extracerebral blood pressure, Equipped with, A system characterized by the following features.

19. The one or more circuits described above include unsupervised feature selection and are configured to estimate intracranial pressure. The system according to claim 18.

20. The aforementioned unsupervised feature selection is, Low-dimensional matrix decomposition of the input data matrix, including, The system according to claim 19.

21. The aforementioned low-dimensional matrix decomposition is principal component analysis. The system according to claim 20.

22. The one or more circuits include a representation learning neural network and are configured to estimate intracranial pressure. The system according to claim 18.

23. The aforementioned representation learning neural network is an autoencoder network. The system according to claim 22.

24. The aforementioned representation learning neural network is Random forest regressor, Equipped with, The system according to claim 22.

25. The one or more circuits include a sequence-sequence regression model method and are configured to estimate intracranial pressure. The system according to claim 18.

26. The aforementioned sequence-sequence regression model method is implemented as a long-short-term memory (LSTM) neural network. The system according to claim 25.

27. The one or more circuits are configured to generate optical interference measurements using multiple sensors and / or camera sensors. The system according to claim 18.

28. One or more of the features extracted from the measurement of extracerebral blood flow are estimates of arterial blood pressure (ABP) data. The system according to claim 16.

29. The one or more circuits are configured to use near-infrared wavelengths to generate the optical interference measurements. The system according to claim 16.

30. The one or more circuits are configured to change the near-infrared wavelength in a sweep pattern in order to generate the optical interference measurement. The system according to claim 18.

31. The one or more circuits are configured to select short-time-of-flight optical interferometric measurements in order to estimate one or more features of the extracerebral blood flow. The system according to claim 18.

32. The one or more circuits are configured to select long-flight optical interference measurements in order to estimate the multiple features of the cerebral blood flow. The system according to claim 18.

33. Supervised learning is used to estimate the intracranial pressure. The system according to claim 18.

34. The random forest regressor is used for the supervised learning described above. The system according to claim 33.

Images