Blood flow rate correlation index estimation device, physical condition abnormality detection system, blood flow rate correlation index estimation method, and blood flow rate correlation index estimation program

Abstract

This blood flow rate correlation index estimation device is provided with: an imaging data acquisition unit (111) that acquires imaging data obtained by imaging a first region and a second region of a human skin; a luminance data calculation unit (113) that acquires, from the imaging data, first luminance data of the first region and second luminance data of the second region; a blood flow rate correlation index calculation unit (121) that calculates a first blood flow rate correlation index correlated with the blood flow rate from the first luminance data and calculates a second blood flow rate correlation index correlated with the blood flow rate from the second luminance data; and a region selection unit (122) that, when one of the first luminance data and the second luminance data includes an invalid value or noise, selects the other luminance data and outputs a selection signal indicating the selected luminance data, or that, when one of the first blood flow rate correlation index and the second blood flow rate correlation index includes an invalid value or noise, selects the other blood flow rate correlation index.

Description

Blood flow correlation index estimation device, abnormal physical condition detection system, blood flow correlation index estimation method, and blood flow correlation index estimation program

[0001] The present disclosure relates to a blood flow correlation index estimation technique for estimating a blood flow correlation index that correlates with blood flow.

[0002] Patent Literature 1 discloses a technique for calculating a blood pressure value from the phase difference between pulse waves acquired from multiple skin regions of a person, such as the hand and face, using a camera. More specifically, the blood pressure value is calculated from the phase difference between the pulse wave acquired from the hand and the pulse wave acquired from the face.

[0003] Japanese Patent Application Laid-Open No. 2022-157247

[0004] According to the technology of Patent Document 1, blood pressure values ​​are calculated based on data acquired from multiple skin areas such as the hands and face, so there is a problem in that blood pressure values ​​cannot be calculated if data cannot be acquired from a certain skin area, for example, when a person changes the direction of their face.

[0005] The present disclosure was made in response to the recognition of such problems, and aims to provide a blood flow correlation index estimation technology that can estimate the value of a desired bioindicator even when significant data cannot be obtained from the skin area to estimate the value of the desired bioindicator.

[0006] One aspect of a blood flow correlation index estimation device according to an embodiment of the present disclosure includes: an imaging data acquisition unit that acquires imaging data obtained by imaging a first region and a second region of human skin; a brightness data calculation unit that acquires first brightness data of the first region and second brightness data of the second region from the imaging data; a blood flow correlation index calculation unit that calculates a first blood flow correlation index that correlates with blood flow from the first brightness data and calculates a second blood flow correlation index that correlates with blood flow from the second brightness data; and a region selection unit that, when either the first brightness data or the second brightness data contains an invalid value or noise, selects the other brightness data and outputs a selection signal indicating the selected brightness data, or, when either the first blood flow correlation index or the second blood flow correlation index contains an invalid value or noise, selects the other blood flow correlation index.

[0007] According to the blood flow correlation index estimation device of an embodiment of the present disclosure, even if significant data cannot be obtained from a certain skin area to estimate the value of the blood flow correlation index as a desired bioindicator, the value of the blood flow correlation index can still be estimated.

[0008] FIG. 3B is a diagram showing an example of the configuration of a blood flow correlation index estimation device and an abnormal physical condition detection system according to embodiment 1. FIG. 3A is a diagram showing a method of calculating pulse wave amplitude from brightness data. FIG. 3B is a diagram showing the operation of a region selection unit. FIG. 3A shows an example in which a good pulse wave amplitude value is calculated for region X, but an invalid value is included in the pulse wave amplitude value for region Y. FIG. 3B shows an example in which a good pulse wave amplitude value is calculated for region Y, but an invalid value is included in the pulse wave amplitude value for region X. FIG. 4A and FIG. 4B are diagrams showing an example of correction when there is a difference in pulse wave amplitude values, which are relative values, between region X and region Y. FIG. 4A shows a case in which there is a difference in pulse wave amplitude values ​​between region X and region Y. FIG. 4B shows an example of correcting the difference in pulse wave amplitude values ​​between region X and region Y by standardization. FIG. 3B is a diagram showing an example of the hardware configuration of a blood flow correlation index estimation device or an abnormal physical condition detection system. FIG. 3C is a diagram showing an example of the hardware configuration of a blood flow correlation index estimation device or an abnormal physical condition detection system. FIG. 3D is a flowchart showing an example of the operation of a blood flow correlation index estimation device or an abnormal physical condition detection system. 10 is a flowchart showing an example of the operation of a blood flow correlation index estimation device or an abnormal physical condition detection system according to a second embodiment.

[0009] Various embodiments of the present disclosure will be described in detail below with reference to the drawings. In the drawings, identical or similar parts are designated by identical or similar reference numerals, and redundant explanations of such parts will be omitted. In addition, in this disclosure, the term "or" is used to mean an inclusive logical OR unless otherwise specified.

[0010] Embodiment 1. <Configuration> A blood flow correlation index estimation device and an abnormal physical condition detection system including the blood flow correlation index estimation device according to embodiment 1 of the present disclosure will be described with reference to Fig. 1. As shown in Fig. 1, the abnormal physical condition detection system 50 includes a blood flow correlation index estimation device 10 and an abnormal physical condition determination unit 20.

[0011] The blood flow correlation index estimation device 10 is a device for estimating the value of a blood flow correlation index that correlates with blood flow as a biomarker value from a time series of images of a person. The abnormal physical condition detection system 50 is a system for determining whether a person's physical condition is abnormal based on the results estimated by the blood flow correlation index estimation device 10.

[0012] In order to achieve this object, one aspect of the blood flow correlation index estimation device 10 includes an imaging data acquisition unit 111 that acquires imaging data obtained by imaging a first region and a second region of human skin, a brightness data calculation unit 113 that acquires first brightness data of the first region and second brightness data of the second region from the imaging data, a blood flow correlation index calculation unit 121 that calculates a first blood flow correlation index that correlates with blood flow from the first brightness data and calculates a second blood flow correlation index that correlates with blood flow from the second brightness data, and a region selection unit 122 that selects either the first brightness data of the first region or the second brightness data of the second region as data to be used by the blood flow correlation index calculation unit for the calculation based on the first brightness data, the second brightness data, or the blood flow correlation index.

[0013] Another aspect of the blood flow correlation index estimation device 10, in addition to the above-mentioned aspect, further includes a correction unit 123 that performs correction between the first blood flow correlation index and the second blood flow correlation index when selecting either the first blood flow correlation index or the second blood flow correlation index.

[0014] Another aspect of the blood flow correlation index estimation device 10 includes a data acquisition unit 11 and an index calculation unit 12. The data acquisition unit 11 includes an imaging data acquisition unit 111, a skin region detection unit 112, and a brightness data calculation unit 113. The index calculation unit 12 includes, for example, a blood flow correlation index calculation unit 121, a region selection unit 122, and a correction unit 123. The index calculation unit 12 may include a biological information calculation unit 124 or a vehicle information acquisition unit 125.

[0015] These functional units are connected to one or more other functional units so that information can be transmitted from the base of the arrow to the tip of the arrow, as shown by the arrows in Figure 1. Each functional unit will be described in more detail below.

[0016] (Imaging Data Acquisition Unit) The imaging data acquisition unit 111 acquires, as an example, time-series imaging data of images captured by a camera (not shown). The camera (not shown) is placed, for example, in the interior or cockpit of a mobile body such as an automobile, ship, or aircraft, or in the interior of a facility such as a nursing home or medical institution, and acquires imaging data including the skin (e.g., face) of a subject to be evaluated. Examples of subjects to be evaluated include occupants of the mobile body, including operators such as drivers, people receiving care, or patients.

[0017] The camera (not shown) captures images at a frame rate of, for example, 30 fps (frames per second). The captured images may be RGB images (color images) or IR images (monochrome images). In the case of IR images, the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 is equipped with an IR light source (not shown) that emits IR light to acquire the IR images from the subject. The captured image data may be images acquired in real time or images stored in a buffer (not shown). When the blood flow correlation index estimation device 10 is used as a device constituting a driver monitoring system that monitors the driver of a vehicle, it is possible to acquire the captured image data in real time. When the blood flow correlation index estimation device 10 is used as a device constituting a patient diagnosis support system, it is possible to acquire the captured image data from a buffer.

[0018] The imaging data acquisition unit 111 supplies the acquired imaging data to the skin region detection unit 112 .

[0019] (Skin Area Detection Unit) The skin area detection unit 112 detects the skin area of ​​the person being evaluated (hereinafter, may be simply referred to as "skin area") using a known image processing technique from the imaging data supplied from the imaging data acquisition unit 111. In the following explanation, the skin area will be specifically described as a face area, but the skin area may also be an area including skin other than a face area.

[0020] For example, the skin region detection unit 112 detects the face of the person to be evaluated according to a known technique, and detects multiple skin regions within the face region based on facial feature amounts. Examples of skin regions included within the face region include the forehead region, cheek region, and chin region.

[0021] Furthermore, the skin region detection unit 112 may subdivide each skin region. For example, the forehead region may be subdivided into 10 regions, and each of the left and right cheek regions may be subdivided into 5 regions. For ease of explanation, when a skin region is subdivided into multiple regions, each of the subdivided skin regions is referred to as a sub-skin region.

[0022] (Brightness Data Calculation Unit) The brightness data calculation unit 113 calculates brightness data indicating the brightness (brightness change) of each skin region detected by the skin region detection unit 112. The brightness data is acquired as data indicating brightness change like a waveform, for example, as shown in Fig. 2. Since the brightness of the skin region in the captured image changes due to changes in blood flow, the change in brightness data over time is observed as a waveform, as shown in Fig. 2.

[0023] As one example, the luminance data calculation unit 113 may calculate luminance data indicating one brightness change for one forehead region. As another example, the luminance data calculation unit 113 may calculate luminance data indicating one brightness change for one cheek region. As another example, the luminance data calculation unit 113 may calculate luminance data indicating one brightness change for one chin region.

[0024] When one skin region is subdivided into multiple skin regions, luminance data indicating one brightness change may be calculated for each sub-region. For example, when the forehead region is subdivided into 10 sub-skin regions, the luminance data calculation unit 113 may calculate 10 pieces of luminance data corresponding to the 10 sub-skin regions.

[0025] Furthermore, luminance data representing a change in brightness for a single skin region formed by a plurality of sub-regions may be calculated from the average or median values ​​for the sub-regions. For example, if the forehead region is subdivided into ten sub-skin regions, the average or median value of the luminance data for the ten sub-skin regions may be calculated as the luminance data for the forehead region.

[0026] When calculating the luminance data of the skin region including the sub-skin region, the luminance data calculation unit 113 may perform a Fourier transform on the time series of the luminance data to convert it into a frequency domain signal, pass the converted signal through a predetermined band-pass filter corresponding to the frequency of the pulse wave to remove noise, and then perform an inverse Fourier transform on the signal in the desired frequency domain to calculate the luminance data.

[0027] As described above, calculating the luminance data by the luminance data calculation unit 113 includes cases where the acquired luminance data is used as is, cases where statistical values ​​such as the average or median are calculated for the acquired luminance data, and cases where a Fourier transform process is performed on the time series of the acquired luminance data.

[0028] (Blood flow correlation index calculation unit) The blood flow correlation index calculation unit 121 calculates the value of a blood flow correlation index that correlates with blood flow (also referred to as blood flow rate or cardiac output) from the luminance data calculated by the luminance data calculation unit 113. Blood flow rate refers to the amount of blood pumped out by cardiac contraction. The blood flow correlation index may also be referred to as a blood flow correlation amount.

[0029] An example of an index correlated with blood flow is pulse wave amplitude. The pulse wave amplitude can be calculated from the time-series luminance data calculated by the luminance data calculation unit 113. FIG. 2 illustrates a method for calculating the pulse wave amplitude from luminance data. For example, when data is acquired using a camera image, the brightness of the skin fluctuates depending on, for example, the amount of hemoglobin in the blood vessels within the skin region. In other words, this time-series luminance data can be considered to contain a pulse wave component. The pulse wave amplitude is a value that can be calculated as the difference between the upper and lower limits within a certain section of the time-series data. Furthermore, by shifting the section for calculating the pulse wave amplitude by an arbitrary time width (stride), discrete data of the pulse wave amplitude can be acquired in real time. In other words, the difference between the upper and lower limits is acquired as the pulse wave amplitude from the time-series luminance data using a sliding window of an appropriate width, and the sliding window is then slid along the time axis, thereby acquiring discrete data of the pulse wave amplitude in real time.

[0030] (Region Selection Unit) The region selection unit 122 removes regions where noise or invalid values ​​(including influences due to occlusion, etc.) occur in the luminance data calculated by the luminance data calculation unit 113 or the blood flow correlation index data calculated by the blood flow correlation index calculation unit 121, or selects the best region. Removing a region means that the data in that region will not be used, and includes, for example, discarding the data in that region. Selecting a region means that the data in that region will be used, and ultimately means that the information output by the region selection unit 122 will reflect the data in that region.

[0031] The operation of the region selection unit 122 will be described with reference to Fig. 3. Fig. 3 shows an example in which a blood flow correlation index is calculated for four skin regions on the face (forehead region, chin region, left cheek region, and right cheek region). The example shown in Fig. 3A shows a case in which, when a face is imaged from the front, a good index without noise or invalid values ​​is calculated for the forehead region X, but a certain number of invalid values ​​occur for the right cheek region Y. In such a case, the region selection unit 122 performs processing to select the forehead region X and discard the right cheek region Y.

[0032] 3B , for example, if a change in the facial orientation of the subject results in an invalid value being generated in the forehead region X and a good index being calculated in the right cheek region Y, the region selection unit 122 performs processing to switch the region to be referenced for index calculation from the forehead region X to the right cheek region Y. In other words, if the blood flow correlation index for the forehead region X contains an invalid value or noise, the region selection unit 122 selects the blood flow correlation index for the right cheek region Y that does not contain an invalid value or noise.

[0033] The rejection process or switching process may be performed not only when an invalid value of 0 occurs, but also when the value is equal to or less than a predetermined threshold value. Such a threshold value may be set to a value smaller than the threshold value for determining whether or not the physical condition is abnormal.

[0034] The above description has been given on the case where noise or invalid values ​​are contained in the pulse wave amplitude, which is an example of a blood flow correlation index, but the region switching process may be performed in the same way when noise or invalid values ​​are contained in the brightness data of the skin region, rather than in the blood flow correlation index. In this case, the brightness data of the skin region is supplied to the region selection unit 122 from the brightness data calculation unit 113, and the region selection unit 122 outputs a region selection signal indicating the selected region to the blood flow correlation index calculation unit 121. Based on the region selection signal from the region selection unit 122, the blood flow correlation index calculation unit 121 calculates or outputs only the blood flow correlation index related to the selected region.

[0035] In this way, the region selection unit 122 selects the other luminance data when either the first luminance data (e.g., luminance data of region X) or the second luminance data (e.g., luminance data of region Y) contains an invalid value or noise, or selects the other blood flow correlation index when either the first blood flow correlation index (e.g., pulse wave amplitude of region X) or the second blood flow correlation index (e.g., pulse wave amplitude of region Y) contains an invalid value or noise.

[0036] The blood flow correlation index estimation device 10 includes the region selection unit 122, so that even if significant data for estimating a blood flow correlation index cannot be obtained from imaging data acquired from a certain skin region, the data can be obtained from imaging data acquired from another skin region. Therefore, the blood flow correlation index estimation device 10 can robustly estimate a blood flow correlation index.

[0037] Examples of situations where meaningful data cannot be obtained include when the subject changes their face direction, when there is a change in ambient light, or when the subject is wearing an accessory.

[0038] (Correction Unit) The blood flow correlation index calculation unit 121 calculates the "relative value" of the blood flow correlation index for each skin region. In other words, although the source blood flow is the same, the blood flow correlation index measured from data obtained by observing each skin region is a relative value that shows a different value for each skin region due to various influences, such as internal human factors such as muscle movement and external human factors such as external light. To eliminate such relativity, it is desirable to correct the value of the pulse wave amplitude when switching regions using the region selection unit 122. Therefore, the index calculation unit 12 of the blood flow correlation index estimation device 10 may be provided with a correction unit 123 that corrects the value of the blood flow correlation index, such as the pulse wave amplitude.

[0039] The function of the correction unit 123 will be described with reference to Fig. 4A and Fig. 4B. Fig. 4A and Fig. 4B are diagrams showing an example of correction when there is a difference in the blood flow correlation index value, which is a relative value, between region X and region Y.

[0040] Differences in the blood flow correlation index values ​​may affect the processing of the abnormal physical condition determination unit 20, which will be described later. For example, in the example shown in Figures 4A and 4B, the blood flow correlation index drops sharply the moment the region is switched. As a result, the abnormal physical condition determination unit 20 may erroneously determine that the physical condition of the person being evaluated is abnormal.

[0041] To prevent such false detection, the correction unit 123 eliminates the difference (relativity) that occurs when the region is switched. As an example, the correction unit 123 standardizes or normalizes the blood flow correlation index value in the resting section where the subject is at rest, thereby eliminating the difference (relativity) that occurs when the region is switched. Note that FIG. 4B shows an example of standardization. As another example, the correction unit 123 may, when switching regions, refer to the blood flow correlation index value (previous value) of the region before switching and correct the blood flow correlation index value after switching to match the previous value.

[0042] (Abnormal Physical Condition Determination Unit) The abnormal physical condition determination unit 20 determines whether an abnormal physical condition has occurred based on a decrease or increase in the blood flow correlation index value (including after correction). For example, if the subject is a vehicle driver, a sudden decrease or increase in the blood flow correlation index value within a range in which the blood flow correlation index does not generate an invalid value can be considered a sign of cerebral disease or fainting. Blood flow is closely related to blood pressure, and an increase in blood flow leads to an increase in blood pressure. Since fainting is a phenomenon in which consciousness is lost due to a short-term decrease in cerebral blood flow, observing the correlation index of blood flow (blood pressure) is effective in detecting fainting. Note that "sudden" refers to a change of more than a predetermined amount over a certain period of time.

[0043] As another example, if the person being evaluated is an ordinary person, and a tendency for gradual changes in the blood flow correlation index value during daily life is detected, the abnormal physical condition determination unit 20 may determine that the person being evaluated is in abnormal physical condition.

[0044] (Biological information calculation unit) The biological information calculation unit 124 calculates biological information that does not correlate with blood flow, such as pulse rate or pulse interval (RR interval), from the luminance time series data. That is, the biological information calculation unit 124 calculates biological information other than the blood flow correlation index. Techniques for calculating pulse rate or pulse interval from luminance time series data are known and are described, for example, in Japanese Patent Laid-Open No. 2018-68720 or Japanese Patent Laid-Open No. 2020-527088.

[0045] The biological information calculation unit 124 supplies the calculated biological information to the abnormal physical condition determination unit 20. By using such biological information, the abnormal physical condition determination unit 20 can detect abnormal physical conditions in more detail and prevent false detection. For example, based on the blood flow correlation index (including the corrected blood flow correlation index) output from the blood flow correlation index estimation device 10 and biological information other than the blood flow correlation index output from the biological information calculation unit 124, the abnormal physical condition of the person is determined to be abnormal only when both are abnormal.

[0046] The data used by the biometric information calculation unit 124 is not limited to the above-described brightness data, and may include data acquired from other biometric sensors (not shown).

[0047] The biological information calculation unit 124 may be provided as a component of the abnormal physical condition detection system 50 instead of as a component of the blood flow correlation index estimation device 10 .

[0048] (Vehicle Information Acquisition Unit) The vehicle information acquisition unit 125 acquires data such as the steering angle or speed of the vehicle as vehicle information, for example, from CAN (Controller Area Network) data. The vehicle information acquisition unit 125 supplies the acquired vehicle information to the physical condition abnormality determination unit 20. By using the vehicle information, the physical condition abnormality determination unit 20 can detect physical condition abnormalities in more detail and can also prevent erroneous detection. Note that the vehicle information acquired by the vehicle information acquisition unit 125 may include, for example, data acquired from a sensor (not shown) that observes the surrounding environment or the interior of the vehicle, such as the interior of the vehicle.

[0049] The vehicle information acquisition unit 125 may be provided as a component of the abnormal physical condition detection system 50 instead of as a component of the blood flow correlation index estimation device 10 .

[0050] In this way, by providing the biometric information calculation unit 124 or the vehicle information acquisition unit 125, it becomes possible to determine abnormal physical conditions with higher accuracy.

[0051] Next, referring to Figures 5A and 5B, an example of the hardware configuration of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 will be described. Each function of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 is realized by a processing circuitry. The processing circuitry may be a dedicated processing circuit 100a as shown in Figure 5A, or a processor 100b as a computer that executes a program stored in memory 100c as shown in Figure 5B.

[0052] When the processing circuitry is a dedicated processing circuit 100a, the dedicated processing circuit 100a may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (application specific integrated circuit), an FPGA (field-programmable gate array), or a combination thereof. The functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 may be realized by a plurality of separate processing circuits, or the functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 may be realized together by a single processing circuit.

[0053] When the processing circuitry is the processor 100b, the functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 are realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the memory 100c. The processor 100b realizes the functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 by reading and executing the programs stored in the memory 100c. Here, examples of the memory 100c include non-volatile or volatile semiconductor memory such as RAM (random access memory), ROM (read-only memory), flash memory, EPROM (erasable programmable read-only memory), and EEPROM (electrically erasable programmable read-only memory), as well as magnetic disks, flexible disks, optical disks, compact disks, minidisks, and DVDs.

[0054] Note that some of the functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 may be realized by dedicated hardware, and other functions may be realized by software or firmware. In this way, the processing circuit can realize the functions of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 by hardware, software, firmware, or a combination of these.

[0055] <Operation> Next, the operation of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing an example of the operation of the blood flow correlation index estimation device or the abnormal physical condition detection system.

[0056] (Step ST1) In step ST1, the imaging data acquisition unit 111 acquires imaging data of time-series images from a device such as a camera (not shown). The imaging data is supplied to the skin area detection unit 112, which detects a skin area from the imaging data.

[0057] (Step ST2) In step ST2, the brightness data calculation unit 113 calculates brightness data indicating the brightness (brightness change) of each skin region detected by the skin region detection unit 112.

[0058] (Step ST3) In step ST3, the blood flow correlation index calculation unit 121 calculates the value of the blood flow correlation index from the luminance data calculated by the luminance data calculation unit 113. For example, the blood flow correlation index of region X that correlates with the blood flow is calculated from the luminance data of region X, and the blood flow correlation index of region Y that correlates with the blood flow is calculated from the luminance data of region Y.

[0059] (Step ST4) In step ST4, the region selection unit 122 removes regions where noise or invalid values ​​occur in the brightness data calculated by the brightness data calculation unit 113 or the blood flow correlation index data calculated by the blood flow correlation index calculation unit 121, or selects the best region.

[0060] (Step ST5) In step ST5, the correcting unit 123 corrects the value of the blood flow correlation index when the region selecting unit 122 switches the region.

[0061] (Step ST6) In step ST6, the abnormal physical condition determining unit 20 of the abnormal physical condition detection system 50 determines whether an abnormal physical condition has occurred based on the decrease or increase in the blood flow correlation index value.

[0062] Next, with reference to Fig. 7, the operation of the blood flow correlation index estimation device 10 or the abnormal physical condition detection system 50 when it includes the biological information calculation unit 124 or the vehicle information acquisition unit 125 will be described. Fig. 7 is a flowchart showing an example of the operation of the blood flow correlation index estimation device or the abnormal physical condition detection system. In Fig. 7, the operations related to the steps with the same reference numerals as in Fig. 6 have already been described. Here, steps not included in Fig. 6 will be described.

[0063] (Step ST1-2) In step ST1-2, the vehicle information acquisition unit 125 acquires vehicle information. The processing by the vehicle information acquisition unit 125 may be performed in parallel with the processing in steps ST1 to ST5.

[0064] (Step ST3-2) In step ST3-2, the biological information calculation unit 124 calculates biological information that does not have a correlation with blood flow volume from the time-series data of brightness. The processing by the biological information calculation unit 124 may be performed in parallel with the processing in steps ST3 to ST5.

[0065] In step ST6, the abnormal physical condition determination unit 20 may determine the occurrence of an abnormal physical condition based on the vehicle information or biological information in addition to the vehicle information or biological information.

[0066] Second Embodiment Next, an abnormal physical condition detection system according to a second embodiment of the present disclosure will be described with reference to Fig. 8. Fig. 8 is a diagram showing the abnormal physical condition detection system according to the second embodiment. The abnormal physical condition detection system according to the second embodiment includes a blood flow correlation index estimation device 10 as a local device and an abnormal physical condition determination device 40 as a server. The blood flow correlation index estimation device 10 and the abnormal physical condition determination device 40 are connected via a communication network (not shown) so that the blood flow correlation index value output by the blood flow correlation index estimation device 10 can be received by the abnormal physical condition determination device 40. The following describes an example in which the abnormal physical condition detection system according to the second embodiment is applied to medical use.

[0067] The blood flow correlation index estimation device 10 is as described in accordance with the first embodiment. The blood flow correlation index estimation device 10 acquires time-series images (video) of a patient receiving medical services and outputs a time-series of blood flow correlation index values. The output blood flow correlation index values ​​may be obtained after correcting the blood flow correlation index values ​​acquired between multiple regions.

[0068] The abnormal physical condition determination device 40 includes a data storage unit 30 that stores blood flow correlation index values ​​transmitted from the blood flow correlation index estimation device 10, and an abnormal physical condition determination unit 20 that determines whether an abnormal physical condition has occurred based on the data stored in the data storage unit 30. The data storage unit 30 is realized by a memory. The abnormal physical condition determination unit 20 is realized by a processing circuit or processor, similar to FIG. 5A or 5B .

[0069] The results of the determination by the abnormal physical condition determination device 40 are provided to the patient or the doctor. The doctor may make a diagnosis based on the results of the determination by the abnormal physical condition determination device 40.

[0070] It is possible to combine the embodiments, and to modify or omit each embodiment as appropriate.

[0071] The blood flow correlation index estimation device of the present disclosure can be used as a device for monitoring the physical condition of people such as vehicle occupants and patients.

[0072] 10 Blood flow correlation index estimation device, 11 Data acquisition unit, 12 Index calculation unit, 20 Abnormal physical condition determination unit, 30 Data storage unit, 40 Abnormal physical condition determination device, 50 Abnormal physical condition detection system, 100a Processing circuit, 100b Processor, 100c Memory, 111 Imaging data acquisition unit, 112 Skin area detection unit, 113 Brightness data calculation unit, 121 Blood flow correlation index calculation unit, 122 Area selection unit, 123 Correction unit, 124 Biometric information calculation unit, 125 Vehicle information acquisition unit.

Claims

1. A blood flow correlation index estimation device comprising: an imaging data acquisition unit that acquires imaging data of a first region and a second region of human skin; a brightness data calculation unit that acquires first brightness data of the first region and second brightness data of the second region from the imaging data; a blood flow correlation index calculation unit that calculates a first blood flow correlation index that correlates with blood flow from the first brightness data and calculates a second blood flow correlation index that correlates with blood flow from the second brightness data; and a region selection unit that, when either the first brightness data or the second brightness data contains an invalid value or noise, selects the other brightness data and outputs a selection signal indicating the selected brightness data, or, when either the first blood flow correlation index or the second blood flow correlation index contains an invalid value or noise, selects the other blood flow correlation index.

2. A blood flow correlation index estimation device as described in claim 1, further comprising a correction unit that performs correction between the first blood flow correlation index and the second blood flow correlation index when selecting either the first blood flow correlation index or the second blood flow correlation index.

3. The blood flow correlation index estimation device according to claim 2, wherein the first region and the second region are regions on the face of the person.

4. The blood flow correlation index estimation device according to claim 3, wherein the person is a driver of a vehicle.

5. An abnormal physical condition detection system comprising: a blood flow correlation index estimation device according to any one of claims 1 to 4; and an abnormal physical condition determination unit that determines whether the person's physical condition is abnormal based on information output from the blood flow correlation index estimation device.

6. An abnormal physical condition detection system comprising: a blood flow correlation index estimation device according to any one of claims 1 to 4; a biological information calculation unit that calculates biological information other than the blood flow correlation index from the first luminance data or the second luminance data; and an abnormal physical condition determination unit that detects whether the person's physical condition is abnormal based on the information output from the blood flow correlation index estimation device and the biological information.

7. An abnormal physical condition detection system comprising: a blood flow correlation index estimation device according to any one of claims 1 to 4; a vehicle information acquisition unit that acquires vehicle information; and an abnormal physical condition determination unit that detects whether the person's physical condition is abnormal based on the information output from the blood flow correlation index estimation device and the vehicle information.

8. A blood flow correlation index estimation method performed by a blood flow correlation index estimation device comprising an imaging data acquisition unit, a brightness data calculation unit, a blood flow correlation index calculation unit, and an area selection unit, comprising: a step in which the imaging data acquisition unit acquires imaging data of a first area and a second area of ​​human skin; a step in which the brightness data calculation unit acquires first brightness data of the first area and second brightness data of the second area from the imaging data; a step in which the blood flow correlation index calculation unit calculates a first blood flow correlation index correlating with blood flow from the first brightness data, and calculates a second blood flow correlation index correlating with blood flow from the second brightness data; and a step in which the area selection unit selects the other brightness data when either the first brightness data or the second brightness data contains an invalid value or noise, and outputs a selection signal indicating the selected brightness data, or selects the other blood flow correlation index when either the first blood flow correlation index or the second blood flow correlation index contains an invalid value or noise.

9. A blood flow correlation index estimation program that causes a computer to execute the following functions: acquiring imaging data of a first region and a second region of human skin; acquiring first luminance data of the first region and second luminance data of the second region from the imaging data; calculating a first blood flow correlation index that correlates with blood flow from the first luminance data, and calculating a second blood flow correlation index that correlates with blood flow from the second luminance data; and selecting the other luminance data when either the first luminance data or the second luminance data contains an invalid value or noise, and outputting a selection signal indicating the selected luminance data, or selecting the other blood flow correlation index when either the first blood flow correlation index or the second blood flow correlation index contains an invalid value or noise.

Images