Patent Application
20090234245
This patent specification relates to the monitoring of a physiological condition of a patient using non-invasive measurement techniques. More particularly, this patent specification relates to the monitoring of intracranial pressure (ICP) using non-invasive optical techniques.
Intracranial pressure refers to the pressure exerted by the cranium on the tissue and fluid matter contained inside the cranium, which includes brain tissue, cerebrospinal fluid, and blood circulating in the brain. Typical values of ICP for a patient at rest are in the range of 10-15 mm Hg (0.013-0.020 atm). Elevated ICP levels are generally undesirable and are often a result of a traumatic head injury, an infectious disease such as meningitis, or another pathological condition such as brain tumor. For an adult, an elevated ICP above 40 mm Hg is likely to cause severe harm, and even pressures between 25 and 30 mm Hg are usually fatal if prolonged. Detection of ICP variations is recognized as an important tool in monitoring the state of injured patients, diagnosing symptoms of potentially diseased patients, and monitoring patient health during surgery or other therapeutic interventions.
Although various proposals have been made for non-invasive ICP monitoring, it is still generally recognized that reliable detection of ICP variations requires invasive measurement devices. However, such invasive techniques involve exposing and potentially traumatizing the brain tissue, which can increase the risk of infection, hemorrhage, leakage of cerebrospinal fluid, and other problems that can actually worsen the patient's condition.
Described in this patent specification are methods, systems, and related computer program products for non-invasive detection of ICP variations using optical techniques in the visible and/or near infrared regime. According to one preferred embodiment, optical radiation is propagated transcranially into the intracranial compartment, and optical radiation is detected that has migrated through at least a portion of the intracranial compartment and back out of the cranium. At least one signal representative of the detected optical radiation is processed to extract therefrom at least one component signal that varies in time according to at least one of an intrinsic physiological oscillation in the patient and an externally driven oscillation in the patient. For one preferred embodiment, the intrinsic physiological oscillation comprises at least one of an intrinsic respiratory oscillation and a cardiac oscillation. For one preferred embodiment, the externally driven oscillation comprises at least one of an external skull vibrator oscillation and a ventilated respiratory oscillation. The at least one extracted component signal is then processed to generate an output signal representative of the ICP variations in the intracranial compartment.
According to another preferred embodiment, a method for ICP monitoring is provided in which an absolute ICP of a patient is monitored using an invasive ICP monitoring device, such as a subarachnoid bolt. Simultaneously with the invasive ICP monitoring, a non-invasive ICP monitoring device is placed in optical communication with the head of the patient, the non-invasive ICP monitoring device using optical radiation to transcranially detect variations in the magnitudes of periodic intracranial matter oscillations intrinsically and/or extrinsically induced, the magnitude variations being indicative of intracranial matter compliance variations brought about by ICP changes. The absolute ICP from the invasive ICP monitoring device is used to calibrate the non-invasive ICP monitoring device. When the invasive ICP monitoring device is removed, ICP monitoring is continued by maintaining the non-invasive ICP monitoring device in optical communication with the head of the patient.
According to another preferred embodiment, a method for non-invasive ICP monitoring is provided, comprising applying a plurality of discrete mechanical impulses to the head of the patient at a respective plurality of discrete points in time. During each of a plurality of time intervals immediately subsequent to each respective discrete point in time, optical radiation is applied to the patient that propagates transcranially into the intracranial compartment, and optical radiation that has migrated transcranially outward from the intracranial compartment is detected. A plurality of time signals representative of the optical radiation detected during the respective time intervals is then processed to generate an output signal representative of the ICP variations. For one preferred embodiment, the processing comprises, for each of the time signals, computing at least one transient characteristic thereof induced by the mechanical impulse associated therewith. On an impulse over impulse basis, a decreasing value is assigned for the ICP output signal when the computed transient characteristic(s) change toward values indicative of greater intracranial matter compliance, while an increasing value is assigned for the ICP output signal when the computed transient characteristic(s) change toward values indicative of lesser intracranial matter compliance.
System 102 comprises an optical source 104 that emits radiation having a wavelength in the range of about 500 nm-1000 nm, i.e., in the upper visible and near infrared wavelengths. Light from the optical source 104 is carried by an optical fiber 106 to a source port 114 of an optical coupling device 112 on the forehead of the patient. Light that has migrated through at least a portion of the intracranial compartment and outward again through the cranium is collected at a detection port 116 of the optical coupling device 112 and guided to an optical detector 108 by an optical fiber 110. For one preferred embodiment, the optical coupling device 112 can be similar to one or more of the optical coupling devices disclosed in U.S. Pat. No. 5,596,987, which is incorporated by reference herein. Preferably, the optical coupling device 112 is designed to be a disposable, one-time-use patch that secures to the forehead using known adhesives. The optical coupling device 112 including the source port 114 and detection port 116 can alternatively be attached to an accessible skin surface elsewhere on the scalp other than the forehead.
The detector 108 generates a signal that is representative of the light collected at the detection port 116. For a relatively simple continuous wave embodiment in which the source 104 emits a monochromatic unmodulated carrier wave, the detector 108 provides a voltage signal VOUT representing an instantaneous intensity of the light collected at detection port 116. For one embodiment, the optical source 104 comprises a 4 mW laser diode emitting at 760 nm, and the optical detector 108 comprises a Hamamatsu R928 photomultiplier tube. Although the optical source 104, optical detector 108, and optical coupling device 112 are illustrated as distinct components in the example of
As used herein, intracranial compartment refers to the space inside the cranium, while intracranial matter refers broadly to the matter that occupies the intracranial compartment. The intracranial compartment encompasses, and the intracranial matter includes, the dura mater, subdural cavity, arachnoid layer, subarachnoid cavity, pia mater, and brain tissue, along with cerebrospinal fluid contained in the subdural cavity, the blood running throughout to all of the living tissue cells, and the arteries, capillaries, veins, etc., that carry the blood.
As used herein, intrinsic physiological oscillation refers to a physiological characteristic or behavior that is brought about autonomically by the patient's body, that exhibits some form of periodicity, and that directly or indirectly brings about some form of corresponding motion, even if slight, in the intracranial matter of the patient. The corresponding motion can be in the form of positional shifts ranging from very small, localized positional shifts to regional or widespread positional shifts, as well as positional shifts ranging from ordered or patterned positional shifts to disordered or random positional shifts. By way of non-limiting example, as the term positional shift is used herein, intracranial matter that is exhibiting a volume change (e.g., expansion or contraction), whether it be on a local basis or a widespread basis, is also exhibiting a positional shift, since at least some individual portion of that intracranial matter is necessarily moving relative to at least some other individual portion of that intracranial matter as part of that volume change. Likewise, by way of further non-limiting example, as the term positional shift is used herein, a wall of an intracranial artery that is undergoing expansion and contraction as part of a cardiovascular oscillation cycle is also exhibiting positional shifts, since at least some individual portion of that wall is necessarily moving relative to at least some other individual portion of that wall as part of those expansions and contractions.
One example of an intrinsic physiological oscillation is the patient's intrinsic respiratory oscillations, i.e., their natural breathing, which generally occurs at a periodic rate somewhere between 3 breaths per minute (0.05 Hz) and 30 breaths per minute (0.5 Hz). It has been observed that there is some degree of motion, in the form of slight positional shifts/volume changes, in at least a portion of the intracranial matter that occurs in conjunction with the respiratory oscillations of the patient. Another example of an intrinsic physiological oscillation is the patient's cardiac oscillations, which generally occur at a rate somewhere between 30 beats per minute (0.5 Hz) to 180 beats per minute (3 Hz). It has been observed that there is some degree of motion, in the form of slight positional shifts, that occur with the cardiac oscillations (heartbeat) of the patient.
As used herein, externally driven oscillation refers to a physiological characteristic or behavior that is brought about by an external force or input, that exhibits some form of periodicity, and that directly or indirectly brings about some form of corresponding motion, even if slight, in the intracranial matter of the patient. One example of an externally driven oscillation is a ventilated respiratory oscillation that occurs when the patient is placed on a ventilator. Just as with natural breathing, each ventilator-induced breath brings about some degree of positional shift/volume change in at least a portion of the intracranial matter relative to the cranium. However, unlike natural breathing, the operation of a ventilator is at a fixed periodic rate set by an attending clinician. Another example of an externally driven oscillation is an external skull vibrator oscillation brought about by a mechanical vibrator positioned in mechanical communication will the patient's skull. With advantages to be described further hereinbelow, there is provided in one preferred embodiment a non-invasive ICP monitoring system that includes at least one mechanical vibrator operating at a subsonic frequency in the range of about 3 Hz-30 Hz that is positioned so as to vibrate the patient's skull at that rate. Preferably, the intensity of the mechanical vibration is high enough to cause some degree of corresponding motion in at least a portion of the intracranial matter, but gentle enough not to cause too much discomfort to the patient. Toward this end, the duty cycle of the mechanical vibrator can be restricted to being “on” for only a few seconds, perhaps 4-5 seconds, every two or three minutes, and “off” otherwise.
As may be evident from the incorporated references, the particular physics and mathematics of the scattering and attenuation of the light as it propagates in a banana-shaped migration path from the source port 114 to the detection port 116 can be quite complicated, even when various simplifying assumptions are made regarding the various bone, tissue, and fluid types traversed. However, in accordance with a preferred embodiment, ICP variations are detected in a relatively elegant manner that transcends the particular scheme (CWS, PMS, TRS, etc.) by which the interrogating light waves are modulated, introduced, detected, and evaluated. As described above, it has been observed that at least a portion of the intracranial matter will experience some type of periodic motion relative to the cranium, in the form of positional shift and/or volume change, in correspondence with the above-described intrinsic physiological oscillations. Alternatively or in conjunction therewith, at least some periodic motion of the intracranial matter can be induced in correspondence with externally driven oscillations. Furthermore, it has been found that the amount of this periodic motion will become more restricted at higher ICP pressures and less restricted at lower ICP pressures. If a signal is extracted from the detected radiation that varies in magnitude (or other measurable amount) with an intrinsic physiological oscillation or an externally driven oscillation, then that extracted signal can be used to detect ICP variations regardless of the designed physiological significance (if any) of that extracted signal. Generally speaking, larger variations in that extracted signal will be indicative of a lower ICP, because the intracranial matter is less restricted in its periodic motion when the ICP is lower. Likewise, smaller variations in the extracted signal will be indicative of a higher ICP, because the intracranial matter is more restricted in its periodic motion when the ICP is higher.
For one preferred embodiment, a single signal is extracted from the detected radiation that varies in magnitude (or other measurable amount) with a single intrinsic physiological oscillation or a single externally driven oscillation. For another preferred embodiment, multiple signals are extracted from the detected radiation that vary in magnitude (or other measurable amount) with multiple respective intrinsic physiological oscillations, multiple respective externally driven oscillations, or a combination of at least one respective intrinsic physiological oscillation and at least one respective externally driven oscillation.
Also provided in accordance with this preferred embodiment is a second processor 119 (which can alternatively be analog filter circuit, and which can optionally be integral with the first processor 118) that processes the component signal Cresp in digital form to provide an output signal Prel indicative of the ICP variations in the intracranial compartment. As part of the processing by the second processor 119, an envelope magnitude (i.e., the vertical distance between the plot lines 130a and 130b) of the component signal Cresp is determined. The output signal Prel is assigned a greater value when the envelope magnitude has a lesser value, and the output signal Prel is assigned a lesser value when the envelope magnitude has a greater value. System 102 further comprises a user display 120 providing a graphical representation 122 and/or a numerical representation 124 of the ICP output value Prel as a percentage of a baseline value 126.
It is to be appreciated that envelope magnitude (i.e., the vertical distance between upper and lower envelope lines) represents one of a variety of different amplitude characteristics of the component signal Cresp that can be measured and used in the determination of the output signal Prel without departing from the scope of the present teachings. More generally, any amplitude characteristic of the component signal Cresp (i.e., any metric that characterizes an AC strength of the component signal Cresp) may be used in place of the envelope magnitude, such as an RMS value, a time average of a rectified version, a standard deviation, a square (or cube, etc) of the peak-to-peak value, and so on, without departing from the scope of the preferred embodiments. Thus, descriptions provided herein with respect to envelope magnitude of the component signal Cresp, which are provided for purposes of clarity of presentation, are applicable for other amplitude characteristics of the component signal Cresp as well.
The particular nature of the inverse relationship between the envelope magnitude of the component signal Cresp and the output value Prel (e.g., whether it is a reciprocal relationship, a negative arithmetic relationship, or other inverse relationship) could be determined empirically based on test scenarios by a person skilled in the art without undue experimentation in view of the present disclosure. By way of example, a set of test data can be developed in clinical data-gathering trials by applying the system 102 to a population of patients during periods in which their absolute ICP levels are being monitored by an invasive ICP monitoring device, such as a subarachnoid bolt, which is currently recognized as the “gold standard” in ICP measurement. The outcome of the clinical data-gathering trials can be used to establish a relationship between (i) the percentage of envelope magnitude change from an initial envelope magnitude baseline, and (ii) the percentage of ICP variation from the corresponding initial absolute ICP reading. This can then be used to provide the ICP output value Prel as a percentage of the baseline value 126. Depending on the results of the clinical data-gathering trials, it may be possible to establish a set of normative data based on different patient characteristics (e.g., height, weight, body surface area to weight ratio, etc.) to provide a more precise mapping between percent envelope magnitude change and percent ICP change. Indeed, it may even be possible, and would certainly be within the scope of the preferred embodiments, to establish a set of normative data that allows absolute ICP levels to be computed based on certain patient information as combined with the envelope magnitude changes and/or envelope magnitude levels, in which case the Prel output shown in
During the peak (right side) of a respiratory cycle, the intracranial matter is deformed toward the cranial bone 202 by a slightly greater amount than during the valley (the drawings are exaggerated for clarity). Thus, for example, there is a greater distance y1 between sections 211 and 213 during the valley (left side) and a lesser distance y2 during the peak (right side). It is these slight shifts of the intracranial matter that cause the variations of the detected optical signal as extracted at the respiration frequency range. Notably, although it is believed that much of the intracranial matter shifting is due to subdural cavity deformation between the dura mater and arachnoid layers, the true physiological nature of the deformation (e.g., which tissues are deforming by what amount, is the deformation conformal versus irregular, etc.) is generally irrelevant for the purposes of measuring the ICP variations in accordance with the preferred embodiments. Rather, the main requirement is simply that “something” is deforming, in “some” manner that affects the detected optical signal in “some” measurable way according to the respiration cycle of the patient.
As used herein, compliance refers to the property of intracranial matter that is illustrated in the examples of
Generally speaking, all of the intracranial matter compliance metrics (e.g., envelope magnitudes) will share a common characteristic in that each will generally increase as the ICP decreases, and that each will generally decrease as the ICP increases. However, it has been found that a rich variety of clinically interesting and relevant information can arise from the fact that these different intracranial matter compliance metrics (e.g., envelope magnitudes) will generally exhibit different differential characteristics with changing ICP as a function of the prevailing absolute level of ICP. By way of example, letting the variable ER represent the respiratory intracranial matter compliance metric (e.g., envelope magnitude of the extracted respiratory component of the detected optical signal), and letting the variable EC represent the cardiac intracranial matter compliance metric (e.g., envelope magnitude of the extracted cardiac component of the detected optical signal), it has been found that ER tends to diminish rapidly with increasing ICP when the absolute ICP is at moderate levels. However, as the absolute ICP increases further, ER tends asymptotically toward zero, such that at high levels of absolute ICP, ER metric ceases to change in any measurable way with increased ICP. In contrast, the cardiac envelope EC tends to be quite robust against increases in absolute ICP, and maintains appreciable nonzero values even for high levels of absolute ICP. In accordance with a preferred embodiment, both of the metrics ER and EC are computed from the detected signal information, and their values relative to each other are analyzed (such as by taking their ratio, difference, etc.) to yield increased precision in the ICP determination process and/or to derive other useful information regarding the health of the patient. The specific ways in which ER and EC can be advantageously processed can be determined, for example, by using data from a large clinical data-gathering trial, where ER and EC are tracked along with absolute ICP and other vital signs, and patterns and/or statistical correlations in the data can be developed. Indeed, it would not be outside the scope of the preferred embodiments for a set of normative data to be developed using multivariate correlations among ER, EC, EV (e.g., envelope magnitude of the extracted subsonic vibratory component of the detected optical signal), and other intracranial matter compliance metrics such that the non-invasive ICP monitoring device can be automatically calibrated based on these computed values for providing absolute ICP level determinations.
Thus, provided in the system 401 according to a preferred embodiment is an optical coupling patch 402 and source/detector system 404 for providing a voltage signal VOUT representing an instantaneous intensity of light collected at a detection port of the optical coupling patch 402, in a manner similar to like elements of
Notably, many different combinations of the above-described elements 408, 496, 497, 498, and 499 can be hooked up to the console 406 without departing from the scope of the preferred embodiments, including an option in which none of them are hooked up and only the signal VOUT is provided to the console. Generally speaking, as more normative clinical data is gathered, the selected ones of these hookups providing the most useful signals will be identified, and increasingly precise results, even up to and including calibrated absolute ICP measurements, can be obtained. However, even in a simplest embodiment in which no external hookups are provided except for VOUT, the system 401 is still useful as an indicator as to whether the ICP is increasing, decreasing, or staying the same. Preferably, the processor 414 is configured to be easily upgradable, such as by firmware flash or internet download, so that the latest and best capabilities are integrated as more and more normative clinical data is gathered.
The user input device 412 allows a user, such as a clinician, to select the basis upon which non-invasive ICP measurement is to be made. Depending upon which buttons the user selects, the processor 414 will “listen” to the appropriate external signals, extract the relevant components from VOUT, and provide a best estimate Prel (or, potentially, Pabsolute) for display to the clinician.
As used herein, lock-in detector refers to a device or algorithm that receives an input signal and a periodic reference signal, and synchronously extracts frequency components from the input signal that correspond to the frequency content of the periodic reference signal. Generally speaking, if a periodic reference signal is available, lock-in detection is highly superior to passive bandpass filtering with respect to signal-to-noise performance, and so the processor 414 generates the signal CR using the bandpass filter 452 as a “last resort” when the user has chosen neither the ventilator nor the respiratory monitor. However, the scope of the preferred embodiments is not so limited, and in other preferred embodiments, plural versions of the CR signal can be generated using both the lock-in detector 454 and bandpass filter 452, and both versions can be considered as distinct inputs to the evaluation module after envelope detection. It still another preferred embodiment, three versions of the CR signal can be created, including one version from the bandpass filter 452, a second version from the lock-in detector 454 using the ventilator reference signal, and a third version from the lock-in detector 454 using the respiratory monitor reference signal.
The signal CR, which is analogous to the periodic component signal Cresp of
Processor 414 further comprises a bandpass filter 456 that is designed to extract a cardiac component CC from the signal VOUT. The bandpass filter 456 is selected at switch SW3 if the user has not chosen the EKG signal on the input device 412. However, if the user has indeed selected the EKG signal, then a lock-in detector 458 is selected at switch SW3, with a reference signal being from the EKG output. The signal CC is then fed to an envelope detector 470 for extracting the envelope signal EC, which represents a measure of the intracranial matter compliance with respect to the cardiac oscillations of the patient. In another preferred embodiment (not shown), there is an option to turn off the cardiac channel entirely, in which case neither bandpass filter 456 nor the lock-in detector 458 is active and no cardiac component is input to the evaluation module 474.
Processor 414 further comprises a lock-in detector 460 that is designed to extract an externally driven vibratory component CV from the signal VOUT. There is generally no need for a passive bandpass filter here because a reference signal should always be available, although the scope of the preferred embodiments is not so limited. The signal CV is then fed to an envelope detector 466 for extracting the envelope signal EV, which represents a measure of the intracranial matter compliance with respect to the externally driven subsonic vibratory oscillations of the patient. The switch SW5 is opened to turn off the subsonic vibratory oscillation channel if the user has not selected it on the input device 412.
Processor 414 further comprises a lock-in detector 462 that is designed to extract an “other” oscillatory component CO from the signal VOUT. Generally speaking, there may be a variety of other periodic inputs that could lead to corresponding intracranial matter oscillations, including those that are not yet currently known. By way of somewhat fanciful example, large periodic doses of therapeutic radiation might someday be applied that cause corresponding intracranial matter oscillations. The extraction of such “other” oscillatory components from the signal VOUT and processing them to detect a metric of corresponding intracranial compliance is not outside the scope of the preferred embodiments. As illustrated in
Finally, evaluation module 474 receives those of ER, EV, EO, and EC that are available according to the user's input and computes therefrom the output Prel (or, potentially, Pabsolute) for display on the display output 410. Similar to the discussion supra with respect to
It is to be appreciated that the scope of the preferred embodiments is not limited to the continuous wave scenario of FIGS. 1 and 4A-4B. In another preferred embodiment (not shown), the emitting and detecting performed by the source(s) and detector(s) can be in accordance with phase modulation spectroscopy (PMS) or time resolved spectroscopy (TRS) principles, provided only that a one-dimensional signal (e.g., a time-varying voltage) representative of the detected output radiation (e.g. phase shift, time of flight, etc.) is provided to the first processor 118 (
In yet another preferred embodiment, (not shown), plural arrays of sources and detectors can be positioned and operated according to CWS, PMS, TRS, or other principles such that a two-dimensional map or image of a spatially varying property within the intracranial compartment is generated, the two dimensional image being time-varying and morphing, even if slightly so, according to the intrinsic physiological oscillation(s) and/or an externally driven oscillation(s) in the patient. Image processing can then be performed on the time-varying image to generate a metric related to an amount of morphing that is happening in correspondence with those oscillations. In one simple example, the amount of morphing can be identified as the time-varying distance between two landmark locations in the two-dimensional image. This metric can then be treated like the voltage VOUT in
Referring again to
During each of a plurality of time intervals (INT0, INT1, INT2) immediately subsequent to each respective discrete point in time (t0, t1, t2) optical radiation is applied to the patient that propagates transcranially into the intracranial compartment, and optical radiation that has migrated transcranially outward from the intracranial compartment is detected. A plurality of time signals (WTRANS,0(t), WTRANS,1(t), WTRANS,1(t)) representative of the optical radiation detected during the respective time intervals (INT0, INT1, INT2) is then processed to generate an output signal representative of the ICP variations.
For one preferred embodiment, the processing comprises, for each of the time signals (WTRANS,0(t), WTRANS,1(t), WTRANS,1(t)), computing at least one transient characteristic thereof induced by the mechanical impulse (700, 701, 702, respectively) associated therewith. Preferably, on an impulse over impulse basis, a decreasing value is assigned for the ICP output signal when the computed transient characteristic(s) change toward values indicative of greater intracranial matter compliance, while an increasing value is assigned for the ICP output signal when the computed transient characteristic(s) change toward values indicative of lesser intracranial matter compliance. For a particular time signal WTRANS,j(t), examples of transient characteristics can be the peak difference between WTRANS,j(t) and the steady state value WSS (i.e., the value or characteristic when there has been no thumping for a substantial time), the time-to-peak or rise time after the impulse, the overall time center of mass of the curve WTRANS,j(t), the relaxation time between the peak value at the steady-state value, or any of a variety of other transient characteristics that characterize how much and/or how fast the intracranial matter is shaking, shifting, etc. responsive to the mechanical thumping. Generally speaking, the best type of optical modulation/filtering scheme used to derive WTRANS,j(t), the type and degree of thumping, the particular selection and/or combinations to transient characteristics to compute, the particular manner in which those values are calibrated to relative or absolute ICP metrics, and other relevant factors could be determined by a person skilled in the art (e.g., empirically using structured clinical experiments) in view of the present disclosure without undue experimentation.
Whereas many alterations and modifications of the preferred embodiments will no doubt become apparent to a person of ordinary skill in the art after having read the foregoing description, it is to be understood that the particular embodiments shown and described by way of illustration are in no way intended to be considered limiting. Thus, reference to the details of the described embodiments are not intended to limit their scope.
