Patent Application
20140358000
The present invention relates to pulsatile flow and, more particularly, to selecting pulse cycles representative of the flow.
Commercial duplex ultrasound systems are used extensively to localize blood vessels and obtain flow characteristics from the blood vessels. For example, in obstetrics, applications exist for uterine arteries, the umbilical artery, the mid cerebral artery and in cardiac applications, the carotid artery and so on. A duplex ultrasound system combines the modality of real-time, two dimensional, pulse-echo imaging of anatomical structures with that of a Doppler ultrasound system from which the Doppler frequency shift or the velocity information is obtainable from a blood vessel.
The use of ultrasound in vascular applications to perform Doppler velocimetry requires the accurate computation of flow parameters to produce consistent, reproducible and reliable diagnosis.
The accuracy with which flow parameters are computed is dependent on the cycles chosen by a sonologist or doctor, with good cycles being selected manually.
Doppler exams typically require a great degree of skill to obtain a clinically useful measurement. For example, correct orientation of the probe with respect to the vessel is essential to ensure that the beam-flow angle is less than 60 degrees. Errors in measurements are amplified when angles of greater than 60 degrees are used in the determination of velocities. The standard workflow on a clinical ultrasound scanner allows a sonographer to determine the orientation of the probe with respect to the vessel using a standard B-mode and Color Flow display. The spectral Doppler measurements are then obtained thus ensuring that the measured velocities are correct.
The use of ultrasound in vascular applications to perform Doppler velocimetry requires availability of skilled personnel.
In emerging market countries such as India, the shortage of specialists limits the availability and access to ultrasound. Hence, an automated method of acquiring and evaluating Doppler signals for clinical diagnosis (without requiring the user to interpret an ultrasound scan image) would be useful to non-radiologists such as OB/GYN or cardiologists who are the primary treatment providers.
In addition, a low-cost system is essential to provide an attractive solution in emerging market environments. Devices that are currently available on the market for antenatal check-ups and labor are the ultrasound scanner and the fetal monitor/cardiotoco graph (CTG) machine. However, both of these devices are relatively expensive.
There exists a need for a low-cost, easy-to-use solution to provide Doppler velocimetry to screen for and monitor high risk pregnancies.
In addition, even with duplex ultrasound systems and likewise in medical applications other than obstetrics, the manual, i.e., visual, selection of good pulse cycles requires specialized skill and is a tedious and time consuming task. In particular, a good cycle is one which represents the actual hemodynamic profile in a vessel.
Also, the judgment as to what constitutes a good cycle varies among observers.
Especially in emerging markets such as India, automatic selection is needed.
Commonly-assigned patent applications entitled “Automated Doppler Velocimetry Using a Low-Cost Transducer” and “Excitation Schemes for Low-Cost Transducer Arrays”, the entire disclosure of both applications being incorporated herein by reference, disclose a hand-held, stand-alone, Doppler-based, ultrasound probe whose examining face is less finely divided into separate transducer elements, i.e., for relatively few separate elements. As mentioned therein, the probe operates automatically without the need for interpreting a visual display of anatomy.
The present patent application is directed to novel, automatic pulse cycle selection, with particular application to the probe referred to immediately above and to user-interactive-imaging systems such as duplex ultrasound systems.
In accordance with an aspect of the present invention, a device is configured for examining pulsatile flow, for deriving, based on the examined flow, spectral characteristics and for, based on the derived characteristics, determining which one or more pulse cycles are to be selected as representative of the flow.
In accordance with a sub-aspect, the selected cycles are consecutive.
In accordance with another sub-aspect, the selecting chooses a predetermined number of cycles. That number can be five.
In accordance with one other sub-aspect, the device includes a display and is further configured for, responsive to the determining, displaying a selected cycle or selected cycles.
According to a different sub-aspect, the device is further configured for operating on a selected cycle to compute a clinical parameter. The clinical parameters are typically computed on each of the selected five cycles and an average value from these five cycles is taken as clinical parameter value.
In accordance with a further sub-aspect, the device includes a display and is further configured for, responsive to the computing, displaying the clinical parameter.
In accordance with a related sub-aspect, the flow is that of a blood vessel.
According to a complementary sub-aspect, the examining includes receiving ultrasound. The device is further configured for generating, from the received ultrasound, the cycles subject to selection.
In a yet another sub-aspect, a handheld, stand-alone, diagnostic apparatus incorporates the device.
In a related sub-aspect, the device features transducer elements and is configured not to collectively use any of these elements to focus, nor to steer, any beam used for the examining to perform the deriving.
In accordance with an additional sub-aspect, the device includes a user interface for specifying a blood flow parameter, for use in the determining, and/or a medical application for which the determining is performed.
According to one other additional sub-aspect, the device includes a user interface comprising a display. The device is further configured for making the selection by choosing a plurality of segments that each are made up of multiple ones of the cycles and for, on the display, displaying the segments for user selection via the user interface.
According to a further, supplementary sub-aspect, the device is further configured for distinguishing the displayed segments by highlighting them.
In a yet different sub-aspect, a spectrogram includes the cycles subject to the selection.
In a further sub-aspect, the cycles subject to the selection respectively have a plurality of specified parameters. The determining excludes from the selection a cycle if any of the specified parameters of the cycle deviates, by more than a predetermined amount, from a respective average derivable from said spectrogram.
In a yet further sub-aspect, the average is a median.
As a particular sub-aspect, a cycle having a plurality of specified parameters is further configured for: a) by specified parameter, using as an exponent the absolute value of the deviation of the parameter from an average to form a term; and b) summing the terms over the plurality of parameters to yield a cycle quality metric.
In a further sub-aspect, the base for said exponent is a function of the base of the natural logarithm.
Alternatively or in addition, in a further sub-aspect, the determining includes summing the metrics to yield a cycle-series segment quality metric.
In a yet different sub-aspect, an example of a characteristic is a Doppler-spectral-waveform caliper measurement on a cycle. The determining entails computing a deviation of the measurement.
In yet one other sub-aspect, a selected cycle is representative of frequency shift versus time.
In yet an additional sub-aspect, the device is configured for, based on at least one of a group of conditions, filtering out a cycle of said cycles subject to the selection, said group consisting of: a) existence in said cycle of at least one of more than one peak and more than one valley; b) frequency response in a spectral band of said cycle not exceeding a predetermined threshold; c) said cycle not reaching its peak systolic value within a predetermined duration; and d) said cycle having less than a predetermined length.
Details of the novel, automated pulse cycle selection device are set forth further below, with the aid of the following drawings, which are not drawn to scale.
The description of what is proposed herein with regard to automatic pulse cycle selection is preceded with what largely is a review of the Doppler-based probe disclosed in the commonly-assigned patent applications mentioned herein above.
The probe 100 is implementable as an automatic, handheld, stand-alone, self-contained, ultrasound examination device. It has a transducer housing 120 and a handle 122.
Within the transducer housing 120, a non-phased, two-dimensional transducer array 124 is comprised of transducer elements 126, the number of elements being determined by the scan volume and anatomy. Data acquisition occurs individually by element 126, although, as discussed in more detail further below, elements are operable concurrently to shorten the total acquisition time period.
As seen in
The total of merely 32 elements 126 stands in stark contrast to the much greater number of elements that would be required in conventional medical imaging to cover the same 6 cm×6 cm volume.
In this regard, electronic focusing for medical imaging, as with a phased-array transducer, requires an inter-element spacing of a ½ wavelength, i.e., ½λ, or less. Doppler ultrasound for imaging can typically range from between 2×106 and 4×106 cycles per second (2 to 4 MHz). Ultrasound travels through soft body tissue at a speed of about 1540 meters/second. Wavelength, i.e., λ, is equal to velocity divided by frequency. Here, this is 1540 m/s divided by approximately 2×106 cycle/s=0.8 millimeter. Medical ultrasound imaging for a display would therefore require an inter-element spacing of less than 0.4 mm, and an element surface area of less than (0.4 mm)2 which is less than 0.2 mm2. Therefore, with a small element size on the order of ½λ, thousands of elements 126 would be required to build a 2D array that, like the one seen in
The spacing (size) of elements in
More generally, the elements 126, in accordance with what is proposed herein, are spaced apart by more than ½λ, although inter-element spacing 128 may be λ, 2λ or more, as discussed hereinabove. The area of the face 132 is, correspondingly, at least 0.6 square millimeters (mm2), and may be more, such as 10 mm2, 25 mm2, or 100 mm2 as in
Advantageously, the automatic ultrasound device 100 does not rely on display of medical images to reach a diagnosis; but, instead, features an array composed of fewer transducer elements and therefore fewer channels. Thus, cost of production is low, while, by virtue of automatic operation, reliability is maintained. Reliability may even be improved, as when medical examinations must be performed at a quicker pace. The automatic operation also tends to reduce examination time, thereby relieving workload, and making the examination more convenient.
During Doppler data acquisition, the elements 126 are fired either sequentially, or in one or more groups taking care that the acoustic signal from one element does not significantly affect others that are excited at the same time. For each element 126, the receive period lags the transmit period. The Doppler receive gate is correspondingly positioned in the receive period so as to enable sampling from a corresponding depth within the volume of interest 106.
On a back surface 134 of the housing 120, so as to face the user, are a number of user-interface, input-output panels which include a top panel 136, a left panel 138 and a right panel 140. An on-off switch 142 and an audio speaker face 144 are disposed in the top panel 136. The left panel 138 frames a function navigation/actuation button 146, a display 148, a Doppler power detection indicator 150, fetal heartbeat acquisition indicator 152, a maternal heartbeat acquisition indicator 154, a normal blood-flow indicator 156, and an abnormal blood-flow indicator 158. The right panel 140 includes three initializing-parameter-entry feedback windows 160, 162, 164.
The elements 126 of the array 124 all are operated to image independently.
This stands in contrast to phased arrays for example, which use multiple separate transducer elements collectively to image or steer a beam. In phased arrays, the steering and focusing is performed by appropriately delaying the input and/or output of elements with respect to other elements.
In accordance with what is proposed herein, the transducer elements of a group are fired simultaneously. The group elements continue imaging concurrently, and independently by element, until expiration of the group's data acquisition time period.
A device for the imaging by groups is configured not to collectively use any of the elements 126 to focus, nor to steer, a beam used in the imaging. By way of demonstration, the transducer elements 166, 168, 170, 172 in
The blood-flow waveform 114 is a graph of Doppler frequency shift versus time and is thereby indicative of blood flow velocity versus time.
Clinical Doppler indices, such as the pulsatility index (PI) 116 and the resistance index (RI) 118 are Doppler angle-independent measures of blood pulsatility. The symbols S, D and C annotating the blood-flow waveform 114 represent, respectively, the peak systolic frequency shift, the end diastolic frequency shift, and the length of one cardiac cycle. Another commonly-used clinical Doppler index is the systolic/diastolic ratio S/D. The symbols S, D and C are Doppler-spectral-waveform caliper measurements. These and other clinical, or blood-flow, parameters such as the clinical Doppler indices are examples of spectral characteristics of pulsatile flow, based upon which the quality of a cycle can be judged. Another spectral characteristic is the frequency response within a spectral band, which if sufficiently low indicates lack of quality of the cycle. Likewise, if a cycle does not achieve its peak systolic value, or its complete cycle length, within a given time period, it is deemed to lack quality.
The probe 100 can utilize the above-identified Doppler indices in identifying blood vessels and in assessing normality of blood flow.
The signal processing involved in classifying, and analyzing, a blood vessel 108-112 found by the probe 100 in the volume of interest 106 and more details on the probe and its use are disclosed in the above-mentioned commonly-assigned patent applications.
A typical good pulse cycle 200, corresponding to a single heart beat, has a significant peak 204 between two valleys 206, 208 within an acceptable time duration since the beginning of the cycle. The acceptable time duration varies in maternal and fetal arteries. The time duration of a maternal artery could vary from 0.6 sec to 1.5 sec and for fetal arteries it could be from 0.3 sec to 1.0 sec. It is also checked that there are no local peaks 204 or valleys 208. Thus, each pulse cycle 200 preferably has a single peak 204 and a single valley 208.
Since the rise and fall in frequency shift during the cycle 200 respectively represent rise and fall in blood flow velocity, a good cycle will exhibit a smooth rise and a smooth fall. The fall part, i.e., from peak to valley, of the cycle 200 undergoes another test. It is checked whether a line 210 joining the peak 204 and the valley 208 that immediately follows the peak crosses the waveform 114 at least once. A crossing point 211 is shown in
Also, some of the spectral information may not be continuously interpretable, as the targeted blood vessel 108-112 may not be within the sample volume due to motion of the patient or sonologist, or the motion of the blood vessel 108-112 itself. Some of the pulse cycles 200 may not be of good quality for interpretation either if the ultrasound probe 100 does not make proper angle with the blood vessel 108-112 or due to noise from neighboring tissue or in-built electronics.
As a result, portions of the spectrogram may be weak, and seen on screen as faded or missing, indicating that associated frequency samples of the Fast Fourier Transform (FFT) that processes the incoming Doppler signal are low in magnitude. Other types of anomalies that are observed in the shape of the spectrogram profile include weak peak strength, incomplete cycle 200, absence of peak 204 and sharp tall peak.
Based on these considerations, some of the cycles 200 are initially filtered out. Spectral characteristics of the surviving cycles 200 are extracted and used in scoring segments of five consecutive, surviving cycles. Relying on five cycles is an acceptable clinical practice. The present inventors have empirically found it prudent to select a series of consecutive cycles 200. The cycle-series segment quality metric 202, shown as Em in
The display 302 continuously shifts across the display screen 314, as from left to right. Any segment 304, 306 currently on screen is selectable, e.g., by touching the touch screen. In the rolling display, the dotted line 320 between the two segments 304, 306 represents temporally where cycles 200, or high scoring (and therefore less desirable) segments, have been excluded. The displayed segments 304, 306 are highlighted, for example by brightness, color, or, as shown here, underlining 324, 328. Along with the highlighting, each segment 304, 306 may be accompanied in its translation across the screen 314 with caliper measurements 224 for each cycle 200 and/or segment scores. When the segment 304, 306 is selected, Doppler parameters 116, 118 for the cycles 200 in the segment are computed and appear on screen.
As an alternative to the rolling segments, the system can automatically select the segment or segment(s) 304, 306 with the lowest scores, compute the respective parameters 116, 118 and display the selected segments and parameters.
If sufficient frequency response is lacking over some spectral band of the current cycle 200 under consideration, as judged from a power threshold (step S408), the current cycle is excluded from selection (step S410).
In either case, if there are more cycles 200 from the data that has been captured in step S404 (step S412), the next cycle is processed (step S414), and processing returns to step S408 with that next cycle serving as the current cycle.
When there are no more cycles 200 to consider (step S412), the cycles that have not been excluded collectively constitue a spectrogram having gaps where cycles have been excluded. Thus, the gapped spectrogram typically has one or more time portions in which cycles have not been excluded. An envelope is, by any known and suitable method, computed for the spectrogram time portion(s) that individually comprise five or more consecutive cycles, and thus have the potential for furnishing a segment of five, consecutive, good cycles 200 (step S415). A method for extracting an envelope from a spectrogram is described in the U.S. Pat. No. 7,611,467 to Zhang, the entire disclosure of which is incorporated herein by reference.
Cycle filtering continues. For example, caliper measurements 224 are made for the current cycle 200, both as to its length of time and the time duration until its peak.
If either measurement 224 exceeds respective a predetermined range of normality (S416), the current cycle 200 is excluded (step S418). Otherwise, if neither range is exceeded, query is made as to whether more than one peak exists in the current cycle (step S420), and as to whether more than one valley exists in the current cycle (step S422). For either condition, the current cycle is excluded (step S418).
In any event, if more cycles are available for consideration (step S424), processing returns to step S416 with the next cycle serving as the current cycle (step S426).
When there are no more cycles for the above-described filtering procedure (step S424), the median 226 is calculated over all surviving cycles. This done for each parameter, such a caliper measurement 224, selected directly or by implication is step S402 (step S428). The calculation is an initial step in computing Em, the cycle-series segment quality metric 202, shown in
For a current segment 304, 306 of five, consecutive ones of the cycles 200 remaining from the above-described filtering, query is made on the deviation 222 for each cycle of the segment. In particular, query is made as to whether the absolute value 220 of the deviation 222 of a parameter from its median 226 exceeds a predetermined deviation threshold. This query is made for each of the selected parameters (step S430). The thresholds may be set so as to detect deviations of 25% or more from the respective median 226. If the deviation threshold is exceeded (step S430), thereby indicating that the current cycle is unworkable, processing shifts cycle-wise past the deviant cycle 200 (step S432).
If, despite the shift, five or more of the cycles 200 that have survived filtering before step S424 are currently available for forming a segment (S434), return is made to step S430 with the next segment 304, 306 serving as the current segment (S435).
If, on the other hand, fewer than five cycles 200 are available (S434), return is made to step S404.
If the deviation threshold is not exceeded for the current segment 304, 306 (step S430), the segment is scored to yield the cycle-series segment quality metric 202 (step S436).
If more segments 304, 306 are desired (step S438), processing branches back to step S434.
Otherwise, if no further scored, good segments 304, 306 are desired (step S438), and the user is to select from among these segments (step S440), the segments are displayed, and highlighted, for selection as described herein above (step S442). The segments 304, 306 are optionally accompanied on screen by caliper measurements 224 and segment scores. Parameters such as Doppler indices are then computed for the segment 304, 306 selected by the user (step S444), and both the segment and indices are displayed on screen 314 (step S446). If, on the othe other hand, no further scored, good segments 304, 306 are desired (step S438), and the selection is automatic (step S440), the segment with the minimum score is selected (step S448). Likewise, parameters are then computed for the selected segment 304, 306 (step S450), and the segment and parameters are displayed (step S452). The computation may involve averaging results for the five cycles 200 of the selected segment 304, 306, with the average being included on the display screen 314.
By virtue of a caliper measurement, operation on a cycle from among the cycles 200 selected is performed to compute a clinical parameter, this all being done automatically in the innnovative technique. More generally, the computation typically involves such measurements on each cycle 200, and averaging of results.
Within the probe 100, 308, control circuitry (not shown), serving as the device proposed herein, can take the form of one or more integrated circuits (ICs). The one or more ICs can, alternatively, be configured for installation into existing apparatus such as ultrasound Duplex scanners.
A signal for operating the device, i.e., IC(s), probe or duplex system, in accordance with the techniques proposed herein can be formed internally, formed by varying an electrical current applied to a wire input to the device, or applied to an antenna for wireless transmission of the signal and reception by a receiving antenna of the device.
A device is configured for examining pulsatile flow, for deriving, based on the examined flow, spectral characteristics and for, based on the derived characteristics, determining which one or more pulse cycles are to be selected as representative of the flow. The cycles selected can be consecutive and amount to a predetermined number of cycles, such as five. The cycles subject to selection may initially be filtered out based on waveform anomalies, with the surviving cycles in a consecutive group of sufficient number being judged based on parameters such as waveform caliper measurements and other types of the characteristics. Good cycles are detected by their lack of variation, with respect to the measured parameters, from each respective, parameter median over the spectrogram cycles not initially filtered. The technique may, according to user selection, take into account additional parameters suited to particular medical application. Uses include correctly identifying an artery by name.
Although methodology of the present invention can advantageously be applied in providing medical diagnosis for a human or animal subject, the scope of the present invention is not so limited. More broadly, techniques disclosed herein are directed to efficiently finding, and subjecting to improved fluid-flow analysis, vessels in body tissue, in vivo, in vitro or ex vivo.
What is proposed herein pertains to selecting good cycles of spectral Doppler waveforms, the selected cycles being representative of blood flow, for rendering a clinical diagnosis based on a result of analyzing the selected cycles. The technique is particularly useful for accurately identifying an artery by name as in the commonly-owned patent application entitled “Automatic Blood Vessel Identification by Name” by the same inventors. Applications of the technology proposed herein include carotid and renal arteries screening, anke-brachial index (ABI) measurements for detecting peripheral arterial disease (PAD), transcranial and cardiac examinations, bleed detection in trauma or other hemorrhages, in addition to fetal well-being assessment.
While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments.
For example, waveform anomalies relating to the shape of a cycle peak can be detected and used in the filtering out of cycles, e.g., step S418.
Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. Any reference signs in the claims should not be construed as limiting the scope.
A computer program can be stored momentarily, temporarily or for a longer period of time on a suitable computer-readable medium, such as an optical storage medium or a solid-state medium. Such a medium is non-transitory only in the sense of not being a transitory, propagating signal, but includes other forms of computer-readable media such as register memory, processor cache and RAM.
A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/IB2012/057033 | 12/6/2012 | WO | 00 | 6/11/2014 |
| Number | Date | Country | |
|---|---|---|---|
| 61576630 | Dec 2011 | US |
