The urine flow estimation device and non-temporary storage media can be read by a computer storing the urine flow estimation program.

Abstract

The purpose of the invention is to improve the accuracy of estimating urine flow. Specifically, the invention relates to a urine flow estimator (70) comprising a camera (71) configured to capture images of urine being discharged, an image analyzer (72) configured to estimate the first urine flow based on the images captured by the camera (71), a microphone (73) configured to capture sounds related to urine discharge, an sound analyzer (74) configured to estimate the second urine flow based on the sounds captured by the microphone (73), and a urine flow estimator (76) configured to estimate the third urine flow based on the first and second urine flows.

Description

Urine flow rate estimation device and urine flow rate estimation program

[0001] The present invention relates to a urine flow estimation device and a urine flow estimation program.

[0002] Uroflow measurement is essential for the diagnosis, treatment, and follow-up of urinary disorders caused by conditions such as benign prostatic hyperplasia, overactive bladder, and neurogenic bladder. Medical institutions typically use a device with a cup placed under the toilet to record urine volume and flow. However, because the environment is different from the toilet they use every day, patients often become nervous or are unable to time their urination properly, making measurements difficult. While it would be ideal to perform measurements multiple times under normal circumstances, performing these methods at home is not practical due to cost and hygiene concerns.

[0003] To overcome these problems, a method has been disclosed in which images taken during urination are analyzed to estimate the urine flow rate (Patent Document 1).

[0004] JP 2018-109285 A

[0005] However, the technique disclosed in Patent Document 1 estimates the urine flow rate only by analyzing images taken during urination, and therefore there is room for improvement in order to improve the accuracy of urine flow rate estimation.

[0006] Therefore, an object of one embodiment of the present invention is to improve the accuracy of estimating urine flow rate.

[0007] As one embodiment, the present application discloses a urine flow estimation device including: a camera configured to acquire an image of excreted urine; an image analysis unit configured to estimate a first urine flow rate based on the image acquired by the camera; a microphone configured to acquire sounds associated with urine excretion; a sound analysis unit configured to estimate a second urine flow rate based on the sounds acquired by the microphone; and a urine flow estimation unit configured to estimate a third urine flow rate based on the first urine flow rate and the second urine flow rate.

[0008] Furthermore, as one embodiment, the present application discloses a urine flow rate estimation program for causing a processor to execute the steps of estimating a first urine flow rate based on an image of excreted urine captured by a camera, estimating a second urine flow rate based on a sound accompanying urine excretion captured by a microphone, and estimating a third urine flow rate based on the first urine flow rate and the second urine flow rate.

[0009] Furthermore, as one embodiment, the present application discloses a urine flow estimation device, wherein the image analysis unit (the step of estimating the first urine flow rate) is configured to estimate the first urine flow rate based on a physical quantity correlated with an axis switching position of the discharged urine in an image of the discharged urine acquired by the camera.

[0010] Furthermore, as one embodiment, the present application discloses a urine flow rate estimation device, wherein the physical quantity is a wavelength that is a distance between a position of maximum amplitude forward of an axis switching position of the discharged urine and the urine outlet in an image of the discharged urine acquired by the camera.

[0011] Furthermore, as one embodiment, the present application discloses a urine flow estimation device, wherein the sound analysis unit (the step of estimating the second urine flow rate) is configured to estimate the second urine flow rate based on sound energy acquired by the microphone.

[0012] Furthermore, as one embodiment, the present application discloses a urine flow estimation device, wherein the urine flow estimation unit (the step of estimating the third urine flow) is configured to estimate the third urine flow using a missing data estimation method based on eigenvalue decomposition of a matrix whose components are the first urine flow, the second urine flow, and the third urine flow.

[0013] Moreover, as one embodiment, the present application discloses a urine flow estimation device, wherein the urine flow estimation unit (the step of estimating the third urine flow) is configured to estimate, with respect to time series data of the first urine flow and the second urine flow, the second urine flow as the third urine flow for a region where the first urine flow is equal to or less than a first threshold, estimate the first urine flow as the third urine flow for a region where the first urine flow is greater than the first threshold and a frequency of the first urine flow is greater than a second threshold, and estimate a urine flow based on the first urine flow and the second urine flow as the third urine flow for a region where the first urine flow is greater than the first threshold and a frequency of the first urine flow is equal to or less than the second threshold.

[0014] Furthermore, as one embodiment, the present application discloses a non-transitory computer-readable storage medium that stores a urine flow estimation program that causes a processor to execute the steps of estimating a first urine flow rate based on an image of excreted urine captured by a camera, estimating a second urine flow rate based on a sound accompanying urine excretion captured by a microphone, and estimating a third urine flow rate based on the first urine flow rate and the second urine flow rate.

[0015] According to one embodiment of the present invention, it is possible to improve the accuracy of estimating urine flow rate.

[0016] FIG. 1 is a diagram schematically showing an experimental apparatus including the urine flow estimation device of this embodiment. FIG. 2 is a diagram schematically showing functional blocks of the urine flow estimation device. FIG. 3 is a diagram schematically showing the shape of a nozzle for discharging urine. FIG. 4A is a diagram schematically showing a background image when urine discharged from a nozzle is captured from the side. FIG. 4B is a diagram schematically showing an image of urine discharged from a nozzle captured from the side. FIG. 5 is a diagram schematically showing the correlation between wavelength change and urine flow rate. FIG. 6A is a diagram showing an example of image processing for suppressing erroneous estimation due to splashing water. FIG. 6B is a diagram showing an example of image processing for suppressing erroneous estimation due to splashing water. FIG. 7A is a diagram schematically showing sound energy density for each frequency over time. FIG. 7B is a diagram schematically showing sound energy over time. FIG. 7C is a diagram schematically showing smoothed sound energy over time. FIG. 8 is a diagram showing a urine flow estimation result (No. 1). FIG. 9 is a diagram showing a urine flow estimation result (No. 2). FIG. 10 is a diagram showing the urine flow estimation result (No. 3). FIG. 11 is a diagram showing the urine flow estimation result (No. 4). FIG. 12 is a diagram showing the urine flow estimation result (No. 5). FIG. 13 is a diagram showing errors in the estimation results (No. 1 to 5) by various estimation methods. FIG. 14 is a diagram showing an example of urine flow estimation by the urine flow estimator. FIG. 15 is a flow diagram for explaining the urine flow estimation program of this embodiment.

[0017] The urine flow estimation device and urine flow estimation program of this embodiment will be described below with reference to the drawings.

[0018] FIG. 1 is a schematic diagram of an experimental apparatus including a urine flow rate estimation device according to this embodiment. The experimental apparatus 100 is configured to compress a liquid simulating urine (hereinafter simply referred to as urine) stored in a tank 20 using a compressor 10, and discharge the liquid from a nozzle 60 via a Venturi tube 30. A differential pressure converter 40 is connected to the Venturi tube 30. The differential pressure converter 40 calculates the urine flow rate based on the difference in pressure between two different points on the Venturi tube 30 using Bernoulli's theorem. The calculated urine flow rate is displayed on a monitor 50. Note that in this specification, urine flow rate refers to the urine flow rate per unit time (ml / s).

[0019] The urine discharged from the nozzle 60 is collected in a container 66 and then returned to the tank 20 by a pump 68. A light source 64 is disposed near the nozzle 60 so that a clear image of the urine being discharged from the nozzle 60 can be obtained.

[0020] In this embodiment, a smartphone that can be held by the patient is used as the urine flow estimation device 70. However, the invention is not limited to this. The urine flow estimation device 70 can be any device having the functions described below.

[0021] Fig. 2 is a diagram showing the schematic functional blocks of a urine flow estimation device 70. As shown in Fig. 2, a urine flow estimation device 70 includes a camera 71 configured to capture images of excreted urine, a microphone 73 configured to capture sounds accompanying urine excretion, a memory 75 that stores a urine flow estimation program, and a processor 77 that can read and execute the urine flow estimation program stored in the memory 75.

[0022] The urine flow estimation device 70 includes, as functional blocks realized by the processor 77 executing a urine flow estimation program, an image analysis unit 72 configured to estimate a first urine flow based on an image acquired by a camera 71, a sound analysis unit 74 configured to estimate a second urine flow based on a sound acquired by a microphone 73, and a urine flow estimation unit 76 configured to estimate a third urine flow based on the first urine flow and the second urine flow.

[0023] The urine flow estimation program can be downloaded to the memory of a device such as a smartphone via communication such as the Internet and executed, or can be stored in a non-transitory storage medium that can be read by a processor and sold, etc.

[0024] The camera 71 may be any camera that can acquire an image with sufficient accuracy to recognize the axis switching position of urine when the image of excreted urine is acquired from the side. In this embodiment, an example is shown in which the image of excreted urine is acquired from the side, but the invention is not limited to this, and the image of excreted urine can be acquired from any direction. The camera 71 may be any camera that can acquire an image with sufficient accuracy to recognize the axis switching position of urine when the image of excreted urine is acquired from any direction.

[0025] That is, this embodiment applies the axis switching phenomenon, which is one of the interfacial instability phenomena of fluids. Axis switching is a phenomenon in which, when a liquid flows out of a nozzle whose cross section is elliptical rather than circular, the major and minor axes of the elliptical cross section of the liquid column immediately after the liquid flows out are reversed, resulting in the observation of a jet shaped like a twisted rod (see Rayleigh, L., "On the Capillary Phenomena of Jets," Proc. R. Soc. London, 21, 71-97, (1879)). Since the tip of the urethra is generally elliptical, axis switching also occurs in urinating urine.

[0026] Fig. 3 is a diagram showing the outline of the shape of the nozzle for discharging urine. As shown in Fig. 3, the outlet 62 of the nozzle 60 is a vertically long ellipse that mimics the shape of the tip of the urethra. The ratio of the minor axis (Dmin) to the major axis (Dmax) of the elliptical shape of the outlet 62 is 1:6.

[0027] The image analysis unit 72 is configured to estimate the first urine flow rate based on a physical quantity correlated with the axis switching position of the excreted urine in the image acquired by the camera 71 .

[0028] Specifically, in this embodiment, the image analysis unit 72 uses, as a physical quantity correlated with the axis switching position of the excreted urine, a wavelength that is the distance between the urine outlet and the maximum amplitude position ahead of the axis switching position of the excreted urine in the image acquired by the camera 71. This point will be explained below.

[0029] FIG. 4A is a diagram schematically illustrating a background image when urine discharged from a nozzle is captured from the side. FIG. 4B is a diagram schematically illustrating an image of urine discharged from a nozzle captured from the side. As shown in FIG. 4B, the image analysis unit 72 uses an existing image analysis method to binarize the image captured by the camera 71, extract the contour of the urine, and determine the axis switching position AX (the position of minimum urine amplitude). The image analysis unit 72 also determines the position MX of maximum urine amplitude ahead of the axis switching position AX. The image analysis unit 72 determines the wavelength L, which is the distance between the urine outlet (nozzle outlet 62) and the maximum amplitude position MX.

[0030] The image analysis unit 72 estimates the first urine flow rate based on the wavelength L in the image acquired by the camera 71. Figure 5 is a diagram schematically showing the correlation between wavelength change and urine flow rate. In Figure 5, the horizontal axis represents time (s), the vertical axis (left) represents wavelength (mm), and the vertical axis (right) represents urine flow rate (ml / s). In Figure 5, wavelengths are plotted as white circles, and urine flow rates are plotted as black circles. As shown in Figure 5, there is a correlation between wavelength and urine flow rate.

[0031] The image analysis unit 72 can estimate the first urine flow rate based on the wavelength L in the image acquired by the camera 71 by previously determining a conversion formula from wavelength L (mm) to urine flow rate (ml / s) through experiments, simulations, or the like. Note that, in this embodiment, an example has been shown in which the image analysis unit 72 estimates the first urine flow rate based on wavelength L, but this is not limiting, and the first urine flow rate can also be estimated based on the distance between the urine outlet (nozzle outlet 62) and the axis switching position AX.

[0032] The image analysis unit 72 can set an image analysis region to prevent erroneous estimation due to splashing of liquid caused by excreted urine. Figures 6A and 6B are diagrams showing an example of image processing for preventing erroneous estimation due to splashing of water.

[0033] When urine discharged from nozzle 60 hits container 66 (a toilet bowl in actual use), the discharged urine itself may bounce off container 66, or liquid such as water or urine stored in container 66 may bounce back, which may appear as splashes SP in the image captured by camera 71, as shown in Figures 6A and 6B. In this case, image analysis unit 72 may erroneously detect the splashes SP as the maximum amplitude position MX, which may result in an erroneous estimation of the first urine flow rate.

[0034] Therefore, the image analysis unit 72 can set the area through which the excreted urine passes as the image analysis area AN out of the entire area of ​​the image acquired by the camera 71. In this case, the image analysis unit 72 can prevent erroneous estimation due to splashing water by determining the maximum amplitude position MX and wavelength L based on image analysis within the image analysis area AN.

[0035] The microphone 73 may be configured to capture sounds associated with urine excretion (for example, the sound of urine hitting the container 66). The sound analysis unit 74 is configured to estimate the second urine flow rate based on the energy of the sound captured by the microphone 73. That is, there is a correlation between the energy of the sound (loudness) associated with excreted urine and the urine flow rate. Therefore, the sound analysis unit 74 can estimate the second urine flow rate based on the energy of the sound captured by the microphone 73 using an existing sound analysis method.

[0036] FIG. 7A is a diagram that schematically illustrates the sound energy density for each frequency over time. In FIG. 7A, the horizontal axis represents time (s) and the vertical axis represents frequency (Hz). FIG. 7A illustrates the power spectral density for each frequency of the sound caused by urine using a gray scale. The sound analysis unit 74 can use an existing sound analysis method to determine the power spectral density for each frequency over time, as shown in FIG. 7A. Note that the sound analysis unit 74 can also use an existing sound analysis method to remove environmental noise, such as the sound of a toilet fan.

[0037] Figure 7B is a diagram that schematically shows sound energy over time. In Figure 7B, the horizontal axis represents time (s) and the vertical axis represents sound energy. The sound analysis unit 74 can use an existing sound analysis method to calculate the total sound energy over time (total energy of all frequencies), as shown in Figure 7B.

[0038] Figure 7C is a diagram schematically showing smoothed sound energy over time. In Figure 7C, the horizontal axis represents time (s) and the vertical axis represents sound energy. The sound analysis unit 74 can obtain the smoothed sound energy over time as shown in Figure 7C by smoothing the graph of Figure 7B by, for example, averaging data from multiple points (e.g., 150 points) before and after the measurement point. The sound analysis unit 74 can estimate the second urine flow rate based on the sound energy acquired by the microphone 73 by previously determining a conversion formula from sound energy to urine flow rate (ml / s) through experiments or simulations.

[0039] The urine flow estimating unit 76 is configured to estimate a third urine flow based on the first urine flow and the second urine flow. Specifically, the urine flow estimating unit 76 is configured to estimate the third urine flow using a missing data estimation method based on eigenvalue decomposition of a matrix whose components are the first urine flow, the second urine flow, and the third urine flow. In this embodiment, Gappy POD is used as an example of a missing data estimation method based on eigenvalue decomposition. Estimation of the third urine flow using Gappy POD will be described below.

[0040] Sparse singular value decomposition (Gappy POD) is a type of matrix singular value decomposition. Singular value decomposition decomposes a given matrix into three matrices, each of which can represent the correlation and strength of row and column elements. First, complete training data is constructed, and then singular value decomposition is performed. The eigenvectors obtained by singular value decomposition represent the correlation between measured parameters (sound and wavelength) and urinary flow. This correlation vector can be used to estimate urinary flow (third urinary flow). In particular, sparse singular value decomposition (Gappy POD) can easily determine the correlation between multiple parameters and estimate urinary flow even if not all data is measured. Specifically, in locations where it is difficult to capture images of excreted urine or when environmental noise is loud and the sound associated with urinary excretion cannot be captured, urinary flow can be estimated using only one of the data, even if complete data (sound and image) cannot be obtained.

[0041] Figures 8 to 12 show the urine flow rate estimation results (No. 1 to 5). Figures 8 to 12 show the urine flow rate estimation results for five different urine excretion patterns. In the graphs on the left side of Figures 8 to 12, the horizontal axis is time (s), and the vertical axis is the normalized first urine flow rate (solid line) and the second urine flow rate (dashed line). In the graphs on the right side of Figures 8 to 12, the horizontal axis is time (s), and the vertical axis is the third urine flow rate (ml / s, solid line) and the urine flow rate measured using the differential pressure transducer 40 (ml / s, dashed line).

[0042] Figure 13 shows the error of the estimation results (No. 1 to 5) using various estimation methods. Figure 13 shows the root mean square error (RMSE) of the estimation results for urine flow rate estimation based on Gappy POD, urine flow rate estimation based on SVD (singular value decomposition), urine flow rate estimation based only on sound analysis, and urine flow rate estimation based only on image analysis (wavelength L). The smaller the RMSE, the smaller the error. As shown in Figure 13, urine flow rate estimation based on Gappy POD improves the accuracy of urine flow rate estimation compared to other estimation methods.

[0043] According to this embodiment, the accuracy of urine flow estimation can be improved by not only estimating urine flow rate by image analysis during urination but also by analyzing sound during urination and estimating urine flow rate based on the results of both estimations. Furthermore, according to this embodiment, urine flow rate is estimated using a missing data estimation method based on eigenvalue decomposition, so that urine flow rate can be estimated accurately even if there is a missing data in the image data or sound data.

[0044] In the above embodiment, the urine flow estimator 76 estimates the third urine flow using a missing data estimation method based on eigenvalue decomposition, but the present invention is not limited to this. Fig. 14 is a diagram showing an example of urine flow estimation by the urine flow estimator.

[0045] The urine flow estimating unit 76 can estimate the second urine flow as the third urine flow for a region (regions t1 and t5) in which the first urine flow is equal to or less than a first threshold (e.g., 0.2 ml / s) in the time series data of the first and second urine flow rates. Also, the urine flow estimating unit 76 can estimate the first urine flow as the third urine flow for a region (region t3) in which the first urine flow is greater than the first threshold and the frequency of the first urine flow is greater than a second threshold in the time series data of the first and second urine flow rates.

[0046] Furthermore, for the time series data of the first urine flow rate and the second urine flow rate, the urine flow rate estimation unit 76 can estimate a urine flow rate based on the first urine flow rate and the second urine flow rate as the third urine flow rate for a region (regions t2 and t4) where the first urine flow rate is greater than the first threshold value and the frequency of the first urine flow rate is equal to or less than the second threshold value.

[0047] For example, for the regions t2 and t4, the urine flow estimating unit 76 may estimate the average value of the first and second urine flow rates as the third urine flow rate.Furthermore, for the regions t2 and t4, the urine flow estimating unit 76 may estimate the third urine flow rate using a missing data estimation method based on eigenvalue decomposition of a matrix whose components are the first urine flow rate, the second urine flow rate, and the third urine flow rate.

[0048] According to this embodiment, it is possible to perform highly accurate urine flow estimation by taking into account the characteristics of both image-based and sound-based urine flow estimation. That is, the inventors of the present application have found that urine flow estimation based on image analysis cannot be performed when the flow rate of excreted urine is low (the estimated result is almost zero even though there is an actual flow rate), as shown in the regions t1 and t5 of Fig. 14. On the other hand, the inventors of the present application have found that urine flow estimation based on sound analysis can estimate the total flow rate, but the estimation accuracy is poor when there are small flow rate fluctuations (when the flow rate frequency is high), as shown in the region t3 of Fig. 14.

[0049] Therefore, for the region where the first urine flow rate is low (regions t1 and t5), the urine flow rate estimation unit 76 adopts the urine flow rate based on sound analysis (second urine flow rate) as the third urine flow rate because the reliability of the urine flow rate based on image analysis (first urine flow rate) is low.Furthermore, for the region where the first urine flow rate has small flow rate fluctuations (region t3), the urine flow rate estimation unit 76 adopts the urine flow rate based on image analysis (first urine flow rate) as the third urine flow rate because the reliability of the urine flow rate based on sound analysis (second urine flow rate) is low.

[0050] Furthermore, for regions that do not fall into any of the above categories (regions t2 and t4), the urine flow estimating unit 76 estimates the third urine flow based on both the urine flow based on image analysis (first urine flow) and the urine flow based on sound analysis (second urine flow), since both are highly reliable. As a result, the urine flow estimating unit 76 can improve the accuracy of urine flow estimation by estimating the urine flow taking into account the characteristics of both the urine flow estimation based on image analysis and the urine flow estimation based on sound analysis.

[0051] Next, the urine flow rate estimation program of this embodiment will be described with reference to Fig. 15, which is a flow chart for explaining the urine flow rate estimation program of this embodiment.

[0052] As shown in FIG. 15, the urine flow estimation program causes a processor (processor 77) to execute step 102 of estimating a first urine flow rate based on an image of excreted urine acquired by a camera (e.g., camera 71).

[0053] Step 102 of estimating the first urine flow rate is configured to estimate the first urine flow rate based on a physical quantity correlated with the axis-switching position of the excreted urine in an image of the excreted urine acquired from the side by a camera, similar to the processing by the image analysis unit 72. The physical quantity may be a wavelength that is the distance between the urine outlet and a maximum amplitude position anterior to the axis-switching position of the excreted urine in an image of the excreted urine acquired from the side by a camera.

[0054] In parallel with step 102, the urine flow rate estimation program also causes the processor (processor 77) to execute step 104 of estimating a second urine flow rate based on sounds associated with urine excretion captured by a microphone (e.g., microphone 73).

[0055] The step 104 of estimating the second urine flow rate is configured to estimate the second urine flow rate based on the energy of the sound acquired by the microphone, similar to the processing by the sound analysis unit 74 described above.

[0056] The urine flow rate estimation program also causes the processor (processor 77) to execute step 106 of estimating a third urine flow rate based on the first urine flow rate estimated in step 102 and the second urine flow rate estimated in step 104.

[0057] Step 106 of estimating the third urine flow rate is configured to estimate the third urine flow rate using a missing data estimation technique based on eigenvalue decomposition of a matrix whose components are the first urine flow rate, the second urine flow rate, and the third urine flow rate, similar to the processing by the urine flow rate estimation unit 76 described above.

[0058] Without being limited to this, in step 106 of estimating the third urine flow, for the time series data of the first urine flow and the second urine flow, the second urine flow can be estimated as the third urine flow in a region where the first urine flow is equal to or less than a first threshold, the first urine flow can be estimated as the third urine flow in a region where the first urine flow is greater than the first threshold and the frequency of the first urine flow is greater than a second threshold, and the urine flow based on the first urine flow and the second urine flow can be estimated as the third urine flow in a region where the first urine flow is greater than the first threshold and the frequency of the first urine flow is equal to or less than the second threshold.

[0059] According to the urine flow estimation program of this embodiment, like the urine flow estimation device 70 described above, it is possible to improve the accuracy of urine flow estimation.

[0060] 60 Nozzle 62 Outlet 70 Urine flow rate estimation device 71 Camera 72 Image analysis unit 73 Microphone 74 Sound analysis unit 75 Memory 76 Urine flow rate estimation unit 77 Processor AX Axis switching position L Wavelength MX Maximum amplitude position

Claims

1. A urine flow estimation device comprising: a camera configured to capture an image of excreted urine; an image analysis unit configured to estimate a first urine flow rate based on the image captured by the camera; a microphone configured to capture sounds associated with urine excretion; a sound analysis unit configured to estimate a second urine flow rate based on the sounds captured by the microphone; and a urine flow estimation unit configured to estimate a third urine flow rate based on the first urine flow rate and the second urine flow rate.

2. The urine flow estimation device according to claim 1, wherein said image analysis unit is configured to estimate said first urine flow rate based on a physical quantity correlated with an axis switching position of said discharged urine in an image of said discharged urine acquired by said camera.

3. The urine flow rate estimation device according to claim 2, wherein the physical quantity is a wavelength that is a distance between the urine outlet and a maximum amplitude position ahead of an axis switching position of the discharged urine in an image of the discharged urine acquired by the camera.

4. The urine flow estimation device according to claim 3, wherein the sound analysis unit is configured to estimate the second urine flow rate based on sound energy acquired by the microphone.

5. The urine flow estimation device according to any one of claims 1 to 4, wherein said urine flow estimation unit is configured to estimate said third urine flow using a missing data estimation method based on eigenvalue decomposition of a matrix whose components are said first urine flow, said second urine flow, and said third urine flow.

6. The urine flow estimating device according to any one of claims 1 to 4, wherein said urine flow estimating unit is configured to: estimate, with respect to time series data of said first urine flow and said second urine flow as said third urine flow for a region where said first urine flow is equal to or less than a first threshold; estimate said first urine flow as said third urine flow for a region where said first urine flow is greater than said first threshold and the frequency of said first urine flow is greater than a second threshold; and estimate a urine flow based on said first urine flow and said second urine flow as said third urine flow for a region where said first urine flow is greater than said first threshold and the frequency of said first urine flow is equal to or less than said second threshold.

7. A urinary flow estimation program for causing a processor to execute the steps of: estimating a first urinary flow rate based on an image of excreted urine captured by a camera; estimating a second urinary flow rate based on a sound accompanying the excretion of urine captured by a microphone; and estimating a third urinary flow rate based on the first urinary flow rate and the second urinary flow rate.

Images