WO1986006606A1

WO1986006606A1 - Apparatus for non-invasively measuring the volume of urine in a human bladder

Info

Publication number: WO1986006606A1
Application number: PCT/US1986/000969
Authority: WO
Inventors: Gan Guly; David Guiliani
Original assignee: International Biomedics, Inc.
Priority date: 1985-05-08
Filing date: 1986-05-02
Publication date: 1986-11-20
Also published as: EP0221994A1

Abstract

An apparatus for measuring the volume of urine in a human bladder, including a means for transmitting a plurality of ultrasound signals (10, 24) through the bladder and for receiving the returned ultrasound signals (10, 24). The apparatus further includes means for processing the received analog signals (26), means for converting the analog signal into a series of digital data elements and means for processing the data elements to detect selected portions of the border of the bladder, determine the interwall separation and compute the volume of urine in the bladder (12, 34).

Description

APPARATUS FOR NON-INVASIVELY MEASURING THE VOLUME OF URINE IN A HUMAN BLADDER

Technical Field

This invention generally concerns non-in¬ vasive ultrasonic medical apparatus, and more specific* ally concerns such an apparatus which is used to deter¬ mine the volume of urine in the bladder.

Background of the Invention

It is well known that bladder dysfunction is associated with a number of clinical conditions requir¬ ing treatment. It has been estimated that as many as twelve million urinary conditions requiring treatment occur each year in the United States. In many of these cases, it is important that the volume of urine in the bladder be accurately determined, sometimes on a fre¬ quent, if not substantially continuous, basis. This is especially true in cases involving spinal cord injuries which require bladder retraining and in those cases, such as post-operative recovery, where there is a tem¬ porary loss of bladder sensation and/or a loss of the normal voiding mechanism. Knowing the volume of urine in the bladder in such situations helps both bladder management and aids in the prevention of bladder over- distension.

The most common and reliable current tech¬ nique of bladder volume determination is catheteriza* tion. Catheterization is used both as a diagnostic tool, and to actually empty the bladder when necessary. Typically, catheterization with respect to bladder volume is accurate to within approximately ten percent, and is currently the standard against which other meth¬ ods are judged.

However, it is well known that there are sig¬ nificant disadvantages to catheterization. It is in- vasive and cannot be used for continuous monitoring, and.further, it is uncomfortable for the patient. Still further, it is estimated that a significant number of people are seriously affected each year in the United States as a result of infection stemming from cath- eterization. The risk of infection alone is signifi-¹ cant enough to provide a high incentive for a reason¬ able alternative.

Non-invasive procedures for bladder volume estimation are known, but are either unreliable or expensive or have some other significant disadvantage. Palpation and ascultatory percussion are known to be unreliable, while radiography and dye-excretion tech¬ niques are known to be similarly inaccurate and are now regarded to be obsolete at this point. Radio-isotopic procedures have also been used, and while accurate, are complicated and relatively expensive, as well as not being suitable for routine and/or continuous measure¬ ment.

Up to this point the most promising non-in- vasive technique has been conventional ultrasound-based measurement, where the output of the ultrasound appara¬ tus is a two dimensional image of the bladder, from which actual measurements can be made and the volume calculated. However, the equipment for producing this ultrasound image is quite expensive and cumbersome to use. Hence, such methods have remained within the research arena and are not widely used.

Therefore, in summary, the non-invasive al¬ ternatives to catheterization all have significant disadvantages, so that catheterization remains the most commonly used procedure for determining bladder volume. As outlined above, however, it is invasive, traumatic to the patient, and is accompanied by a risk of infec¬ tion.

Disclosure of the Invention

Accordingly, the present invention is an apparatus for measuring the volume of urine in a human bladder, including means for transmitting an ultrasound signal into the bladder, means for receiving the re- turning ultrasound signal, means for determining from the received signal the distance between selected por¬ tions of the border of the bladder, and means for de¬ termining the volume of urine in the bladder from said distance.

Brief Description of the Drawings

Figure 1 is a simplified cross-sectional view showing the location of the bladder in human tissue and the relationship thereto of the ultrasound beam. Figure 2 is a diagram showing the returned ultrasound beam from the patient.

Figure 3 is a diagram showing the signal of Figure 2 following processing thereof by the present invention. Figure 4 is a functional block diagram of the apparatus of the present invention.

Figure 5 is a cross-sectional view showing a linear array embodiment of the transducer portion of the apparatus of Figure 4. Figure 6 is a flow chart of the signal proc¬ essing software portion of the present invention.

Best Mode For Carrying Out The Invention

Although the present invention uses ultra- sound to determine bladder volume, it is unlike a tra¬ ditional ultrasound apparatus which produces a two dimensional visual image of a target or portion of a target, which image is then interpreted by an operator. The currently used but still experimental techniques using conventional ultrasound apparatus produce two dimensional images of the bladder. The diameter of the bladder is then actually measured by the operator off the image on the screen, and the bladder volume is then calculated using that measurement information.

The present invention, however, is substan¬ tially different in concept, in that it uses an ultra¬ sound signal to automatically determine selected portions of the border of the bladder, i.e. the loca¬ tion of the front and back walls of the bladder, rather than to produce a two-dimensional image. The inventors utilized the known fact that a returning ultrasound signal from a tissue region of the human body has a relatively high signal level, while the returning signal from a fluid-filled region, such as a bladder, has a very low level, almost zero, because a fluid filled region is anechoic. The returning signal, after processing, provides a clear representation of the selected portions of the border of the bladder, from which the diameter of the bladder is then deter¬ mined by the processing circuitry. The volume of the bladder is then calculated automatically by the appa¬ ratus and the result displayed as a number, instead of a two-dimensional image.

Referring now to Figure 4, the apparatus of the present invention includes a scanhead element 10. During use of the invention, the scanhead 10 is placed against the abdomen of the patient and transmits and receives the ultrasound signals. The scanhead is a transducer which may comprise one or more transducing elements. In one embodiment, the scanhead comprises a single focused transducer which is mounted at the end of a pen-like handle. Such an element is commercially available. In use of the transducer, a transmission gel is applied to the surface of the transducer and it is placed directly on the patient's abdomen. In another embodiment, a single focused transducer is mounted in a gimbaled structure, which permits freedom of movement along two orthogonal axes. A pair of stepper motors move the transducer through a predetermined path, under computer control. In Figure 4, microprocessor 12 and the accompanying control software 14 would perform the control function for the stepper motors. The front of the transducer in such an embodiment is covered with a convex neoprene dome, which in use is coated with transmission gel and then placed on the patient's abdomen.

For both of the above scanhead embodiments, it is assumed that the bladder to be measured has a substantially spherical configuration. However, in an- other scanhead embodiment, shown in Figure 5 and re¬ ferred to as a sparse linear array, a plurality of focused transducers 16-16 (typically between 5 and 15) are potted on a substrate 18. The individual trans¬ ducers 16 are mounted at predetermined angles and with predetermined spacing relative to each other such that when the individual transducers are activated in a particular known sequence, accurate volume measurements of any bladder shape are obtained.

In the embodiment of Figure 5, a flexible neoprene facing sheet 20 forms the front surface of the transducer, with an acoustic fluid 21 filling the re¬ gion between facing sheet 20 and substrate 18. The transducers 16-16 may also be positioned in a single plane, in which case the transducers are activated sequentially, under computer control, so as to pro¬ vide coverage over a relatively large region.

The transceiver 24 is a conventional combined transmitter and receiver, which is switched between its two operational modes at selected times by micropro- cessor 12. The signal produced by the transceiver in its transmit mode is a typical ultrasound signal, for example 2.25 megahertz pulses in groups or bursts at a pulse repetition frequency of 0.5 kilohertz. The 9

transceiver in the receive mode should have sufficient dynamic range to capture the desired border informa¬ tion. In the present case, a dynamic range of 50db will likely be sufficient. The receiver section in- eludes a variable gain amplifier which can be adjusted to compensate for tissue attenuation of the returning signal.

The received signal is applied to an analog signal processor 26, which detects, rectifies and amp- lifies it. The analog signal processor 26 includes a detector circuit, low and high pass filters and an am¬ plifier. Figure 2 shows a typical signal output of the analog signal processor. The maximum amplitude of the signal for the embodiment shown is approximately 2.5 volts. The signal contains information about the bor¬ ders of the bladder, specified as FW (front wall) and B (back wall) , and the distance therebetween. The remaining portion of the signal is the returning sig¬ nal from the surrounding tissue. The output of the analog signal processor is applied to a standard digitizer circuit 28 and a dis¬ play circuit module 30, such as a liquid crystal dis¬ play or a CRT. The display of the signal from the analog signal procesor is used to give the operator an indication of proper initial positioning of the scanhed on the abdomen of the patient, as will be explained more in detail hereinafter. A switch on the front panel 31 controls whether or not the apparatus is oper¬ ating in this mode. Digitizer 28 is a conventional analog-to-digital converter which in the embodiment shown comprises two high speed six-bit units in cascade to provide a capability of 12 bits of resolution, al¬ though only 7 bits of resolution are actually used in the embodiment shown, i.e. each line of data comprises 256 individual bits of data. It is also possible that the digitizer will include only one six bit unit should such an arrangement provide acceptable results. The resulting digital signal is stored in a buffer memory 32 and from there is transferred to the microprocessor's main memory. Here the data is oper¬ ated on by the signal processing software 34. The signal processing software is responsible for the most significant portion of the processing of the received ultrasound signals. The control software 14, on the other hand, is responsible for the timing of the transceiver modes as well as the sequencing of the various functions of the microprocessor and related modules. The control software is conventional, but the signal processing software is unique in concept.

In operation of the apparatus of the present invention, the scanhead 10 is placed on the abdomen of a patient, with the patient typically being in a su¬ pine position. The transducer is moved around on the patient's abdomen until the the visual display is similar to that shown in Figure 2, which indicates that the transducer is generally over the area of the blad- der 13. This is the initial positioning step for the apparatus. The transducer is then rocked gently by the operator about its initial position, so that the ultrasound signals proceed through the bladder, at various angles from this point. The ultrasound sig- nals, which are substantially straight lines, thus are directed through a substantial portion of the cross- sectional area of the bladder 11.

For the embodiment shown, fiftyseven differ¬ ent individual signal bursts, each referred to as a line of data, are used in order to give an appropriate coverage. However, it should be understood that fewer lines could be used. As the individual lines of data, such as the signal shown in Figure 2, which is an ap¬ propriate analog signal for one line of data, are re- ceived by the instrument, each line of data is digi¬ tized and individually processed in microprocessor 12 by the signal processing system of the present inven¬ tion. The signal processing system comprises a series of key operations which are performed on the data. The first operation is a threshold determinati¬ on. In this step, the level of noise present in the returning signal is estimated. The overall noise level is first estimated by computing the standard deviation of the entire A-line data. Each A-line corresponds to a single transmission burst. This initial noise esti¬ mate is then used as a rough threshold. A refined noise level estimate is then ob¬ tained by calculating the standard deviation for that segment of the signal with amplitudes equal to or below that of the crude threshold. This refined threshold value then represents an estimate of the noise which accompanies the signal returning from the bladder re¬ gion and is used as the final threshold value for the border detection process.

The next operation is noise cleaning or fil¬ tering. Before performing the actual thresholding op- eration, it is necessary to filter out as much of the noise in the returning signal as possible. However, since the desired border information is associated with large spatial derivatives (and hence relatively high spatial frequencies) a more conventional linear smooth- ing filter, with low pass characteristics, would remove valuable information. Hence, a non-linear median fil¬ ter is used. Median filters are characterized by their ability to remove "spikey" noise, while leaving border information intact. In the current embodiment of the invention, the length of the median filter is chosen such that noise spikes with a base of less than 0.5 centimeter in width are removed.

In the next step of thresholding, each ele¬ ment of the filtered signal vector is compared with the refined threshold value established earlier and des¬ cribed above. Each element with original amplitudes which are greater than or equal to the threshold value are set to a preset constant value, while the ampli- tudes of the other elements are set to zero. Thus, the A-line data is now transformed to a binary signal, with the zero amplitude elements representing the fluid filled bladder region. In the next step, the front to back wall separation is estimated. The front wall is determined by locating the first element of the thresholded A-line with an amplitude value of zero. The back wall is determined by locating the first of a set of at least five consecutive non-zero values. The criterion for detecting the actual location of the back wall was developed experimentally to avoid inaccuracies due to reverberation.

A flow chart for the signal processing is shown in Figure 6. Block 40 represents the beginning or initializing of the border detection process. Block 41 represents the next operation in the process, the se¬ lection of the memory bank containing the data set associated with the ith line of data, or more conven- iently, the ith A line. The mean value _mean(i₎ of this data is computed as well as the standard deviation

^Asdev(i) •

In the next operation, represented by block

42, the ith A line is processed to determine a minimum amplitude value A_mi_n(i₎ for the ith line of data. The line of data encompasses the distance from the pa¬ tient's abdomen to the back wall of the patient's blad¬ der.

In the next operation shown in block 44, a rough threshold value T_crucj_{e (}i_j is computed by adding

^Asdev(i) ^{t0 A}min(i)« ^Tne rough threshold value ^τcrude₍i₎ i^s shown in Figure 2 for that particular signal as an example.

The mean value Bmeanfi₎ for all the data points with amplitudes less than T_cru<3_e,.. __s compute_d and then in the next operation, as shown in block 46, the standard deviation

for this set of data points is computed and the minimum amplitude value ^Bmin(i) ^{for a}H ^tne data points with amplitudes less than crude^ ₎ is determined. The standard deviation is the square root of the mean of the squares of the differences of the individual data points relative to the mean value B^^an^₎. As shown in block 48, a re_* fined threshold value Trefnd(i) is then computed by adding Bmin₍i₎ to Bsdev₍i₎. The refined threshold value is also shown in Figure 2 for that line of data. The next step, as shown in block 50, is to filter the data, to remove as much noise as possible without harming the signal itself. As discussed above, a conventional low pass filter is not used, since such a filter would remove valuable information from the signal. The median filter, on the other hand, leaves the true signal information, such as the borders in Figure 2, intact. The operation of a median filter is discussed in detail in a book titled Digital Imaging Processing by W. Pratt, which is hereby incorporated by reference. Basically, a three point median filter proc- esses three consecutive bits of data at a time, begin¬ ning at the start of a data string. There is thus a filter window of three data elements. For each such group, the value of the middle data point is replaced by the median value of the three data points. The filter window is then moved one data element along the data string, so that substantially each data point in the data string is processed three times by the filter. At the end of the filtering process, the noise spikes have been removed, leaving the border information in- tact.

In the next operation, shown in block 52, each data point in the A-line data set from the median filter is compared against the refined threshold value ^τrefn_d(i_{) *} *f ^tne amplitude of a particular data point is greater than the threshold value

that data point is set to a predetermined high value, ^Aconst« ϊf ^fch^e amplitude of the data point is below the threshold value, the bit is set to zero. The first data point which is below the threshold _ref_n(3 ₍i₎ (which is hence set to zero) indicates the presence of the front wall of the bladder. There will then be a string of zero data points which represents the bladder region. When there has been five consecutive data points above the threshold, succeeding data points above the threshold are set to a high. The initial high point represents the back wall. The selection of five consecutive high data points as an indication of the back wall is somewhat arbitrary, but is a reliable indication that the back wall has actually been reached. After all the data points in one line (A line) of data has been processed, the resulting signal is like a binary signal in appearance, with clearly defined borders. Figure 3 is a processed transforma¬ tion of the signal of Figure 2.

The next operation in the processing .of the i^tn line of data is shown in block 54. In this step, the inter-wall distance between the front and back wall of the bladder is computed. This distance is stored in a second bank in the microprocessor memory with an associated software pointer to indicate which A-line this distance corresponds to. The above de¬ scribed process is then repeated for the next A-line data. The process is further repeated until all the A-lines have been processed.

When all the inter-wall distance estimates are completed, the information is then used with an appropriate geometric model and the estimated volume is computed. In one embodiment an ellipsoid of rotation is used as the geometric model. In the particular embodiment which includes a scanning array transducer, it is possible to compute the volume of irregular bladder shapes, again using conventional geometric computation procedures. The resulting volumetric meas¬ urement is then displayed as a numerical amount on the front of the apparatus. Thus, an apparatus has been described and shown which provides an accurate indication of the volume of a bladder, non-invasively and without trauma to the patient. It is convenient for an operator to use, and initial experimental results have shown the results to have an average difference of 11% compared with catheterization derived volume measurements.

Although a preferred embodiment of the inven¬ tion has been disclosed herein for illustration, it should be understood that various changes, modificatio¬ ns and substitutions may be incorporated in such em¬ bodiment without departing from the spirit of the in¬ vention, as defined by the claims which follow:

Claims

13 Claims

1. An apparatus for measuring the volume of urine in a human bladder, comprising: means for transmitting an ultrasound signal into- the bladder; means for receiving the returning ultrasound signal; means for determining from said received signal the distance between selected portions of the border of the bladder; and means for determining the volume of urine in the bladder from said distance.

2. An apparatus of Claim 1, including means for displaying said volume as a number.

3. An apparatus of Claim 1, wherein said selected portions of the border of the bladder are the front and back of the bladder, respectively.

4. An apparatus of Claim 3, including means for transmitting a plurality of ultrasound signals into the bladder and determining the distance between the selected portions of the border of the bladder for each signal.

5. An apparatus of Claim 4, wherein said plurality of signals are transmitted substantially through the bladder.

6. An apparatus of Claim 5, including means for converting the received signals into a series of data elements, means for establishing a threshold amplitude value, means for comparing each data element against said threshold value and means for then setting the amplitude for each said data element at a high 14

value if the data element value is above said threshold value and to a low value if it is below said threshold value.

7. An apparatus of Claim 6, including means for removing noise from the received signals without degrading the border information thereof.

8. An apparatus of Claim 6, including means for determining the average distance between the se¬ lected portions of the border of the bladder for all the received signals and for determining the volume of urine in the bladder from said average distance value.

9. A method for measuring the volume of urine in a human bladder, comprising the steps of: transmitting an ultrasound signal into the bladder; receiving the returning ultrasound signal; determining from said received signal the distance between selected portions of the border of the bladder; and determining the volume of urine in the blad¬ der from the distance between the selected portions of the border thereof.

10. A method of Claim 9, including the step of displaying said volume of urine as a number.

11. A method of Claim 9, wherein the se¬ lected portions are the front and back portions of the border of the bladder, respectively.

12. A method of Claim 11, wherein the step of transmitting includes transmitting a plurality of ultrasound signals into the bladder and the step of determining the distance between the selected portions of the border of the bladder, includes the step of determining the distance between the selected portions of the border of the bladder for each said ultrasound signal.

13. A method of Claim 12, wherein the step of determining the selected portions of the border includes the step of converting the received signals into a series of data elements, establishing a thresh* old value, comparing each data element against said threshold value, and setting the amplitude of each said data element at a high value if the data element is above the threshold value and to a low value if it is below the threshold value.

14. A method of Claim 13, including the step of removing noise from the received signals without degrading the border information thereof.

15. A method of Claim 13, including the step of determining the average distance between the se¬ lected portions of the border for all the received signals and for determining the volume of urine in the bladder from said average distance value.