Method, device and device arrangement for classifying a cerebral pressure, and corresponding computer program and computer-readable medium

Abstract

The invention relates to a computer-implemented method for classifying a cerebral pressure, the method comprising the following steps: receiving data with time-of-flight values based on time-of-flight values measured through a skull cross section for one or more ultrasonic signals at successive times; classifying a property of the cerebral pressure on the basis of one or more features of a representative curve for a time profile of the time-of-flight values in order to obtain a classified property of the cerebral pressure; and outputting the classified property of the cerebral pressure.

Description

[0001] Method, device and device arrangement for classifying intracranial pressure, as well as a corresponding computer program and a computer-readable medium

[0002] The invention relates to a method, a computer-implemented method for classifying intracranial pressure, a device for classifying intracranial pressure, a device arrangement for classifying intracranial pressure, and a corresponding computer program and a computer-readable medium.

[0003] The brain is constantly supplied with oxygen and nutrients via the blood. If the blood flow to the brain is reduced, it can cause brain damage, particularly the death of nerve cells. Intracranial pressure, i.e. the pressure inside the skull, including the cerebrospinal fluid spaces (also called intracranial pressure, ICP), is important in this context because it influences blood flow to the brain and thus the supply of oxygen and nutrients. Under “normal” conditions, intracranial pressure is permanently below around 10 mmHg, and therefore does not restrict blood flow to the brain. This is because the brain tissue has the natural elasticity to absorb blood. If the brain is damaged by injury or bleeding, it usually swells.However, because the brain is surrounded by a hard bone shell that protects it from damage, its rigidity means it can only give way slightly when the brain swells. Therefore, intracranial pressure can rise when the brain swells, for example, it can remain permanently above 15 mmHg. When intracranial pressure rises, the elasticity of the brain tissue decreases, meaning the tissue in the brain can expand less during the systolic and diastolic phases, and less oxygen-rich blood reaches the brain. This reduces blood flow to the brain. In extreme cases, if intracranial pressure exceeds a certain threshold, i.e. is abnormal, the lack of elasticity in the tissue means the brain only receives poor blood flow, and nerve cells in the brain can die.

[0004] It is therefore important to be able to determine whether the intracranial pressure is in a normal or abnormal range, ie whether it has sufficient elasticity to absorb oxygen-rich blood in order to react accordingly.

[0005] According to the currently prevailing standard, intracranial pressure is determined invasively using pressure probes in the cranial cavity (the so-called gold standard for intracranial pressure measurement). The pressure probes are surgically inserted, usually via a drainage catheter, through the outer bone shell of the skull into the cranial cavity, preferably into a lateral ventricle of the brain. By inserting the pressure probes directly into the brain, the currently prevailing pressure within the skull can be directly measured by the pressure probes. From the change in the measured pressure within the skull during one or more diastolic and systolic phases, intracranial pressure and thus the elasticity of the brain, i.e. the brain's ability to absorb oxygen-rich blood, can be determined.In addition, a cerebrospinal fluid drain, which is available as an effective intracranial pressure-lowering therapy, can be placed in the skull at the same time as the intracranial pressure measurement. Such invasive methods impose considerable stress on the patient during surgery, can lead to complications, and require appropriate aftercare. Furthermore, such invasive methods place high demands on equipment and personnel in medical facilities.

[0006] Previous approaches to improving the gold standard for intracranial pressure measurement have primarily focused on improving the sensitivity of pressure measurement probes, reducing their dimensions, and improving their contact points. In particular, by reducing the size of the pressure measurement probes and additionally inserting a CSF drain into the skull during surgery, the surgical effort required to detect and treat elevated intracranial pressure could be reduced. However, this only slightly reduced the severity of the procedure for the patient.

[0007] Furthermore, non-invasive methods are known from WO 2020 / 219773 Ai in which a transmitter and a receiver are mounted on opposite sides of the skull. The distance and propagation time of a sound signal between the transmitter and receiver are determined. The intracranial pressure is estimated from the correlation between distance and propagation time. However, such methods cannot reliably determine intracranial pressure.

[0008] Against this background, the object of the invention is to provide a particularly simple method with which increased intracranial pressure can be determined.

[0009] This object is achieved by a computer-implemented method for classifying intracranial pressure according to the main claim, the method comprising: receiving data with transit time values ​​based on transit time values ​​measured at successive points in time for one or more ultrasound signals through a cross-section of a skull; classifying a property of the intracranial pressure based on one or more features of a representative curve of a temporal progression of the transit time values ​​to obtain a classified property of the intracranial pressure; and outputting the classified property of the intracranial pressure. All method steps of the computer-implemented method described herein are carried out by one or more computers.

[0010] The transit time values ​​of one or more ultrasound signals can be measured using ultrasound probes prior to receiving the data, so that the patient's presence is not required for the procedural steps performed by the computer. The data can be transferred to the computer, for example, via a data storage device such as a USB stick, or via a wireless or wired connection. The data is processed separately, without physical contact with the patient.

[0011] The invention is based on the inventors' finding that the method according to the invention enables a non-invasive assessment of intracranial pressure. Using the present method, intracranial pressure, in particular the determination of whether it is normal or abnormal, can be determined non-invasively. This means, in particular, that no surgical intervention is necessary to determine whether intracranial pressure is normal or abnormal, thus reducing the burden on the patient and significantly reducing the medical personnel and logistical effort, and ultimately the cost factor.

[0012] In this context, the invention is based in particular on the inventors' finding that a change in the state of the intracranial pressure within the skull directly correlates with a change in the transit time of the one or more ultrasound signals through a cross-section of the skull (as will also be explained in more detail in connection with Fig. 2).

[0013] The heartbeat induces a pulse wave, which pushes blood into the vessels of the brain. The pulse wave induced by the heartbeat leads to a change in volume or expansion of the vessels in the brain, which absorb the blood pushed into the brain. This volume change leads to a change in pressure in the skull, which was previously measured directly via the sensors in the brain, according to the gold standard.

[0014] The inventors have now recognized that a transit time value of one or more ultrasound signals measured through the cross-section of the skull depends on a superposition of all tissue layers and fluids located within the sound field and is thus correlated to them.

[0015] The transit time values ​​of one or more ultrasound signals through a cross-section of a skull measured at successive points in time thus capture the transit time differences resulting from the pressure-induced volume change and pressure-induced compression of the components of the brain, such as blood, cerebrospinal fluid and tissue.

[0016] The curve of the time course of the transit time values ​​measured at successive points in time thus represents the volume and density changes of the components of the brain and thus of the intracranial pressure during one or more

[0017] Pulse waves. Based on the characteristics of the time-of-flight curve, which reflects the pressure-induced volume change and pressure-induced compression of the brain components during one or more heartbeats, intracranial pressure can be determined or classified. In particular, a property, such as whether it is normal or abnormal, can be classified noninvasively.

[0018] In particular, the method according to the invention makes it possible to determine the condition within the skull, specifically the intracranial pressure, without distortion caused by the heartbeat. Measuring the distance between the probes is therefore unnecessary.

[0019] In a learning phase, the underlying algorithm for classifying a property of intracranial pressure was trained in such a way that individual features of the representative curves can be assigned to a property of intracranial pressure.

[0020] During the learning phase, transit time values ​​through the skull were measured over time. In parallel, direct intracranial pressure measurements were determined according to the proven gold standard in several patients with normal and abnormal intracranial pressure. The following physiological inclusion criteria were considered when selecting patients: Pathology: severe traumatic brain injury (Glascow Coma Scale < 10); heart rate range: 50 min-1 < HR < 120 min-1; pulse pressure (difference between systolic and diastolic blood pressure) < 90 mmHg; systolic blood pressure < 190 mmHg; diastolic blood pressure > 30 mmHg.

[0021] Using the features of several representative curves from temporal courses of the transit time values ​​compared to the direct intracranial pressure measurements, which were recorded in parallel with the transit time value measurements on the same patients, it was possible to determine whether a feature in the curves is characteristic of a specific intracranial pressure or whether a feature reliably separates the data from abnormal and normal intracranial pressures. Within the training dataset, features were identified that showed the greatest importance for distinguishing between elevated ICP (e.g., >15 mmHg) and normal ICP (e.g., <10 mmHg).

[0022] During the learning phase, various features were considered on several curves. The considered features can be divided into four groups: statistical features (e.g., means, standard deviation, distribution skewness); features from data aggregation (e.g., minimum runtime values, maximum runtime values, average amplitudes); features from curve discussions (e.g., peak widths, splines, extreme values, inflection points); and frequency-domain features from public expert libraries (such as tsfresh) for time-dependent signals.

[0023] During the learning phase, it was determined which of the examined features in the representative curves from the respective time courses are characteristic of a certain intracranial pressure or a normal or abnormal intracranial pressure, which was determined with the gold standard.

[0024] By determining these characteristics of the curve of the time course of the transit time values ​​measured through the skull cross section, characteristics of a property of the intracranial pressure can be assigned, whereby the intracranial pressure can be classified in a non-invasive manner from the curve of a time course of the transit time values ​​measured through the skull cross section of one or more ultrasound signals.

[0025] In the temporal progression, at least one runtime value is assigned to each of the successive points in time. The temporal progression can be described differently as a curve in a two-dimensional coordinate system with the runtime values ​​on one axis and the respective successive points in time on the complementary axis.

[0026] In one aspect, the temporal course of the transit time values ​​correlates with a temporal course of the intracranial pressure.

[0027] Because the inventors recognized that the temporal progression of the transit time values ​​measured at successive points in time correlates with the volume and density changes of the brain components and thus directly reflects the temporal progression of the pressure in the skull, the characteristic of the intracranial pressure can be classified non-invasively based on the curve of the temporal progression of the transit time values. In particular, it can be deduced whether the intracranial pressure is normal or abnormal. In one aspect, the temporal progression of the transit time values ​​exhibits several peaks, each of which correlates with an increase and a decrease in intracranial pressure during a systolic and a diastolic phase.

[0028] A peak is formed by an increase or lengthening of the transit time values ​​of one or more ultrasound signals through the cross-section of the skull, measured at successive points in time. This means that a peak correlates to a rise in intracranial pressure during a heartbeat. An increase in intracranial pressure during a systolic phase, i.e. a rise in pressure in the skull, results in a longer transit time value. A fall in intracranial pressure during a diastolic phase, i.e. a drop in pressure in the skull, results in a shorter transit time value. A rising area of ​​a peak therefore corresponds to a systolic phase in which blood is pushed into the vessels of the brain, and a falling area of ​​the peak corresponds to a diastolic phase in which blood flows out of the brain.

[0029] The temporal progression of the transit time values ​​thus represents several systolic and diastolic phases, i.e., several consecutive pulse waves in which blood is forced into the vessels of the brain. Thus, several pulse waves are resolved by the temporal progression of the transit time values. This allows for reliable classification of intracranial pressure based on the temporal progression curve.

[0030] In one aspect, a change between transit time values ​​in the time course at at least two consecutive time points directly correlates to a change in intracranial pressure at the at least two consecutive time points.

[0031] Because the inventors have recognized that a change between transit time values ​​of the one or more ultrasound signals directly correlates to a change in the intracranial pressure at the at least two consecutive points in time, the intracranial pressure can be classified non-invasively by the method according to the invention.

[0032] In one aspect, the classified property of the intracranial pressure indicates that normal intracranial pressure, in particular an intracranial pressure of Si or lower, is present, or that abnormal intracranial pressure, in particular an intracranial pressure of S2 or more with S2>S1, is present. The method according to the invention can differentiate between normal and abnormal intracranial pressures based on different threshold values ​​Si, S2 and is therefore not limited to specific values ​​for Si and S2. The threshold values ​​Si, S2 are determined based on the patient population, where Si <S2 gilt. Diese Schwellwerte Si, S2 liegen jeweils üblicherweise zwischen 5 mmHg und 20 mmHg, wobei die Schwellwerte Si=io mmHg und 82=15 mmHg zur Vermeidung der Klassifikation von tatsächlich anormalen Hirndrücken als normal besonders bevorzugt sind.

[0033] Intracranial pressure can thus be classified as normal or abnormal. The corresponding finding of normal or abnormal intracranial pressure can then be output by the process; for example, the classification can be shown on a display or transmitted to another entity in the form of data. Based on the finding, medical personnel can diagnose a patient's disease (without the patient being present).

[0034] At an intracranial pressure that is preferably lower than approximately 10 mmHg, the brain has sufficient elasticity for oxygen uptake. At an intracranial pressure that is preferably 15 mmHg or higher, it is assumed that the intracranial pressure is in the elevated range. In cases of doubt (e.g., due to measurement inaccuracies), if the intracranial pressure is between 10 and 15 mmHg, elevated intracranial pressure is assumed.

[0035] In one aspect, the representative curve has at least one characteristic peak.

[0036] The characteristic peak is thus characteristic of the transit time changes of several pulse waves in the brain. For example, the characteristic peak can be formed by calculating a median of the transit time values ​​of several peaks in the temporal course of the transit time values. In other words, the characteristic peak in the representative curve represents a characteristic increase and a characteristic decrease in the intracranial pressure of several systolic and several diastolic phases. As a result, the at least one characteristic peak better represents a characteristic behavior of the brain. Thus, the intracranial pressure can be classified more reliably based on the temporal course of the transit time values. In one aspect, the property of the intracranial pressure is determined by the computer extracting the one or more features of the curve of the at least one characteristic peak.

[0037] The characteristic peak reflects the transit time changes across multiple pulse waves in the brain, thus providing a more meaningful representation of brain behavior. Extracting one or more features from the curve of at least one characteristic peak ensures more reliable classification.

[0038] In one aspect, a first feature of the one or more features is a peak width of the at least one characteristic peak at a certain percentage of the maximum amplitude of the at least one characteristic peak.

[0039] During the learning phase, the inventors discovered that one characteristic that can be assigned to normal or abnormal intracranial pressure is the peak width in the temporal progression of the runtime values. Thus, normal or abnormal intracranial pressure can be determined based on the peak width of at least one characteristic peak at a certain percentage of the maximum amplitude.

[0040] In one aspect, a second characteristic feature of the plurality of features is a similarity of the curve shape of the at least one characteristic peak to a synthetic function, in particular a continuous wavelet transform.

[0041] In addition, during the learning phase, the inventors discovered that another feature that reflects normal or abnormal intracranial pressure is a similarity of the curve shape of the at least one characteristic peak to a synthetic function.

[0042] During the learning phase, several characteristic peaks were convolved or superimposed with several defined synthetic functions, with various synthetic functions being tested. The similarity of the curve shape can be determined by mathematically convolving the curve shape of a synthetic function. Various wavelet functions were used as synthetic functions, with properties of the wavelet functions being varied using scaling factors, such as their coefficients, and the corresponding functions were thus compressed and expanded. Examples of wavelet functions used are: the db4, dbiö, haar, coif, sym4, sym8, biori.3, and the bior.i functions. Descriptions of wavelet functions can be found, for example, at: https: / / de.mathworks.com / help / wavelet / ref / cwt.html or at https: / / en.wikipedia.org / wiki / Ricker_wavelet.

[0043] By comparing them with corresponding gold standard intracranial pressure measurements, the inventors found that certain scaling coefficients of the wavelet functions reflect normal or near-normal intracranial pressure.

[0044] The second characteristic feature, which is extracted from the representative peak, allows the property of intracranial pressure to be reliably classified. A particularly reliable classification can be achieved by determining a combination of the first and second features from the at least one characteristic peak.

[0045] In one aspect, the at least one characteristic peak is for a plurality of peaks in at least one temporal segment of the temporal course of the runtime values.

[0046] According to this aspect, the at least one characteristic peak is formed based on a temporal segment, i.e., a temporal section, of the temporal progression of the runtime values. In other words, a segment can be understood as a temporal segment that includes the runtime values ​​assigned to a subset of the consecutive time points.

[0047] For example, a segment with the best quality runtime data can be selected. In other words, segments with poor quality runtime data can be excluded. This makes the classification of the intracranial pressure property more reliable. Furthermore, the associated reduction in the amount of data used for evaluation can reduce the required computing power.

[0048] In one aspect, the at least one characteristic peak is determined by the following steps: extracting a plurality of individual peaks from the at least one temporal segment of the temporal course of the runtime values; normalizing times of the runtime values ​​in the individual peaks; grouping the individual peaks by applying a similarity criterion for the individual peaks; identifying the group with the largest number of peaks, and forming the at least one characteristic peak from the peaks in the identified group of peaks.

[0049] The at least one characteristic peak is thus formed from the travel time values ​​of the peaks from a group of most peaks that satisfy the similarity criterion.

[0050] The similarity criterion may, for example, include forming a silhouette coefficient for the individual peaks.

[0051] This ensures that at least one characteristic peak represents a plurality of peaks with relatively meaningful data quality. This makes the classification of the intracranial pressure characteristic based on the measured transit time values ​​more reliable.

[0052] In one aspect, forming the at least one characteristic peak comprises: for each time point, forming a median of the travel time values ​​of the peaks from the identified group.

[0053] This results in a plurality of peaks being formed by dividing the transit time values ​​of at least one characteristic peak by the median of the transit time values. This results in the at least one characteristic peak representing a plurality of peaks. This makes the classification of the intracranial pressure characteristic based on the measured transit time values ​​more reliable.

[0054] In one aspect, the at least one segment is determined by the following steps: extracting one or more temporal segments from the temporal progression of the runtime values, wherein a segment has multiple peaks in the runtime values; determining the at least one temporal segment by selecting one or more temporal segments in which the peaks satisfy a quality criterion.

[0055] According to this aspect, segments are selected in which the travel time curve meets a certain quality. This allows temporal segments in the measured travel time values ​​that contain disturbances caused, for example, by heartbeat pauses or similar events to be excluded and not used for further data processing. By selecting individual segments from the temporal course of the travel time values ​​for classification according to a quality criterion, the classification of the intracranial pressure property becomes more reliable. Furthermore, the computational effort required for data evaluation can be reduced.

[0056] In one aspect, the quality criterion is formed by an autocorrelation function, which evaluates the intrinsic similarity of the temporal course of the runtime values ​​in a segment.

[0057] The autocorrelation function is a measure of the intrinsic similarity of the curve. Since the temporal course of the runtime values ​​is characterized by rhythmic, recurring fluctuations with a high degree of similarity, the autocorrelation function represents a criterion for the presence of a brain pulse curve.

[0058] According to this aspect, segments in which the peaks in the curve of the temporal progression of the transit time values ​​exhibit disturbances, which, for example, correspond to a non-existent heartbeat, can be excluded. This improves the reliability of the classification of the intracranial pressure characteristic. In one aspect, the quality criterion is formed by comparing the maximum amplitudes of the most dominant frequency component with the amplitudes of other frequency components in the temporal progression of the transit time values ​​in a segment.

[0059] In this aspect, the quality criterion can be expressed differently as the peak-to-mean criterion, which describes the ratio between the amplitude of the pulse and its harmonics to the remaining frequency components over time. Thus, it is an operation in the frequency domain. Peak-to-mean is a measure of whether the brain pulse curve represents the dominant component over time and is associated with a correspondingly high amplitude. If the signal-to-noise ratio is too low, the respective segment is not used for further analysis.

[0060] According to this aspect, segments in which the peaks in the time-of-flight curve exhibit disturbances that make the peaks caused by the heartbeat appear to be the less dominant component can be excluded. This improves the reliability of the classification of the intracranial pressure characteristic.

[0061] In one aspect, the quality criterion is formed by an evaluation of frequencies in the temporal course of the runtime values ​​in a segment.

[0062] The quality criterion for this aspect can be described as the upper band noise criterion, which describes the proportion of higher frequencies (>15 Hz) that are not generated by harmonics of the heartbeat. A high proportion indicates increased noise, and the respective segment is not used for further analysis. Noise amplitudes in this frequency range are not attributable to physiological origins.

[0063] This aspect allows segments in which the peaks in the time-of-flight curve exhibit disturbances, such as those caused by non-physiological noise, to be excluded. Furthermore, background frequencies not caused by the heartbeat can be excluded.

[0064] The quality criteria for these aspects can be applied individually or in combination to select a segment. By combining several quality criteria, segments with particularly meaningful data quality can be selected.

[0065] For example, a segment can be selected which satisfies the quality criterion formed by the autocorrelation function and additionally satisfies the quality criterion formed by comparing the maximum amplitudes of the most dominant frequency component or satisfies the quality criterion formed by evaluating frequencies in the time course.

[0066] In other aspects, for example, a segment can be selected that satisfies the quality criterion formed by the evaluation of frequencies over time and additionally satisfies the quality criterion formed by the comparison of the maximum amplitudes of the most dominant frequency component. In particularly preferred aspects, a segment can be selected that satisfies the quality criterion formed by the autocorrelation function and satisfies the quality criterion formed by the comparison of the maximum amplitudes of the most dominant frequency component, as well as the quality criterion formed by the evaluation of frequencies over time. This aspect is particularly preferred because it allows a segment to be selected that satisfies the quality criteria of all three aspects and thus provides particularly meaningful data quality for the one or more selected segments.

[0067] In one aspect, the method further comprises: measuring the travel time values ​​of the one or more ultrasound signals through the cross-section of the skull using at least two probes positioned on opposite sides of the skull at the successive times.

[0068] According to this aspect, the transit time of one or more ultrasound signals can be measured. A transit time value corresponds to the transit time required for the one or more ultrasound signals to traverse the skull between the probes along the cross-section.

[0069] To measure the time of travel, the at least two probes are positioned on opposite sides of the skull. The at least two probes can comprise at least one emitter which is designed to output the one or more ultrasound signals, and at least one detector which is designed to receive the one or more ultrasound signals emitted by the at least one emitter. The at least one emitter and the at least one detector can be positioned on different opposite sides of the skull such that a sound field between the probes is approximately parallel to a frontal cross-section of the skull. By means of the at least one emitter and the at least one detector, the time of travel which the one or more ultrasound signals require to travel from the at least one emitter to the at least one detector through the cross-section of the skull can be measured.

[0070] As described above, by measuring the propagation time of one or more ultrasound signals, a characteristic of intracranial pressure can be non-invasively determined and classified, since the propagation time of the one or more ultrasound signals reflects a state of the brain along the cross-section of the skull. Furthermore, the one or more ultrasound signals can be a longitudinal wave. The inventors have discovered that this type of wave is particularly well suited for determining and classifying intracranial pressure.

[0071] The probes have at least one emitter for outputting the one or more ultrasonic signals and at least one corresponding detector for detecting the output one or more ultrasonic signals, wherein the at least one emitter and the at least one detector are positioned on opposite sides of the skull.

[0072] This positioning allows the transit time to be measured through the cross section along the skull.

[0073] The emitter and detector are each positioned in an area above the external ear canal.

[0074] The inventors have discovered that positioning the emitter and detector on opposite areas of the skull above the ear canal is particularly advantageous for imaging the internal state of the brain and thus allows the properties of the brain to be classified particularly reliably.

[0075] The travel time values ​​of the one or more ultrasound signals through the skull can be measured by: emitting a plurality of wave packets, each containing an n-periodic signal of constant frequency at each of the successive time points; measuring the travel time values ​​through the cross-section of the skull for each of the n periods of the periodic signal at each of the plurality of wave packets.

[0076] This discloses a particularly advantageous measuring method for providing a sufficient number of transit time values ​​for processing, whereby the invention is not limited to this specific measuring method. A wave packet is emitted at a time by a probe, in particular the emitter. A wave packet comprises a periodic signal with n periods, where n is an integer. The wave packet passes through the cross-section of the skull and is detected at the probe on the opposite side of the skull. A transit time value is determined for each of the n periods of the wave packet. This means that n transit time values ​​can be assigned to each time point. This increases the number of available measured values ​​for later processing of the measured values. By increasing the number of measured values, the signal better describes the temporal course of the internal state of the intracranial pressure in the brain.This makes it possible to achieve a more reliable evaluation of the measured transit time values. Ultimately, the classification of intracranial pressure characteristics is more reliable.

[0077] In one aspect, receiving data with traveltime values ​​comprises: receiving a number of n traveltime curves, wherein in each of the n traveltime curves a traveltime value is assigned to one of the consecutive times; for each of the n traveltime curves, subtracting an offset from the traveltime values ​​in the traveltime curve; forming an averaged traveltime curve by averaging the traveltime values ​​of each of the n traveltime curves at one of the consecutive times.

[0078] In a further aspect, in one of the n-time travel curves, a travel time value at a time is formed by a period of an n-periodic wave packet of the one or more ultrasonic signals that was emitted or received at the corresponding time.

[0079] From the previous measurement, n transit time values ​​were determined at each time point (where n corresponds to the number of periods in a wave packet). This allows a number of n-time courses of transit time values ​​to be created, whereby in each of the n-time courses the transit time value of an identical period in a wave packet is assigned to a time point in the wave packet. Each of the time courses thus forms a transit time curve. This creates n transit time curves. An offset is subtracted from each of the n transit time curves. The offset can represent a variety of factors that are not directly correlated with cardiac activity. For example, the offset can be used to subtract frequency-dependent interference signals from the measuring equipment or a patient's respiration. An averaged transit time curve is created by averaging the transit time values ​​at corresponding time points of the n transit time curves.

[0080] This averaged time-of-flight curve can be used for further data processing, for example, by forming one or more segments and / or forming the averaged peak. This aspect processes the measured data in such a way that the data are freed from any background that is not directly correlated with cardiac activity. Averaging provides a meaningful representation of the changing state of the brain along the cross-section. This allows for better identification of morphology over time based on the measured time-of-flight values. Thus, classification based on the time-of-flight values ​​becomes more reliable.

[0081] In one aspect, each of the n runtime curves has a periodicity given by the heart rate.

[0082] Thus, the measured values ​​and those received by the computer reflect the change in intracranial pressure induced by the heartbeat.

[0083] In one aspect, the one or more temporal segments for determining the at least one segment are selected from the averaged runtime curve.

[0084] Thus, the averaged runtime curve is used as the basis for the temporal segments to determine the at least one characteristic peak. This results in better and more reliable data quality for determining the at least one characteristic peak, making the characterization of the intracranial pressure properties more reliable.

[0085] In one aspect, a bandpass filter is applied to the delay values ​​in a delay curve to determine the offset.

[0086] The bandpass filter allows low-frequency components (e.g., a trend or respiratory effects) to be cut off. This allows frequency components not directly correlated with cardiac activity to be cut off, allowing for more reliable classification of intracranial pressure characteristics.

[0087] Furthermore, the object mentioned above is achieved by a device, in particular a computer, for classifying intracranial pressure, configured to carry out the method according to the invention. The computer can comprise one or more processors configured to carry out the method according to the invention. Furthermore, the computer can comprise a storage medium on which instructions are stored which, when executed by the one or more processors, cause the one or more processors to carry out the method according to the invention.

[0088] Furthermore, the object mentioned at the outset is achieved by a device arrangement comprising the device described above and at least two probes for determining the transit time values ​​of one or more ultrasound signals through a cross-section of a skull, wherein the at least two probes comprise at least one emitter for outputting the one or more ultrasound signals and at least one corresponding detector for detecting the outputted one or more ultrasound signals, wherein in particular the at least one emitter and the at least one detector are positionable on opposite sides of the skull, in particular in a plane through a frontal cross-section of the skull.

[0089] In one aspect, the at least one emitter and the at least one detector are each positionable in a region above the external ear canal.

[0090] In one aspect, the at least two probes are configured to output a plurality of wave packets, each having an n-periodic signal of constant frequency, at each of the successive points in time, and to detect travel time values ​​through the cross-section of the skull for each of the n periods of the periodic signal to each of the plurality of wave packets.

[0091] Furthermore, the object mentioned at the outset is achieved by a computer program comprising instructions which, when the method according to the invention is carried out by a computer, cause the computer to carry out the method according to the invention, as well as a computer-readable medium on which the computer program is stored.

[0092] In the following, further properties, features and advantages of the invention will become clear by describing preferred embodiments of the invention with reference to the accompanying exemplary drawings, in which:

[0093] Fig. 1 shows an example of a measuring arrangement during the determination of the travel time values ​​of one or more ultrasound signals through a skull.

[0094] Fig. 2 shows a verification of the measured time courses of the runtime values. Fig. 3 shows examples of the directly measured values ​​and their preprocessing.

[0095] Fig. 4 shows an example of a classifier.

[0096] Fig. 5 shows an example of a detailed view of the classifier.

[0097] Fig. 6 shows an example of the determination of minima on a data set.

[0098] Fig. 7 shows an example of a grouping of individual peaks in a data set.

[0099] Fig. 8 shows an example of a characteristic peak.

[0100] Fig. 9 shows an example of a chronological sequence of the process steps.

[0101] Fig. io shows an example of an arrangement for classifying intracranial pressure.

[0102] The features disclosed in the above description, the figures and the claims may be important both individually and in any combination for the realization of the invention in the various embodiments.

[0103] Reference symbols in the figures refer to the same elements.

[0104] Fig. 1 shows an example of a measuring arrangement for determining the transit time values ​​of one or more ultrasound signals through a skull cross-section.

[0105] Shown are two probes positioned on opposite sides of the skull 1 with a brain 2. The two probes 3 comprise at least one emitter configured to emit one or more ultrasound signals, and at least one detector configured to receive the one or more ultrasound signals emitted by the at least one emitter. The one or more ultrasound signals can be implemented by a longitudinal wave. Furthermore, the one or more ultrasound signals can comprise one or more wave packets with a constant frequency. The at least one emitter and the at least one detector are positioned on different opposite sides of the skull 1.

[0106] Between emission and detection, the one or more ultrasound signals traverse the cross-section of the skull 1. A sound field covering a certain area of ​​the skull cross-section is thus formed between the emitter and the detector, through which the one or more ultrasound signals traverse. The sound field can preferably be aligned parallel to a frontal plane of the skull i.

[0107] As described above, the transit time of one or more ultrasound signals measured at different points in time depends on a change in the internal state of the intracranial pressure of brain 2. A change in the transit time values ​​between two consecutive points in time thus represents a change in the state of the components in skull i.

[0108] In the following, a verification of the measured transit time values ​​is explained by comparing the measured transit time values ​​with a time course of the intracranial pressure measured according to the gold standard on the same patient at the same time.

[0109] Sections (a) and (c) of Fig. 2 show intracranial pressure measurements performed using the gold standard method for intracranial pressure measurement on a patient with normal intracranial pressure (section (a)) and abnormal intracranial pressure (section (c)). Furthermore, sections (b) and (d) of Fig. 2 show transit time measurements performed simultaneously with the intracranial pressure measurements on the same patients with normal intracranial pressure (section (b)) and abnormal intracranial pressure (section (d)). The transit time values ​​shown were measured and preprocessed using the method described in more detail in Fig. 3.

[0110] The intracranial pressure measurements shown in sections (a) and (c) and performed according to the gold standard serve as a reference. As described above, the gold standard for intracranial pressure measurements involves inserting pressure probes into the brain, thus measuring intracranial pressure directly, i.e., relatively reliably. This measurement method is generally accepted and can therefore serve as a reference.

[0111] Section (a) of Figure 2 shows a temporal progression of the directly measured intracranial pressure (ICP) in a patient with normal intracranial pressure. The temporal progression of the intracranial pressure exhibits a characteristic curve with several peaks, each representing an increase in pressure during a systolic phase and a decrease in pressure during a diastolic phase. These peaks clearly exhibit a consistently specific shape, which can be attributed to the elasticity of the patient's brain 2 within a normal range, thus reliably indicating that the intracranial pressure is within a normal range.

[0112] Section (c) of Figure 2 shows a time course of the directly measured intracranial pressure (ICP) in a patient with abnormal intracranial pressure. The time course of the intracranial pressure also exhibits a characteristic curve, which includes several peaks correlating to a systolic and diastolic phase. As can be seen from section (c) of Figure 2, the curve has a significantly different shape compared to the curve of the time course of the intracranial pressure in a patient with normal intracranial pressure. The shape of the curve can therefore reliably indicate abnormal intracranial pressure.

[0113] From the curves of the time courses of the intracranial pressure measurements according to the gold standard, it is possible to reliably conclude whether the intracranial pressure is normal or abnormal.

[0114] In contrast, section (b) of Figure 2 shows a temporal progression of the transit time values ​​of one or more ultrasound signals through the skull cross-section, measured using the described arrangement, which were simultaneously determined on the same patient with normal intracranial pressure as in section (a). The temporal progression of the transit time values ​​also exhibits a characteristic curve with multiple peaks, which can thus also be assigned to a systolic and diastolic phase. As can be seen, the curve of the transit time values ​​has a very similar shape to the curve of the temporal progression of the intracranial pressure shown in section (a). It can therefore be reliably concluded that the curve of the temporal progression of the transit time values ​​directly correlates with the change in intracranial pressure in the skull 1, and based on the curve of the temporal progression of the transit time values, it can be concluded that the patient's intracranial pressure is normal.

[0115] Section (d) shows a time course of the transit time values ​​of one or more ultrasound signals through the skull cross-section, measured using the described arrangement, which were determined simultaneously in the same patient with abnormal intracranial pressure as in section (c). The time course of the transit time values ​​also exhibits a characteristic curve with multiple peaks, which can thus also be assigned to a systolic and diastolic phase. As can be seen, the curve of the transit time values ​​has a very similar shape to the curve of the intracranial pressure shown in section (c). It can therefore be reliably concluded that the curve of the transit time values ​​directly correlates with the change in intracranial pressure in the skull 1. Based on the curve of the time course shown in section (d), it can therefore be reliably concluded that the patient has abnormal intracranial pressure.

[0116] These simultaneous measurements were repeated in several patients with normal and abnormal intracranial pressures, and fundamentally similar correlations were observed between the time courses of intracranial pressure measured according to the gold standard and the time courses of the second-line values. Patient selection was based on the physiological inclusion criteria mentioned above.

[0117] Thus, the curve of the time course of the transit time values ​​measured at successive points in time can be used to reliably determine whether intracranial pressure is normal or abnormal.

[0118] This finding can be explained by the fact that a change in the state of intracranial pressure within the skull 1 directly correlates with a change in the speed of sound and thus with a travel time of the one or more ultrasound signals through the cross-section of the skull i.

[0119] The transit time values ​​of one or more ultrasound signals through a cross-section of a skull i measured at successive points in time thus capture the transit time differences resulting from the pressure-induced volume change and a pressure-induced compression of the components of the brain 2, such as blood, cerebrospinal fluid and tissue.

[0120] The temporal progression of the transit time values ​​measured at consecutive points in time thus reflects the volume and density changes of the components of Brain 2 and thus of the intracranial pressure during a pulse wave, allowing for reliable non-invasive classification of intracranial pressure. A correlation between the transit time values ​​of one or more ultrasound signals through the brain and an internal state of intracranial pressure in Brain 2 has been scientifically proven.

[0121] In connection with Fig. 3, the measurement of the travel time values ​​of the one or more ultrasound signals through the brain 2 as well as the subsequent preprocessing of the data of the measured travel time values ​​are explained in more detail.

[0122] It should be noted that the measurement method and preprocessing of the measured values ​​are presented only as examples, and other measurement methods and data preprocessing are also possible to enable sufficient data quality for classifying intracranial pressure based on a temporal progression of transit time values. For example, other measurement methods can be used that provide sufficient data quality, but do not require corresponding data processing to determine an analyzable temporal progression.

[0123] A wave packet of constant frequency containing one or more ultrasonic signals is emitted by the emitter at one of the consecutive points in time. A wave packet comprises a periodic signal with n periods, where n is an integer. The wave packet passes through the cross-section of the skull 1 and is detected by the detector on the opposite side of the skull 1. For each of the n periods of the wave packet, a travel time value from the emitter to the detector is determined. This means that each point in time is assigned n travel time values.

[0124] From these measurements, n traveltime values ​​are determined at each point in time, with each of the n traveltime values ​​being assigned a point in time. Then, a number of n time profiles of traveltime values ​​are created, with each of the n time profiles assigning a traveltime value of the same period in a wave packet to a point in time within the wave packet. Each of the n time profiles thus forms a traveltime curve. This creates n traveltime curves corresponding to the number of n periods in a wave packet.

[0125] Fig. 3 (a) shows one of the n propagation time curves with the directly measured propagation time values. The x-axis indicates the time points in seconds, and the y-axis indicates the propagation time values ​​in microseconds. In the propagation time curve, each successive time point is assigned a propagation time value of a period of the periodic signal of a wave packet. The propagation time values ​​fluctuate over the approximately 5 seconds shown due to the heartbeat. In addition, there are n further propagation time curves (not shown), with each of the n propagation time curves assigning a propagation time value with the same period to the corresponding time point.

[0126] An offset is subtracted from each of the n time-of-flight curves. The offset is subtracted by applying a bandpass filter to the time-of-flight values ​​in a time-of-flight curve. The bandpass filter can be used to cut off low-frequency components (for example, a trend or breathing effects).

[0127] Fig. 3 (b) shows an example of the runtime curve from Fig. 3 (a) after subtracting an offset (e.g. constant and trend), where the X-axis shows the times in units of seconds and the Y-axis shows the runtime values ​​in nanoseconds.

[0128] An averaged runtime curve is then calculated by averaging the runtime values ​​of each of the n runtime curves at one of the consecutive points in time. In the averaged runtime curve, an averaged runtime value from the corresponding n runtime values ​​is assigned to a point in time. The corresponding averaged runtime curve is shown as an example in Fig. 3©, where the points in time are given in seconds on the x-axis and the runtime values ​​in nanoseconds on the y-axis.

[0129] The averaged time-of-flight curve can be used as a basis for further data processing. However, with sufficient signal quality, time-of-flight values ​​from data that have not undergone such a measurement and processing process can also be used as a basis for further processing.

[0130] Furthermore, the temporal profiles of the runtime values ​​can undergo an optional preparatory step for classification. In the preprocessing step for classification, the directly measured temporal profiles of the runtime values ​​can be divided into temporal segments of consecutive points in time with the correspondingly assigned runtime values. These segments serve as the basis for data preprocessing, and several segments are formed with a correspondingly averaged runtime curve. In other aspects, temporal segments of consecutive points in time with the correspondingly assigned runtime values ​​can be formed from the averaged curve and used as the basis for further processing.

[0131] From the segments thus formed, one or more segments can be selected which meet a quality criterion.

[0132] The quality criterion can be formed by an autocorrelation function, which evaluates the intrinsic similarity of the temporal course of the travel time values ​​in a segment. Additionally or independently, the quality criterion can be formed by comparing the maximum amplitudes of the most dominant frequency component with the amplitudes of other frequency components in the temporal course of the travel time values ​​in a segment. Additionally or independently, the quality criterion can be formed by evaluating frequencies in the temporal course of the travel time values ​​in a segment.

[0133] The selected segments can be used as a basis for further processing, in particular by loading them into a classifier.

[0134] Fig. 4 shows an example of the functionality of a classifier for classifying the property of the time course of the runtime curve. The classifier is implemented by the computer.

[0135] Data with traveltime values ​​of one or more ultrasound signals measured at consecutive points in time through a cross-section of a skull i are input into the classifier. The input data can be directly measured traveltime values ​​that have not undergone any preprocessing steps, data that have undergone a preprocessing step, in particular the averaged traveltime curve, or segments selected according to the quality criterion.

[0136] In the classifier, a characteristic of intracranial pressure is classified based on one or more features of a representative curve of the time course of the runtime values, in order to subsequently output a classified characteristic of intracranial pressure. Figure 5 shows a detailed view of the classifier to illustrate the classification of the intracranial pressure characteristic.

[0137] As described, data with runtime values ​​measured at consecutive points in time are fed into the classifier.

[0138] In one aspect, several of the selected segments are loaded into the classifier and combined to create a temporal progression of the runtime values. In another aspect, the raw runtime values ​​are loaded into the classifier. In an additional aspect, the averaged runtime curve is loaded into the classifier.

[0139] In a corresponding temporal progression of the runtime values, individual peaks are first determined by detecting minima in the temporal progression between neighboring peaks. A corresponding data set is shown in Fig. 6, with successive time points in IO on the x-axis. A(-2) seconds are indicated, and the corresponding runtime values ​​in nanoseconds are shown on the y-axis. The detected minima are indicated by the unfilled points.

[0140] In a next step in Fig. 5, the peaks thus determined are extracted and temporally normalized to a common starting time.

[0141] In an optional subsequent step, peaks that are too short or too long can be eliminated. The elimination of peaks that are too short or too long is determined by determining whether a peak width at the detected minima lies below or above a reference value, which is determined by a median or mean of the peak widths of the cutout peaks at their minima.

[0142] In a first step, the extracted and normalized peaks are grouped. Grouping is achieved by applying a similarity criterion to the individual peaks. The group of peaks with the most peaks is selected for further processing.

[0143] The grouping of the cut-out and normalized peaks is shown as an example in Fig. 7, where the time-normalized time points are shown on the X-axis in units of 10 A (-2) seconds, and on the Y-axis, the runtime values ​​normalized to a maximum peak amplitude are shown without units. In this example, the group of peaks with amplitude maxima within a range of approximately 2 seconds is selected for further processing (represented by solid lines in the curves).

[0144] By calculating the median of the respective travel time values ​​of the peaks of the selected curve at a given time, the at least one characteristic peak of the representative curve can be formed. Thus, the median is calculated from the travel time values ​​of individual peaks at a given time. The at least one characteristic peak thus characterizes peaks of a plurality of peaks in the temporal progression of the travel time.

[0145] A peak determined in this way is shown as an example in Fig. 8, where the x-axis shows the normalized time points in units of IO A (-2) seconds are given and on the Y-axis the normalized runtime values ​​of the characteristic peak are given without units.

[0146] From the at least one characteristic peak, features of the curve of the at least one characteristic peak are then determined.

[0147] As described above, these features were determined during a learning phase. A feature of the one or more features can, for example, be a similarity of the curve shape of the at least one characteristic peak to a synthetic function, in particular a continuous wavelet transform. Furthermore, a further feature can be a peak width of the at least one characteristic peak at a certain percentage of the maximum amplitude of the at least one characteristic peak.

[0148] Based on one or more features of the characteristic peak, the properties of the intracranial pressure can be classified, particularly whether the intracranial pressure is normal or abnormal. Based on the classification, a classification can be determined and output by the computer. Based on the output classification, a diagnosis of a disease can be made.

[0149] Fig. 9 shows a chronological sequence of the steps of the previously described method. In a first step, the travel time values ​​of one or more ultrasound signals through a skull i are measured.

[0150] In a second step, the measured runtime values ​​can be preprocessed for further processing.

[0151] In a third step, the runtime values ​​can be prepared for classification.

[0152] Preprocessing the runtime values ​​and preparing them for classification in the second and third steps are optional. For example, the measured runtime values ​​can be loaded directly into the classifier.

[0153] In a fourth step, a property of the intracranial pressure is classified based on one or more features of a representative curve of a time course of the transit time values ​​in order to obtain a classified property of the intracranial pressure.

[0154] In a fifth step, the classification of the property of the intracranial pressure is output.

[0155] Fig. io shows an example of an arrangement for classifying intracranial pressure.

[0156] The device arrangement comprises two probes 3. The two probes 3 comprise an emitter for emitting one or more ultrasound signals and a detector for detecting the one or more ultrasound signals. The probes 3 can, for example, be part of a measuring device. Using the probes 3, the transit time values ​​of the one or more ultrasound signals from the emitter to the detector through the skull 1 are measured.

[0157] The measured transit time values ​​for the ultrasound signals are transmitted to a computer 4. The transmission takes place via a suitable medium. For example, data with the transit time values ​​can be transmitted to the computer 4 via a data storage device, such as a USB stick. In other examples, the data with the transit time values ​​can be transmitted to the computer 4 via a cable. In further examples, the data can be transmitted wirelessly, for example via a WLAN or Bluetooth connection, to the computer 4. The probes 3 or the measuring device are in a corresponding communication connection with the computer 4 for this purpose. The computer 4 carries out the previously described method for classifying intracranial pressure based on the received data with transit time values. The output can be sent to a display connected to the computer 4. The classified property of the intracranial pressure can then be displayed to a user on the display.

[0158] Furthermore, a computer-readable medium 5, such as a storage medium or a data carrier, is provided that can be connected to the computer. The computer-readable medium 5 comprises instructions that, when executed by the computer, cause the computer to perform the steps according to the described method.

[0159] The above-described aspects and embodiments of the technology described herein may be implemented in numerous ways. For example, the embodiments may be implemented using hardware, software, or a combination thereof.

[0160] When a feature is implemented in software, the software code may be executed on any suitable computer, processor, or collection of processors, whether provided in a single computer or distributed across multiple computers. Such processors may be implemented as integrated circuits, with one or more processors in an integrated circuit component, including commercially available integrated circuit components known in the art by names such as CPU chips, GPU chips, microprocessor, microcontroller, or coprocessor. Alternatively, a processor may be implemented as an ASIC or as a programmable logic device configuration. As yet another alternative, a processor may be a part of a larger circuit or semiconductor device.As a specific example, some commercially available microprocessors have multiple cores, so one or a subset of these cores can form a processor. However, a processor can be implemented using circuitry in any suitable format.

[0161] Furthermore, a computer may be embodied in any of a number of forms, such as a rack-mounted computer, a desktop computer, a laptop computer, or a tablet computer. Furthermore, a computer may be embedded in a device that is not generally considered a computer but has suitable processing capabilities, including a personal digital assistant (PDA), a smartphone, or other suitable portable or fixed electronic device.

[0162] Such computers may be interconnected by one or more networks of any suitable form, including a local area network or a wide area network, such as a corporate network or the Internet. Such networks may be based on any suitable technology and may operate according to any suitable protocol, and may include wireless networks, wired networks, or fiber optic networks.

[0163] Furthermore, the various methods or processes outlined herein may be encoded as software executable on one or more processors using any of a variety of operating systems or platforms. Moreover, such software may be written using any of a variety of suitable programming languages ​​and / or programming or scripting tools, and may also be compiled as executable machine language code or intermediate code running on a framework or virtual machine. The terms "(computer) program" or "software" are used herein in a general sense to refer to any type of computer code or set of computer-executable instructions that can be used to program a computer or other processor to implement various aspects of the present invention as discussed above.It should also be noted that, according to one aspect of this embodiment, one or more computer programs which, when executed, perform methods of the present invention need not reside on a single computer or processor, but may be distributed in a modular manner among a number of different computers or processors to implement various aspects of the present invention.

[0164] The invention may be embodied as a computer-readable (storage) medium (or multiple computer-readable media) (e.g., a computer memory, one or more floppy disks, compact discs (CDs), optical disks, digital video disks (DVDs), magnetic tapes, flash memory, circuit configurations in field-programmable gate arrays or other semiconductor devices, or other tangible computer storage media) encoded with one or more programs that, when executed on one or more computers or processors, perform methods that implement the various embodiments of the invention discussed above. As can be seen from the foregoing examples, a computer-readable storage medium can store information for a sufficient time to provide computer-executable instructions in a non-transitory form.Such a computer-readable storage medium or media may be transportable so that the program or programs stored thereon may be loaded onto one or more different computers or other processors to implement various aspects of the present invention, as discussed above. As used herein, the term "computer-readable (storage) medium" encompasses only a non-transitory computer-readable medium that may be considered a physical article (i.e., an article of manufacture) or a machine. Alternatively or additionally, the invention may be embodied as a computer-readable medium other than a computer-readable storage medium, such as a propagating signal.

[0165] Computer-executable instructions or commands can take many forms, such as program modules executed by one or more computers or other devices. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. Typically, the functionality of the program modules can be combined or distributed as desired in various embodiments.

[0166] Features of various aspects or embodiments have been described above by way of example so that those skilled in the art can better understand the present invention. However, it is clear that other aspects and embodiments than those described in detail may also be subject of the invention, the invention being defined by the scope of the appended claims.

[0167] 1 skull

[0168] 2 brain 3 probes

[0169] 4 computers

[0170] 5 Computer-readable medium

Claims

Claims 1. A computer-implemented method for classifying intracranial pressure, the method comprising: Receiving data with traveltime values ​​based on traveltime values ​​measured at successive times for one or more ultrasound signals through a cross-section of a skull (i); Classifying a property of the intracranial pressure based on one or more features of a representative curve of a time course of the transit time values ​​to obtain a classified property of the intracranial pressure; and Output the classified property of intracranial pressure.

2. The method according to claim 1, wherein the temporal course of the transit time values ​​correlates to a temporal course of the intracranial pressure within the skull (i).

3. Method according to one of the preceding claims, wherein the temporal course of the transit time values ​​has several peaks, which each correlate to an increase and a decrease in the intracranial pressure during a systolic and a diastolic phase.

4. Method according to one of the preceding claims, wherein a change between transit time values ​​in the time course at at least two consecutive points in time correlates directly to a change in the intracranial pressure at the at least two consecutive points in time.

5. Method according to one of the preceding claims, wherein the classified property of the intracranial pressure indicates a normal intracranial pressure, in particular an intracranial pressure of 10 mmHg or lower, or indicates an abnormal intracranial pressure, in particular an intracranial pressure of 15 mmHg or more.

6. Method according to one of the preceding claims, wherein the representative curve has at least one characteristic peak. 7- The method of claim 6, wherein the characteristic of the intracranial pressure is determined by extracting the one or more features of the curve of the at least one characteristic peak by the computer.

8. The method of claim 7, wherein a first characteristic feature of the one or more features is a peak width of the at least one characteristic peak at a certain percentage of the maximum amplitude of the at least one characteristic peak.

9. The method according to one of claims 7 or 8, wherein a second characteristic feature of the one or more features is a similarity of the curve shape of the at least one characteristic peak to a synthetic function, in particular a continuous wavelet transform.

10. The method according to any one of claims 6 to 9, wherein the at least one characteristic peak is characteristic of a plurality of peaks in at least one time segment of the time course of the runtime values.

11. The method according to claim 10, wherein the at least one characteristic peak is determined by the following steps: Extracting several individual peaks from the at least one temporal segment of the temporal course of the runtime values, Normalizing the time points of the run second values ​​in the individual peaks, Grouping the individual peaks by applying a similarity criterion for the individual peaks, Identify the group with the largest number of peaks, and Forming the at least one characteristic peak from the peaks in the identified group of peaks.

12. The method of claim 11, wherein forming the at least one characteristic peak comprises: for each time point, forming a median of the travel time values ​​of the peaks from the identified group.

13. The method according to any one of claims 10-12, wherein the at least one segment is determined by the following steps: Extracting one or more temporal segments from the temporal course of the runtime values, where a segment has several peaks in the runtime values, Determining the at least one temporal segment by selecting one or more temporal segments in which the peaks meet a quality criterion.

14. The method according to claim 13, wherein the quality criterion is formed by an autocorrelation function which evaluates an intrinsic similarity of the temporal course of the runtime values ​​in a segment.

15. The method according to claim 13 or 14, wherein the quality criterion is formed by comparing the maximum amplitudes of the most dominant frequency component with amplitudes of other frequency components in the temporal course of the transit time values ​​in a segment.

16. Method according to one of claims 13-15, wherein the quality criterion is formed by an evaluation of frequencies in the temporal course of the runtime values ​​in a segment.

17. The method according to any one of the preceding claims, wherein receiving data with runtime values ​​comprises: Receiving a number of n runtime curves, wherein in each of the n runtime curves a runtime value is assigned to one of the successive points in time, For each of the n runtime curves, subtracting an offset from the runtime values ​​in the runtime curve, Forming an averaged runtime curve by averaging the runtime values ​​of each of the n runtime curves at one of the consecutive times.

18. The method according to claim 17, wherein in one of the n-time travel curves a time travel value at a time is formed by a period of an n-periodic wave packet of the one or more ultrasonic signals which was emitted or received at the corresponding time.

19. The method according to claim 17 or 18, wherein each of the n runtime curves has a periodicity given by the heart rate.

20. The method according to any one of claims 17-19, wherein the one or more time segments for determining the at least one segment are selected from the averaged runtime curve.

21. Method according to one of claims 17-20, wherein a bandpass filter is applied to the delay values ​​in a delay curve to determine the offset.

22. Device, in particular a computer (4), arranged to carry out the method according to one of the preceding claims.

23. Device arrangement, comprising the device according to claim 22 and at least two probes (3) for determining the transit time values ​​of the one or more ultrasound signals through a cross-section of a skull (1), wherein the at least two probes (3) comprise at least one emitter for outputting the one or more ultrasound signals and at least one corresponding detector for detecting the output one or more ultrasound signals, wherein in particular the at least one emitter and the at least one detector are positionable on opposite sides of the skull (1), in particular in a plane through a frontal cross-section of the skull (1).

24. The device arrangement of claim 23, wherein the at least one emitter and the at least one detector are each positionable in a region above the external ear canal.

25. Device arrangement according to one of claims 23 or 24, wherein the at least two probes (3) are designed to output a plurality of wave packets each having an n-periodic signal of constant frequency at one of the successive times, to detect travel time values ​​through the cross-section of the skull (1) for each of the n periods of the periodic signal to each of the plurality of wave packets.

26. A computer program comprising instructions which, when the method is executed by a computer (4), cause the computer (4) to execute the method according to any one of claims 1 to 21.

27. A computer-readable medium (5), wherein the medium comprises instructions which, when the method is executed by a computer (4), cause the computer (4) to execute the method according to any one of claims 1-21, or wherein the computer program according to claim 26 is stored on the medium (5).

Images