Biometric authentication by vascular studies

Abstract

Presented herein are methods of authenticating unknown individuals based on changes in spatial properties and directionality of blood flow through blood vessels over time. At a high level, these methods rely on monitoring blood vessel dynamics to identify unknown individuals. For example, an authentication platform can examine digital images of an anatomical region to establish how properties of a vasculature within the anatomical region change due to morphing. Examples of properties include locations, sizes, volumes, and pressures of blood vessels included in the vasculature, as well as velocities and accelerations of blood flow through the vasculature.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to U.S. Provisional Application No. 63 / 127,054, filed December 17, 2020, entitled “Vein Map Authentication with Image Sensor,” the entirety of which is incorporated herein by reference. Technical Field

[0003] This disclosure relates to biometric authentication in computer security, and more specifically, to techniques for examining physiological characteristics to achieve minimally destructive authentication. Background Technology

[0004] Biometric authentication procedures verify an individual's identity using biometric technology. The term "biometric" refers to physical or behavioral characteristics that can be used as a means of verifying identity. Biometric technology is difficult to fake and convenient because the individual does not need to remember a password or manage a token. Instead, the authentication mechanism is an integral part of the individual.

[0005] Historically, fingerprints have been the most common biometric method. However, with technological advancements, other biometric patterns have emerged. As an example, vascular pattern recognition (also known as "vein pattern recognition") uses near-infrared light to create images of subcutaneous blood vessels (or simply "vessels"). These subcutaneous vessels collectively define a "vascular pattern" or "vein map" that can be used for authentication. Vascular pattern authentication has shown promise because vascular patterns are not only unique to the respective individuals, but also undergo only minimal changes as individuals age.

[0006] Vein pattern authentication typically involves identifying and subsequently analyzing vascular patterns on the back of the hand. For example, near-infrared light generated by a light-emitting diode (LED) can be emitted toward the back of the hand to penetrate the skin. Due to differences in absorption rates between blood vessels and other tissues, the near-infrared light will be reflected to different depths of the skin. Based on the analysis of the reflected near-infrared light, the vascular pattern can be inferred, and based on the vascular pattern, features such as branch locations and angles can be determined (and then used for authentication).

[0007] Vein pattern authentication is touted as a contactless option for biometric authentication, offering relative immunity to forgery because vascular patterns are difficult to reproduce. Furthermore, vein pattern authentication shows significant improvements over other biometric authentication methods in terms of false acceptance (also known as "false positive rate") and false rejection (also known as "false negative rate"). However, vein pattern authentication has several drawbacks. For example, individuals are often reluctant to expose their bodies to unfamiliar light sources, which typically come with the scanning equipment required for vein pattern authentication. Additionally, such scanning equipment is difficult or even impossible to deploy in some environments and is prohibitively expensive for many businesses. Attached Figure Description

[0008] Figure 1 This includes advanced illustrations of traditional authentication procedures, in which unknown individuals are prompted to show their hands to a blood vessel scanner.

[0009] Figure 2 The illustration shows how each cardiac cycle is displayed as a peak in a photoelectrovascular volumetric graph.

[0010] Figure 3 This includes a high-level representation of the system, which can be used to authenticate the identity of unknown individuals whose vascular systems are available for imaging.

[0011] Figure 4 The illustration shows an example of an electronic device capable of implementing an authentication platform designed to authenticate the identity of unknown persons based on image data generated by an image sensor.

[0012] Figure 5 The illustration shows how the underlying vascular system in an anatomical region (here, the face) is altered by performing postures that cause physical deformation of the surrounding tissues.

[0013] Figures 6A-6C Several different methods for determining, calculating, or otherwise obtaining pulse waves are described.

[0014] Figure 7 This includes a flowchart of a procedure for authenticating users of an authentication platform based on the analysis of visual evidence of vascular dynamics in anatomical regions.

[0015] Figure 8 This includes a flowchart of the processes performed by the authentication platform during the usage phase (also known as the "implementation phase").

[0016] Figure 9 This includes a flowchart of another process for determining whether an unknown person should be identified as a given individual through vascular studies.

[0017] Figure 10 This includes a visual illustration of the process by which an authentication platform determines whether to authenticate an unknown person as a given individual.

[0018] Figure 11 The flowchart includes the process of creating a model that is trained to predict blood flow through the vascular system of an anatomical region during deformation.

[0019] Figure 12 The block diagram illustrates an example of a processing system in which at least some of the operations described herein can be implemented.

[0020] For those skilled in the art, the various features of the technology described herein will become clearer by studying the "Detailed Description" section in conjunction with the accompanying drawings. Embodiments are illustrated in the drawings by way of example and not limitation. While various embodiments are depicted for illustrative purposes, those skilled in the art will recognize that alternative embodiments can be employed without departing from the principles of the present technology. Therefore, although specific embodiments are shown in the drawings, various modifications can be made to the present technology. Detailed Implementation

[0021] To join an authentication process that relies on matching vein patterns, an individual (also referred to as a “user”) may initially be prompted to present their hand to a vascular scanner. The term “vascular scanner” can be used to refer to an imaging instrument that includes (i) an emitter operable to emit electromagnetic radiation (e.g., in the near-infrared range) into the body, and (ii) a sensor operable to sense electromagnetic radiation reflected by physiological structures within the body. Typically, a digital image is created based on the reflected electromagnetic radiation as a reference template. At a high level, the reference template represents a “ground truth” vascular pattern that can be used for authentication.

[0022] Figure 1 This includes high-level illustrations of traditional authentication procedures, where unknown individuals are prompted to present their hands to a blood vessel scanner. For example... Figure 1 As shown, a vascular scanner emits electromagnetic radiation into the hand and then creates a digital image (also known as a "scan") based on the electromagnetic radiation reflected by the blood vessels in the hand. This image indicates the vascular pattern of the hand, which can then be verified against a reference template created for a given individual during the registration phase (also known as the "registration phase"). If the digital image matches the reference template, the unknown person will be authenticated as the given individual. However, if the digital image does not match the reference template, the unknown person will not be authenticated as the given individual.

[0023] Because vascular scanners do not require direct contact with the body during scanning, vein pattern matching has become an attractive option for biometric authentication. However, vein pattern matching has proven susceptible to spoofing. As an example, Jan Krissler and Julian Albrecht demonstrated at the 2018 Chaos Communications Conference how a wax prosthetic hand could be used to bypass a vascular scanner. While spoofing is unlikely to succeed under most real-world conditions, any concerns associated with vulnerabilities could hinder the adoption of reliable biometric authentication technologies.

[0024] Therefore, this paper introduces methods for identifying unknown persons based on the temporal changes in the spatial properties and directionality of blood flow through blood vessels. At a high level, these methods rely on monitoring vascular dynamics to identify unknown persons. The term "vascular dynamics" refers to changes in the vascular system and its properties caused by deformation of the surrounding subcutaneous tissue, for example, due to the execution of posture. Examples of vascular properties include the location, size, volume, and pressure of blood vessels, as well as the velocity and acceleration of blood flowing through them.

[0025] As further discussed below, these authentication methods can be viewed as a form of extended photoplethysmogram (PPG) monitoring. The term "photoplethysmogram" refers to a volumetric map obtained optically, used to detect changes in blood volume in subcutaneous tissue. With each cardiac cycle, the heart pumps blood to the periphery of the body. Although this pressure pulse is somewhat suppressed as the blood reaches the skin, it is sufficient to cause the blood vessels in the subcutaneous tissue to dilate to a detectable degree. The volumetric changes caused by the pressure pulse can be detected by illuminating the skin and then measuring the amount of light transmitted or reflected to an image sensor. In PPG, each cardiac cycle is represented by a peak, such as... Figure 2 As shown.

[0026] Historically, pulse oximeters have been commonly used for PPG monitoring. PPG meters typically include at least one light-emitting diode (LED) that emits light through a part of the body (such as a fingertip or earlobe) towards the LED. However, PPG can also be obtained by analyzing digital images of anatomical regions of interest. In this scenario, pressure pulses might be identified by subtle changes in the color of the skin and subcutaneous tissue. However, the subtle characteristics of pressure pulses can be difficult to establish. For example, the timing and phase of pressure pulses can be difficult to detect by analyzing digital images of the face due to the complex structure of the underlying vascular system and the intricate effects of body posture and facial expressions. Deformation of subcutaneous tissue caused by body posture and facial expressions affects the resistance to blood flow through the venous network of the face, which in turn affects the signal generated by the image sensor observing the subcutaneous tissue innervated by the venous network. While the relationship between subcutaneous tissue deformation and the signal generated by the image sensor is difficult to quantify, deformation has a predictable effect on the signal (and can therefore be used as a means of authentication).

[0027] To determine whether to authenticate an unknown person as a given individual, the authentication platform (also known as an "authentication system") can determine the degree to which the vascular dynamics of the unknown person are comparable to those of the given individual. For example, suppose the unknown person wishes to authenticate themselves as the given individual. In this scenario, the unknown person might be prompted to perform a posture that causes deformation of the subcutaneous tissue (and thus the vascular system) in an anatomical region. This posture might be related to an anatomical region. For example, if the authentication platform is checking the vascular dynamics of the face, the unknown person might be prompted to smile or frown, while if the platform is checking the vascular dynamics of the hands, the unknown person might be prompted to clench their hands.

[0028] When an unknown person performs a gesture, the camera on an electronic device can generate digital images of the anatomical region. For example, the camera can generate digital images rapidly and continuously at a predetermined rhythm. As another example, the camera can generate video of the anatomical region, in which case the digital images can represent frames of the video. Based on the analysis of the digital images, the authentication platform can generate a "biometric signature" or "vascular signature" for the unknown performer. For example, the authentication platform can generate a vein model that programmatically indicates the deformation of the vascular system during the gesture. At a high level, the vein model specifies how the spatial properties of the vascular system change with posture. Alternatively, the authentication platform can estimate metrics of vascular properties based on the analysis of the digital images. For example, the authentication platform can attempt to quantify how the directionality of blood flow through the vascular system changes with posture.

[0029] The authentication platform can then compare the biometric signature with a registered biometric signature (also known as a “reference biometric signature”) associated with a given individual to determine whether the unknown person should be authenticated as that individual. For example, if the authentication platform generates a vein model that programmatically indicates how the unknown person’s vascular system deforms during posture execution, the platform can (i) obtain a vein map associated with the given individual and (ii) estimate the expected deformation during posture execution based on the vein map. As another example, if the authentication platform estimates a metric that indicates how vascular properties change during posture execution, the platform can (i) obtain a vein map associated with the given individual and (ii) estimate the expected metric during posture execution based on the vein map. As further discussed below, the vein map can be stored in a digital profile containing information about the given individual’s vascular system. For example, the digital profile may include vein maps of different anatomical regions, metrics of different vascular properties, and so on.

[0030] In summary, the authentication platform enables the presentation of a notification instructing the person to be authenticated to perform a posture that causes deformation of an anatomical region, acquire a digital image of the anatomical region generated by an electronic device as the person performs the posture, estimate the characteristics of blood flow through subcutaneous vessels in the anatomical region based on the digital image, and then determine whether to authenticate the person as that individual based on a comparison of the estimated characteristics with a digital profile associated with that individual. The estimated characteristics can be, for example, the directionality, velocity, volume, phase, or pressure of blood flow through the subcutaneous vessels.

[0031] Biometric signature-based authentication offers many of the same benefits as vein pattern matching: high accuracy, reliability, and consistency, because the information being "read" is inside the body. However, these methods are easier to implement because they do not require specialized equipment (e.g., a vascular scanner). Instead, authentication can be performed based on the analysis of digital images generated by electronic devices. While electronic devices may include specialized software, firmware, or hardware, commodity-standardized hardware (e.g., digital image sensors used in mobile phones, tablets, etc.) may be sufficient to capture high-quality digital images.

[0032] At a high level, the authentication platform is designed to facilitate a method by which individuals can link a vein map, as an authentication factor, to a measurement of blood flow as a PPG signal. Therefore, authentication can be achieved using electronic devices that cannot detect an individual's blood vessels but can detect spatially resolved PPG signals (e.g., through analysis of digital images). Specifically, the method described herein (i) enables highly secure authentication without requiring specialized equipment, (ii) allows authentication based on (e.g., deformable) knowledge factors as well as biometric information of unknown persons and given individuals, and (iii) allows authentication robust to deception and theft because new deformities can be easily identified and requested.

[0033] For illustrative purposes, embodiments may be described in the context of monitoring the vascular system in a given anatomical region. For example, embodiments may be described in the context of examining digital images of the face, palm, or fingers. However, the methods described herein are similarly applicable to the vascular system in other parts of the human body.

[0034] While not strictly necessary, the following description of implementation is within the context of instructions executable by electronic devices. The term "electronic device" is generally used interchangeably with the term "computing device," and can be used to refer to computer servers, point-of-sale (POS) systems, tablet computers, wearable devices (e.g., fitness trackers and watches), mobile phones, and so on.

[0035] While some aspects of this technology, such as certain modules, can be described as being performed entirely or primarily by a single electronic device, some implementations are carried out in a distributed environment where modules are shared among multiple electronic devices linked via a network. For example, an unknown person might be prompted to initiate an authentication process by a mobile phone generating digital images of anatomical regions, although the decision to authenticate the unknown person might be made by an authentication platform residing on a computer server, to which the mobile phone sends the digital images.

[0036] the term

[0037] In this specification, the terms "an embodiment" or "one embodiment" refer to a feature, function, structure, or characteristic described that is included in at least one embodiment of the present technology. The appearance of such phrases does not necessarily refer to the same embodiment, nor does it necessarily refer to mutually exclusive alternative embodiments.

[0038] Unless the context explicitly requires otherwise, the terms “including,” “comprising,” and “consisting of” shall be interpreted in an inclusive sense, not in an exclusionary or exhaustive sense (i.e., in the sense of “including but not limited to”). The term “based on” shall also be interpreted in an inclusive sense, not in an exclusionary or exhaustive sense. Thus, unless otherwise noted, the term “based on” means “at least partially based on.”

[0039] The terms “connection,” “coupling,” and their variations are intended to encompass any connection or coupling between two or more elements, whether direct or indirect. Connections / couplings can be physical, logical, or a combination thereof. For example, objects can be electrically or communicatively coupled to each other without sharing a physical connection.

[0040] The term "module" can refer to a software component, firmware component, or hardware component. A module is typically a functional component that generates one or more outputs based on one or more inputs. As an example, a computer program may include multiple modules responsible for performing different tasks or a single module responsible for performing all tasks.

[0041] When used to refer to a list of multiple items, the term "or" is intended to cover all of the following interpretations: any item in the list, all items in the list, and any combination of items in the list.

[0042] The order of steps performed in any process described herein is exemplary. However, these steps may be performed in various orders and combinations, unless contrary to physical possibility. For example, steps may be added to or removed from the processes described herein. Similarly, steps may be substituted or reordered. Thus, the description of any process is intended to be open-ended.

[0043] Authentication based on analysis of vascular information

[0044] This section introduces the use of vascular dynamics as an authentication platform for biometric evidence proving that an unknown person is a given individual. As discussed further below, the spatial properties and directionality of blood flow through vessels in an anatomical region can be estimated based on the analysis of one or more digital images of that region. These spatial properties and directionality of blood flow vary with deformation of the surrounding subcutaneous tissue (e.g., due to postural changes), and these variations can be used to determine whether an unknown person can be authenticated as a given individual.

[0045] Authentication platforms can be used to secure biometric-driven transactions, such as payments authorized via a hands-free interface. For example, suppose an unknown person wishes to authenticate themselves to complete a transaction. Instead of prompting the unknown person to position a body part (e.g., their hand) near a vascular scanner, biometric authentication is performed using an electronic device the unknown person has already used to initiate the transaction. For instance, if the unknown person initiates the transaction using a mobile phone they own, the phone can generate a digital image of an anatomical region (e.g., the face), which can be analyzed by the authentication platform. As discussed further below, the authentication platform can reside on the mobile phone or another electronic device (e.g., a computer server) with which the mobile phone communicates. While this authentication method relies on the analysis of blood vessels beneath the skin, the mobile phone does not need to touch the skin. Instead, the unknown person can simply be prompted to use the mobile phone to generate a digital image of an anatomical region for authentication purposes. Thus, authentication platforms can allow people to authenticate themselves in a minimally invasive manner by relying on information about vascular dynamics.

[0046] In some embodiments, the authentication platform operates independently to authenticate the identity of an unknown person, while in other embodiments, it operates in conjunction with another system. For example, a payment system may interface with the authentication platform to ensure transactions are completed securely and without hassle. As an example, the authentication platform may facilitate contactless payment procedures in which an unknown person is allowed to initiate or complete a transaction by making a body part available for imaging. As described above, an unknown person can make a body part available for imaging simply by positioning it within the field of view of an electronic device's camera. While electronic devices are typically used to initiate or complete transactions, this is not always necessary.

[0047] Note that while the embodiments may discuss authentication within the context of initiating or completing a transaction, authentication can function in a variety of contexts. For example, suppose a group of individuals are invited to attend a network-accessible meeting where sensitive information will be shared. Each person attempting to enter the network-accessible meeting may need to be authenticated by an authentication platform before being allowed access.

[0048] Overview of the authentication platform

[0049] Figure 3 This includes a high-level representation of system 300, which can be used to authenticate the identity of an unknown person whose vascular system is available for imaging. For example... Figure 3As shown, system 300 includes an authentication platform 302, which may have access to a user interface (UI) 304, an image sensor 306, a light source 308, a processor 310, or any combination thereof. As further discussed below, these elements of system 300 may be embedded in the same electronic device or distributed among multiple electronic devices. For example, authentication platform 302 may reside partially or entirely on a network-accessible server system, while UI 304, image sensor 306, light source 308, and processor 310 may reside on another electronic device responsible for generating digital images of unknown persons.

[0050] UI 304 represents the interface through which an unknown person can interact with system 300. UI 304 can be a voice-driven graphical user interface (GUI) displayed on the display of an electronic device. Alternatively, UI 304 can be a non-voice-driven GUI displayed on the display of an electronic device. In such embodiments, UI 304 can visually indicate body parts presented for authentication purposes. For example, UI 304 can visually prompt the unknown person to position their body so that image sensor 306 can observe anatomical areas. As an example, UI 304 can include a "live view" of the digital image generated by image sensor 306 so that the unknown person can easily align their face with image sensor 306. As another example, UI 304 can present a diagram indicating where the unknown person should place their hands so that image sensor 306 can image the palm or back of the hand. Furthermore, UI 304 can present the authentication decision ultimately made by authentication platform 302.

[0051] Image sensor 306 can be any electronic sensor capable of detecting and transmitting information to generate a digital image. Examples of image sensors include charge-coupled device (CCD) sensors and complementary metal-oxide semiconductor (CMOS) sensors. Image sensor 306 can be implemented in a camera module (or simply "camera"). In some embodiments, image sensor 306 is one of a plurality of image sensors implemented in an electronic device. For example, image sensor 306 may be included in a front-facing or rear-facing camera built into a mobile phone.

[0052] Typically, digital images are generated by combining an image sensor 306 with ordinary visible light. However, the image data representing a digital image can take various formats, color spaces, and so on. For example, the image sensor 306 can be implemented in a camera designed to output image data according to a red-green-blue (RGB) color model, so that each pixel is assigned a separate red, green, and blue chromaticity value. As another example, the image sensor 306 can be implemented in a camera designed to output image data according to one of the YCbCr color spaces, so that each pixel is assigned a single luminance component value (Y) and a pair of chromaticity component values ​​(Cb, Cr).

[0053] Light source 308 includes one or more light emitters capable of emitting light in the visible or non-visible light range. For example, light source 308 may include a light emitter capable of emitting white light when a digital image is generated by image sensor 306. Additionally or alternatively, light source 308 may include a light emitter capable of emitting ultraviolet or infrared light. Examples of light emitters include light-emitting diodes (LEDs), organic LEDs (OLEDs), resonant-cavity LEDs (RCLEDs), quantum dots (QDs), lasers (e.g., vertical-cavity surface-emitting lasers (VCSELs)), superluminescent diodes (SLEDs), and various phosphors.

[0054] Those skilled in the art will recognize that if image sensor 306 is instructed (e.g., by processor 310) to combine light emitted from light source 308 (whether visible or non-visible) to generate a digital image, then image sensor 306 must be designed to detect electromagnetic radiation within an appropriate range. Other examples of image sensors besides CCD and CMOS include monolithically integrated germanium (Ge) photodiodes, indium gallium arsenide (InGaAs) photodiodes, mercury cadmium tellurium (HgCdTe) photodiodes, and other photodetectors (e.g., photodiodes) designed for the infrared and ultraviolet regions of the electromagnetic spectrum.

[0055] Therefore, image sensor 306 and light source 308 can operate together to generate digital images of anatomical regions under certain lighting conditions. For example, image sensor 306 can generate a series of digital images when light source 308 emits light within the visible light range. As another example, image sensor 306 can generate at least one digital image when light source 308 emits light within the visible light range, and at least one digital image when light source 308 emits light outside the visible light range.

[0056] As described above, image sensor 306 and light source 308 can be embedded in a single electronic device. In some embodiments, the electronic device is associated with an unknown person. For example, image sensor 306 and light source 308 can be embedded in a mobile phone associated with an unknown person. In other embodiments, the electronic device is not associated with an unknown person. For example, image sensor 306 and light source 308 can be embedded in a POS system through which an unknown person is attempting to complete a transaction.

[0057] This electronic device can be referred to as a "vascular monitoring device" because it is responsible for monitoring changes in the vascular system within an anatomical region of interest. During the imaging portion of the authentication session, the vascular monitoring device can collect image data related to the anatomical region. As further discussed below, the authentication platform 302 may be able to identify pulse waves by examining the image data. The term "pulse wave" can refer to a change in color along the surface of the anatomical region caused by the movement of blood through the underlying subcutaneous tissue. While color changes may be difficult (or even impossible) to detect with the human eye, the authentication platform 302 may be able to identify these changes by analyzing the image data. Because pulse waves are associated with the cardiac cycle, information about the vascular system within the anatomical region (and the cardiovascular system as a whole) can be gathered from the pulse waves.

[0058] like Figure 3 As shown, the authentication platform 302 may include a flow prediction algorithm 312, a flow measurement algorithm 314, a pattern matching algorithm 316, an authentication algorithm 318, and a biometric database 320. The biometric database 320 can store biometric data representing collected information related to vascular attributes, which can be used to identify known individuals. The biometric data in the biometric database 320 may vary depending on the authentication method employed by the system 300. The biometric data in the biometric database 320 may be encrypted, hashed, or otherwise obfuscated to prevent unauthorized access.

[0059] For example, the biometric database 320 may include digital profiles of various individuals, and each digital profile may include a vein map of the corresponding individual that can be used for authentication. Each vein map may consist of, or be constructed from, two-dimensional or three-dimensional image data of a corresponding anatomical region. For example, suppose the authentication platform 302 is programmed to determine whether to authenticate an unknown person as a given individual based on the deformation of the vascular system in the anatomical region. In such a scenario, the authentication platform 302 may prompt the unknown person to perform a posture and then establish the deformation of the vascular system in the anatomical region by analyzing image data generated by image sensor 306. The authentication platform 302 may then compare this deformation with a vein model associated with the given individual. At a high level, the vein model may programmatically indicate how the vascular system of the given individual deforms during the performance of the posture. In other words, the vein model may represent a series of discrete locations indicating how the shape of a single blood vessel or ensemble of blood vessels changes over time as the posture is performed, resulting in deformation of the surrounding subcutaneous tissue.

[0060] This vein model can be created in several different ways. In some embodiments, during the registration phase, a given individual is prompted to perform a pose while being imaged, and the vein model is created based on analysis of the resulting digital images. In other embodiments, anatomical regions of the given individual are imaged so that a vein map can be generated by the authentication platform 302. In such embodiments, the authentication platform 302 can simulate the deformation of the vascular system during pose execution based on the vein map.

[0061] Digital profiles may include a single vein model associated with a single pose, multiple vein models associated with a single pose, or multiple models associated with different poses. Similarly, digital profiles may include a single vein model associated with a single anatomical region, multiple vein models associated with a single anatomical region, or multiple models associated with different anatomical regions. During the registration phase, individuals may be allowed to specify which anatomical region(s) and which pose(s) can be used for authentication. While authentication platform 302 may require the creation of at least one vein model for each anatomical region and pose pair, individuals may be allowed to create multiple vein models (e.g., for improved robustness).

[0062] Additionally or alternatively, the digital profile may include reference values ​​for different vascular properties that can be used for authentication (e.g., the velocity of blood flowing through a vessel in a given anatomical region). Thus, the biometric database 320 may include data indicating temporal changes in vascular properties of an individual vessel or group of vessels during the execution of a posture. Authentication may be based on the similarity between values ​​of vascular properties such as pressure and flow velocity, rather than on the similarity between spatial deformations of the vascular system, or in addition to the similarity between spatial deformations of the vascular system.

[0063] As described above, the biometric database 320 may include one or more biometric signatures. The nature of each biometric signature may depend on how authentication is performed. For example, each biometric signature may represent a vein model created for an individual during the registration phase. Alternatively, each biometric signature may represent one or more values ​​that indicate the temporal changes in vascular properties as deformation of subcutaneous tissue occurs in the anatomical region. As an example, a biometric signature may include a vector of length N, where each element is a value specifying the rate at which blood flows through the vascular system of the anatomical region during the execution of the posture. N may represent the number of samples obtained during the execution of the posture. In other words, N may represent the number of digital images of the anatomical region generated during the execution of the posture, since the flow rate can be estimated independently for each digital image.

[0064] Biometric signatures in biometric database 320 can be associated with a single individual, in which case authentication platform 302 may be limited to authenticating an unknown person as that individual. Alternatively, these biometric signatures can be associated with multiple individuals, in which case authentication platform 302 may be able to authenticate an unknown person as any of these individuals. Furthermore, as mentioned above, a single individual can have multiple biometric signatures in biometric database 320. These biometric signatures can correspond to different types (e.g., values ​​of vein models and vascular attributes), different anatomical regions, or different postures. For example, an individual may choose to create multiple biometric signatures for different anatomical regions during the registration phase, and there may be a different biometric signature for each anatomical region. As another example, an individual may choose to create multiple biometric signatures for different postures during the registration phase, and there may be a different biometric signature for each posture.

[0065] When executed by processor 310, the algorithm implemented in authentication platform 302 allows individuals to generate biometric signatures during the registration phase. Subsequently, the algorithm implemented in authentication platform 302 allows verification to occur during the usage phase. The registration and usage phases will be further described below.

[0066] Flow prediction algorithm 312 can be responsible for determining the relative timing of a pulse wave in an anatomical region by analyzing one or more digital images of the anatomical region. For example, flow prediction algorithm 312 can determine the timing or phase of the pulse wave in certain spatial coordinates (e.g., a specified anatomical region) with or without physical deformation based on digital images. When determined without physical deformation, this measurement may be referred to as "measured venous flow pattern" or "measured flow pattern," while when determined with physical deformation, the measurement may be referred to as "measured deformed venous flow pattern" or "measured deformed flow pattern." As an example, the relative arrival timing of the pulse wave can be estimated based on the identification of characteristics of the pulse wave (e.g., dicrotic notch). Based on this information, flow prediction algorithm 312 can estimate the rate of blood flow through the vascular system in the anatomical region. Alternatively, flow prediction algorithm 312 can estimate another vascular property, such as the phase of a pressure pulse, the direction of blood flow, the volume of blood flow, or the pressure in the vascular system of the anatomical region.

[0067] The flow measurement algorithm 314 is responsible for predicting the propagation pattern of the pulse wave that will occur in the anatomical region during deformation. This propagation pattern may be referred to as the "predicted deformation venous flow pattern" or "predicted deformation flow pattern." To achieve this, the flow measurement algorithm 314 can create a PPG by modeling, estimating, or otherwise predicting how the pulse wave will propagate through the anatomical region in the deformation state. The flow measurement algorithm 314 can take the measured flow pattern, the venous map, and the deformation venous map as inputs. As mentioned above, the deformation venous map can be determined based on at least one digital image of the anatomical region in the deformation state, or the deformation venous map can be determined by modifying the venous map to simulate deformation.

[0068] In some embodiments, the flow measurement algorithm 314 is a machine learning algorithm. For example, the flow measurement algorithm 314 may be based on a neural network whose parameters are predetermined based on best practice examples or adjusted experimentally.

[0069] Predicted deformation flow patterns can be expressed using two-dimensional or three-dimensional coordinates relating to the surface of the anatomical region. Furthermore, predicted deformation flow patterns can be correlated with (i) timing information and (ii) phase information. Timing information can relate to the relative time it takes for a pressure pulse to reach that coordinate after arriving at the anatomical region. For example, timing information can correspond to identifiable features of the pressure pulse, such as a dicrotic notch or another portion of the pulse wave representing the pressure pulse. Phase information can relate to the relative phase of the pressure pulse that may exist at each coordinate at a single point in time. Due to the varying effects of pressure pulses across the entire anatomical region, each coordinate within the anatomical region may have a different pulse waveform.

[0070] Pattern matching algorithm 316 can be responsible for calculating the strength of the match between the predicted deformation flow pattern and the measured deformation flow pattern. In other words, pattern matching algorithm 316 can be responsible for determining the degree of similarity between the predicted deformation flow pattern and the measured deformation flow pattern. This degree of similarity can be expressed using a metric called a "match score". The match score can be expressed using any suitable numerical scale. For example, the match score can use any integer value between 0 and 100 or any decimal value between 0 and 1 to indicate the degree of similarity.

[0071] The authentication algorithm 318 can be responsible for determining whether to authenticate an unknown person as a given individual based on a matching score. For example, the authentication algorithm 318 can be programmed to authenticate an unknown person as a given individual if the matching score exceeds a predetermined threshold. If the matching score does not exceed the predetermined threshold, the authentication algorithm 318 may not authenticate the unknown person as a given individual. Typically, the authentication algorithm 318 is designed to output a binary signal indicating whether the authentication is appropriate (e.g., pass or fail). However, the authentication algorithm 318 can be designed to output a non-binary signal. As an example, the output produced by the authentication algorithm 318 may indicate that: (i) the unknown person should be authenticated as the given person, (ii) the unknown person should not be authenticated as the given person, or (iii) further authentication attempts are needed. If the authentication algorithm 318 cannot definitively determine whether the authentication is appropriate, the authentication platform 302 may take further action (e.g., by prompting the unknown person to perform another pose or suggesting another anatomical region for imaging).

[0072] Figure 4 An example of an electronic device 400 capable of implementing an authentication platform 414 is illustrated, designed to authenticate the identity of an unknown person based on image data generated by an image sensor 408. As described above, the image data may represent one or more digital images of anatomical regions of the body. In some embodiments, these digital images are generated based on ambient light reflected from the anatomical regions to the image sensor 408. In other embodiments, a light source 410 emits light toward the anatomical regions to illuminate them as these digital images are generated by the image sensor 408. Note that the light source 410 may also be configured to emit a series of discrete “pulses” or “flashes” at certain time intervals.

[0073] In some embodiments, the authentication platform 414 is embodied as a computer program executed by an electronic device 400. For example, the authentication platform 414 may reside on a mobile phone capable of acquiring image data and determining whether authentication is appropriate based on the image data. As another example, the authentication platform 414 may reside on a POS system capable of acquiring image data and making a determination based on the image data. In other embodiments, the authentication platform 414 is embodied as a computer program executed by another electronic device to which the electronic device 400 is communicatively connected. In such embodiments, the electronic device 400 may transmit image data to other electronic devices for processing. For example, while authentication of an unknown person may be sought by a POS system used to initiate a transaction, the image data may be generated by a mobile phone located near the unknown person. This image data may be provided to the POS system or another electronic device (e.g., a computer server) for processing, or the image data may be processed by the mobile phone before being delivered to the POS system or other electronic devices. Those skilled in the art will recognize that aspects of the authentication platform 414 may also be distributed among multiple electronic devices.

[0074] Electronic device 400 may include processor 402, memory 404, UI output mechanism 406, image sensor 408, light source 410, and communication module 412. Communication module 412 may be, for example, a wireless communication circuit designed to establish communication channels with other electronic devices. Examples of wireless communication circuits include integrated circuits (also referred to as "chips") configured for Bluetooth, Wi-Fi, NFC, etc. Processor 402 may have general-purpose characteristics similar to a general-purpose processor, or processor 402 may be an application-specific integrated circuit (ASIC) that provides control functions to electronic device 400. Figure 4 As shown, the processor 402 can be directly or indirectly coupled to all components of the electronic device 400 to achieve communication purposes.

[0075] Memory 404 can be composed of any suitable type of storage medium, such as static random-access memory (SRAM), dynamic random-access memory (DRAM), electrically erasable programmable read-only memory (EEPROM), flash memory, or registers. In addition to storing instructions executable by processor 402, memory 404 can also store image data generated by image sensor 408 and data generated by processor 402 (e.g., when executing modules of authentication platform 414). Note that memory 404 is only an abstract representation of the storage environment. Memory 404 can consist of actual memory chips or modules.

[0076] As described above, the light source 410 can be configured to emit light (more specifically, electromagnetic radiation) onto the anatomical regions of the body of the unknown person to be authenticated, within a visible or invisible range. Typically, the light source 410 will only emit light when instructed. For example, if the authentication platform 414 determines that authentication is necessary, the authentication platform 414 can generate an output prompting the processor 402 to (i) instruct the light source 410 to emit light and (ii) instruct the image sensor 408 to generate image data.

[0077] Communication module 412 can manage communication between components of electronic device 400. Communication module 412 can also manage communication with other electronic devices. Examples of electronic devices include mobile phones, tablet computers, personal computers, wearable devices, POS systems, and network-accessible server systems consisting of one or more computer servers. For example, in an embodiment where electronic device 400 is a mobile phone, communication module 412 can facilitate communication with a network-accessible server system responsible for examining image data generated by image sensor 408.

[0078] For convenience, authentication platform 414 may be referred to as a computer program residing in memory 404. However, authentication platform 414 may consist of software, firmware, or hardware components implemented in or accessible to electronic device 400. According to the embodiments described herein, authentication platform 414 may include components as referenced above. Figure 3Various algorithms are discussed. Typically, these algorithms are executed by separate modules of the authentication platform 414, which are individually addressable (and thus can be executed independently without interfering with other modules). These modules can be integral parts of the authentication platform 414. Alternatively, these modules can be logically separate from the authentication platform 414, but operate "together" with it. Together, these algorithms enable the authentication platform 414 to authenticate the identity of an unknown person based on analysis of vascular dynamics determined from image data generated by the image sensor 408.

[0079] For example, suppose an unknown person wishes to authenticate themselves as a given individual. In this scenario, the unknown person might be prompted to perform a posture that causes deformation of the vascular system in an anatomical region. As the unknown person performs the posture, image sensor 408 can generate image data representing a digital image of the anatomical region. Based on the analysis of the image data, authentication platform 414 can generate a "biometric signature" for the unknown executor. For example, authentication platform 414 can generate a vein model that programmatically indicates how the vascular system deforms during the posture, or authentication platform 414 can estimate measures of vascular properties based on the analysis of the image data.

[0080] The authentication platform 414 can then compare the biometric signature with registered biometric signatures associated with a given individual to determine whether the unknown person should be authenticated as that individual. Typically, registered biometric signatures are stored in a biometric database 416. Figure 4 In this system, the biometric database 416 is located in the memory 404 of the electronic device 400. However, the biometric database 416 may alternatively or additionally be located in a remote memory accessible via a network by the electronic device 400. If the biometric signature is sufficiently similar to a registered biometric signature, then the authentication platform 414 can authenticate an unknown person as a given individual.

[0081] Other elements may also be included as part of the authentication platform 414. For example, the UI module may be responsible for generating content to be output by the UI output mechanism 406 for presentation to unknown individuals. The form of this content may depend on the nature of the UI output mechanism 406. For example, if the UI output mechanism 406 is a speaker, then the content may include audio instructions requiring the electronic device 400 to be positioned such that anatomical areas are observable by the image sensor 408. As another example, if the UI output mechanism 406 is a display, then the content may include visual instructions requiring the electronic device 400 to be positioned such that anatomical areas are observable by the image sensor 408. The UI output mechanism 406 may also be responsible for outputting (e.g., transmitting or displaying) authentication decisions made by the authentication platform 414.

[0082] Vascular deformation caused by posture execution

[0083] Within a given anatomical region, the blood vessels beneath the skin define the vascular system. As an example, the face includes several anatomical regions (e.g., the forehead, cheeks, and chin) where the vascular system can be visually monitored. Figure 5 The illustration shows how the underlying vascular system in an anatomical region (here, the face) is altered by performing postures that cause physical deformation of the surrounding tissues. Figure 5 The diagram also illustrates how blood flowing through the vascular system can be monitored in terms of pulse waves. In other words, the movement of blood within the vascular system can be visually monitored as it flows towards the small arteriolar ends of capillaries in anatomical regions. Figure 5 In the diagram, any cells numbered 1-4 are used to show an example sequence of pulse wave arrival.

[0084] As mentioned above, the measured flow pattern may change due to: (i) physical movement of blood vessels, which alters their position; and (ii) deformation of surrounding tissues, which alters the blood dynamics of the vascular system of interest. For example, tissue compression may cause an increase in capillary pressure, thereby altering the relative pulse phase and pulse wave velocity. Typically, the vascular system deforms in a predictable manner each time the posture is repeated. If (i) posture, (ii) undeformed vascular pattern (e.g., Figure 5 (i) the top left image in the image) and (iii) the measured flow pattern (e.g., Figure 5 If the lower left image in the image is known, then the measured deformation flow pattern (e.g., Figure 5 (lower right image in the image) and / or deformed vascular patterns (e.g., Figure 5 The upper right image in the image can be determined. The following will refer to... Figure 9 Step 904 further elaborates on this process.

[0085] Flow patterns were established through the analysis of pulse waves.

[0086] An important aspect of the methods described in this paper is to establish patterns of blood flow through the vascular system in a given anatomical region by analyzing image data. Figures 6A-6C Several different methods for determining, calculating, or otherwise obtaining pulse waves are described. Figure 6A The illustration depicts a method where object recognition is used to define a region of interest (ROI), and a digital image of that region is generated. For example... Figure 6BAs shown, red and green pixel values ​​can be extracted for a Region of Interest (ROI) within at least one period. This can be done for multiple patches within the ROI, where the pixel values ​​are averaged across each patch. The patches can have a fixed size and be distributed within the ROI according to a segmentation function. Alternatively, the patches can have an adjustable size (e.g., based on the size of the ROI or the amount of available computational resources). The average pixel values ​​from each patch can then be used to estimate pulse wave values, which can be used to determine the phase of the pulse wave. Figure 6C As shown, these pulse wave values ​​can indicate the relative phase period in any unit.

[0087] Note that since red and green pixel values ​​relate to variations occurring at different depths (e.g., different vascular structures), these components may be affected differently by physical deformation. This difference can be a useful component, and the algorithm described in this paper can be trained to detect it and then used to create predictions. Therefore, it may be beneficial to ensure that red and green pixel values ​​are not only provided to the authentication platform but can also be used independently to compute predictions and estimates of the phase.

[0088] Authentication methods

[0089] Figure 7 This includes a flowchart of procedure 700 for authenticating users of an authentication platform based on the analysis of visual evidence of vascular dynamics in anatomical regions. As further discussed below, the authentication process has three phases: a training phase, a registration phase, and a usage phase. These phases can be designed to allow for minimally disruptive authentication without requiring users to interact with the electronic device in unusual ways. Instead, users can simply perform gestures, while the electronic device generates digital images of the anatomical regions deformed as a result of these gestures.

[0090] To illustrate the point, the authentication procedure can be described in the context of monitoring the vascular system as the surrounding subcutaneous tissue deforms due to the execution of the posture. However, the surrounding subcutaneous tissue can deform in other ways. For example, if the vascular system to be monitored is located in the finger, the user can be prompted to position their finger near the electronic device, so that the tactile feedback generated by the haptic actuator (or simply "actuator") located inside the electronic device can deform the surrounding subcutaneous tissue.

[0091] To begin the training phase, the flow prediction algorithm may undergo supervised, semi-supervised, or unsupervised learning, where training data is retrieved, created, or otherwise acquired and then provided to the flow prediction algorithm for training purposes. Training data may include measured flow patterns and / or measured deformation flow patterns, which are associated with corresponding vein maps and / or deformation vein maps. Typically, training data is associated with a single anatomical region, because an understanding of deformation in one anatomical region (e.g., the face) may not be applicable to another anatomical region (e.g., the hand). However, training data may be associated with multiple individuals. Thus, for a variety of individuals, training data may include measured flow patterns, measured deformation flow patterns, vein maps, deformation vein maps, or any combination thereof.

[0092] In some embodiments, pose learning is accomplished via transfer learning based on a model learned during normal time (e.g., a state without pose, also referred to as the "normal state"). For example, each layer of the neural network can be divided into: (i) a first layer relating to features of the corresponding individual that are not pose-dependent, and (ii) a second layer relating to features of the corresponding individual that are pose-altered. By fixing the first layer and learning the second layer only for each pose of interest, it is expected that the amount of image data used for learning (and the learning process) can be reduced.

[0093] As mentioned above, a portion of the training data (e.g., vein maps and deformed vein maps) may be image data. In some embodiments, the image data is generated from various capture angles or locations, or with various image sensors (e.g., corresponding to different electronic devices), to provide greater robustness to these variations during the usage phase.

[0094] Furthermore, training data can be divided into a training set and a test set; for example, 80% of the training data may be allocated to the training set and 20% to the test set. Those skilled in the art will recognize that these values ​​are provided for illustrative purposes. More or less than 80% of the training data may be allocated to the training set. Similarly, more or less than 20% of the training data may be allocated to the test set. The share of training data allocated to the training set is typically larger than the share allocated to the test set (e.g., in multiples of 2, 3, 5, etc.). The training set can be used to train the flow prediction algorithm, as described below, while the test set can be used to confirm that the flow prediction algorithm has correctly learned how to predict flows.

[0095] At a high level, flow prediction algorithms comprise a set of algorithms designed to produce an output (also called a “prediction”) related to the flow of blood through the vascular system of an anatomical region, given certain inputs. These inputs may include vein maps, deformed vein maps, or image data of the anatomical region. In some embodiments, this set of algorithms represents one or more neural networks. The neural network learns by processing examples, each associated with known inputs and outputs, to form a probabilistically weighted association between the inputs and outputs. These probabilistically weighted associations may be referred to as “weights.” During the training phase, randomly selected weights may initially be used by one or more neural networks of the flow prediction algorithm. These weights can be adjusted as the flow prediction algorithm learns from measured flow patterns, vein maps, and deformed vein maps. Thus, the flow prediction algorithm can adjust these weights as it learns how to output predicted deformed flow patterns.

[0096] Each predicted deformation flow pattern output by the flow prediction algorithm can be scored based on its deviation from the corresponding measured deformation flow pattern, which serves as the ground truth. Typically, this is done for each example included in the training set. More specifically, the pattern matching algorithm can calculate a score using a timing or phase threshold for each time coordinate (e.g., + / - 3, 5, or 10 milliseconds). If the timing difference between the measured and predicted deformation flow patterns exceeds the threshold, that time coordinate may be marked as a failure. The score can be calculated by the pattern matching algorithm based on the percentage of time coordinates that have been classified as failures. Furthermore, the pattern matching algorithm can compare the score to a predetermined threshold, classifying each example in the training set as "pass" or "fail" depending on whether the corresponding score exceeds the predetermined threshold. The pattern matching algorithm can calculate the overall success rate of the flow prediction algorithm based on the percentage of pass examples in the training set.

[0097] Note that the weights of one or more neural networks in a flow prediction algorithm can be adjusted according to any scheme, where the adjustment is made to optimize for success. An example of a known scheme is the Monte Carlo method. Therefore, this part of the training phase can be repeated for a predetermined number of epochs, or this part of the training phase can be repeated until the overall success rate reaches an acceptable value (e.g., 90%, 95%, or 98%).

[0098] To ensure the proper functioning of the flow prediction algorithm, a test set can be used. The algorithm can then be applied to vein maps and deformed vein maps included in the test set to generate predicted flow patterns or predicted deformed flow patterns. As described above, the pattern matching algorithm can calculate a score indicating the performance of the flow prediction algorithm based on a comparison between the predicted flow pattern or predicted deformed flow pattern and the measured flow pattern or measured deformed flow pattern, respectively.

[0099] During the registration phase (also known as the "setup phase"), vein maps and deformed vein maps can be generated for the user during normal use of the electronic device. For example, upon receiving input instructing the initiation of the registration phase, the electronic device can generate a digital image of an anatomical region while it deforms (e.g., due to a user's posture). The deformation can be prompted via UI, requesting (e.g., via text) the user to smile, frown, or purse their lips while the electronic device generates a digital image of their face. For example, the UI can display a graphical representation of the deformation using a general model of the anatomical region or the human body to visually guide the user. As another example, the UI can display a graphical representation of a posture that will cause deformation of the anatomical region. For example, this graphical representation can serve as a visual instruction to interact with the electronic device responsible for generating the digital image in a certain way (e.g., swiping a finger across the screen, gripping the casing in a certain way, etc.).

[0100] From these digital images, venous maps and deformed venous maps of anatomical regions can be generated. For example... Figure 7 As shown, vein maps and deformed vein maps are typically stored in biometric databases and then retrieved when an unknown person attempts to authenticate themselves as a user.

[0101] Figure 8 This includes a flowchart of process 800 performed by the authentication platform during the usage phase (also known as the "implementation phase"). Initially, the authentication platform receives input from a source instructing that an unknown person be authenticated as a given individual (step 801). In some embodiments, the source is a computer program that executes on the same electronic device as the authentication platform. For example, if the authentication platform resides on a mobile phone, the authentication request may originate from a mobile application through which the unknown person is attempting to perform an activity requiring authentication. In other embodiments, the source originates from another electronic device. For example, suppose an unknown person attempts to complete a transaction using a POS system associated with a merchant. In this scenario, the POS system may require authentication. While the POS system can be responsible for generating the image data required for authentication, the authentication platform may reside on a computer server that communicates with the POS system via a network.

[0102] The authentication platform can then receive (i) a vein map and (ii) a deformed vein map associated with a given individual (step 802). As described above, the deformed vein map can be associated with a pose performed by the given individual during the registration phase. Furthermore, the authentication platform can cause a notification to be presented prompting an unknown person to perform a pose (step 803). This notification is intended to prompt the unknown person to perform the same pose as the one performed by the given individual during the registration phase.

[0103] As described above, performing a posture may cause deformation of the anatomical region of interest. As an unknown person performs the posture, electronic devices can monitor this deformation. For example, the image sensor of the electronic device can generate image data by observing the anatomical region, and the authentication platform can obtain this image data for analysis (step 804). In some embodiments, the image data includes digital images generated before, during, or after the deformation of the anatomical region. For example, the electronic device can generate a first series of digital images during a first time interval before the deformation occurs, and a second series of digital images during a second time interval while the deformation is "held." Thus, the electronic device can generate digital images when the anatomical region is in its natural state (also known as a "relaxed" state) and deformed state. Typically, the first and second time intervals are long enough that at least one complete cardiac pulse cycle can be observed. While the duration of a cardiac pulse cycle varies depending on various physiological factors, it typically falls within the range of 0.5–2.0 seconds. Therefore, the first and second time intervals can be at least 1, 2, or 3 seconds. Longer durations can optionally be used to capture more than one cardiac pulse cycle.

[0104] The authentication platform can then analyze the image data to determine (i) the measured flow pattern and (ii) the measured deformation flow pattern (step 805). To achieve this, the authentication platform can apply a flow measurement algorithm to the image data. When applied to the image data, the flow measurement algorithm can initially perform a registration operation (also known as a “mapping operation”) to determine the pixel locations in the image data corresponding to certain anatomical coordinates in the vein map and the deformation vein map. This mapping operation ensures that the values ​​in these data sets are associated with the same locations in the anatomical regions. The flow measurement algorithm can then average the red or green frequency components of the image data over various pixel regions (e.g., 3x3, 6x6, or 9x9 pixel regions). The selected frequency band can roughly correspond to the frequency of the pressure pulses that bring blood to the anatomical region. However, other frequencies can be used alternatively if appropriate. The average red or green frequency components can be calculated for the image data corresponding to different time points, thereby creating a time series of the intensity of the specified frequency band of interest. For example, the flow measurement algorithm can average the red or green frequency components of different digital images (e.g., frames representing videos generated by electronic devices). Additionally, flow measurement algorithms can perform pattern recognition to determine the relative timing of individual identifiable phases of a pressure pulse (e.g., dicrotic notch) based on analysis of temporal variations in the average values ​​of red or green frequency components. After identifying this phase of the pressure pulse, the flow measurement algorithm can assign timing values ​​to some or all pixel regions relative to the earliest detected occurrence of the identified pressure pulse phase. At a higher level, these timing values ​​can represent a flow pattern indicating how blood flows through the vascular system of an anatomical region, determined from the analysis of image data. If the image data is associated with an anatomical region in its natural state, this flow pattern can be termed a "measured flow pattern." If the image data is associated with an anatomical region in its deformed state, this flow pattern can be termed a "measured deformed flow pattern."

[0105] The authentication platform can then apply the flow prediction algorithm to (i) the measured flow patterns generated for unknown individuals, (ii) the vein map of a given individual, and (iii) the deformed vein map of a given individual to generate a predicted deformed flow pattern (step 806). At a high level, the predicted deformed flow pattern can be a data structure containing timing values ​​that represent predictions about how blood might flow through a given individual's vascular system when a posture is performed.

[0106] The authentication platform can then apply a pattern matching algorithm to (i) the measured deformation flow pattern and (ii) the predicted deformation flow pattern to generate a metric indicating similarity (step 807). As mentioned above, this metric can be referred to as a "match score." At a high level, this metric indicates how comparable the measured deformation flow pattern and the predicted deformation flow pattern are on a per-value basis.

[0107] The authentication platform can then determine whether to authenticate the unknown person as a given individual based on this metric (step 808). For example, the authentication platform can apply an authentication algorithm that compares the metric to a predetermined threshold. If the metric exceeds the predetermined threshold, the authentication algorithm can produce an output indicating that the unknown person should be authenticated as a given individual. However, if the metric does not exceed the predetermined threshold, the authentication algorithm can produce an output indicating that the unknown person should not be authenticated as a given individual.

[0108] Figure 9 This includes a flowchart of another process, 900, to determine whether an unknown person should be identified as a given individual through vascular studies. Meanwhile, Figure 10 This includes a visual illustration of the process by which an authentication platform determines whether to authenticate an unknown person as a given individual. Initially, the authentication platform may receive input indicating a request to authenticate an unknown person as a given individual (step 901). Figure 9 Step 901 can be with Figure 8 Step 801 is essentially similar. For illustrative purposes, process 900 is described in the context of examining a digital image generated by an electronic device owned by an unknown person. However, those skilled in the art will recognize that process 900 can be similarly applied to scenarios where an unknown person approaches but does not own the electronic device (e.g., the electronic device is a POS system).

[0109] Subsequently, the authentication platform can obtain a digital profile associated with the given individual claimed by the unknown person (step 902). For example, the authentication platform can access a biometric database that stores digital profiles associated with different individuals, and then the authentication platform can select a digital profile from the digital profiles based on this input. Typically, this input identifies (e.g., using a name or identifier, such as an email address or phone number) the given individual claimed by the unknown person, so the authentication platform can simply identify the appropriate digital profile from the digital profiles stored in the biometric database.

[0110] A digital profile may include one or more vascular patterns (also known as “vein maps”) associated with a given individual. In addition to being associated with a given individual, each vascular pattern may also be associated with a given anatomical region. For example, a digital profile may include individual vascular patterns of the face, palm, fingers, etc. Furthermore, a digital profile may include vascular patterns of the same anatomical region in different states. For example, a digital profile may include (i) a first vascular pattern that provides spatial information about the blood vessels in the anatomical region when it is in a natural state; and (ii) a second vascular pattern that provides spatial information about the blood vessels in the anatomical region when it is in a deformed state (e.g., due to an execution posture).

[0111] Then, while the camera of the electronic device is pointed towards the anatomical region, the authentication platform can present an instruction to an unknown person to perform a posture that causes deformation of the anatomical region (step 903). As the unknown person performs the posture, the camera can generate a series of digital images. These digital images can be generated discretely and continuously at a predetermined rhythm (e.g., every 0.1, 0.2, or 0.5 seconds). Alternatively, the camera can generate a video of the anatomical region, in which case the digital images can represent frames of the video. In this scenario, the digital images can be generated at a predetermined rate (e.g., 20, 30, or 60 frames per second).

[0112] The authentication platform can then estimate the flow patterns of the unknown person based on digital images generated by the camera. More specifically, the authentication platform can estimate (i) a first flow pattern of blood when the anatomical regions of the unknown person are in their natural state, and (ii) a second flow pattern of blood when the anatomical regions of the person are in a deformed state (step 904), based on the analysis of the digital images. As described above, these flow patterns will be estimated based on different digital images generated by the camera of the electronic device. The first flow pattern can be generated based on the analysis of digital images of the anatomical regions in their natural state (e.g., those generated before or after performing a pose), while the second flow pattern can be generated based on the analysis of digital images of the anatomical regions in their deformed state (e.g., those generated when performing or maintaining a pose). As stated above regarding Figure 8 The discussion suggests that first and second flow patterns can be estimated based on a programmed analysis of the pixels of the corresponding digital images in order to identify color variations (e.g., in the red or green component) that indicate blood flow through blood vessels in anatomical regions.

[0113] Furthermore, the authentication platform can predict a third flow pattern (step 905) of blood flow through the anatomical region of a given individual if that individual were to perform the pose, based on the digital profile and the first flow pattern. As described above, the digital profile may include (i) a first vascular pattern, which provides spatial information about the blood vessels in the anatomical region of the given individual when the anatomical region is in a natural state, and (ii) a second vascular pattern, which provides spatial information about the blood vessels in the anatomical region of the given individual when the anatomical region is in a deformed state. By applying an algorithm to the first vascular pattern, the second vascular pattern, and the first flow pattern, the authentication platform may be able to generate a third flow pattern as output. At a high level, the algorithm can simulate blood flow through the blood vessels during the deformation of the anatomical region caused by the pose.

[0114] Then, the authentication platform can determine whether to authenticate the unknown person as a given individual based on a comparison of the second and third flow patterns (step 906). For example, suppose the first, second, and third flow patterns are represented as matrices. The first flow pattern can be represented as a first vector or matrix, where each element includes a value indicating the estimated blood flow through the corresponding portion of the unknown person's anatomical region in its natural state. The second flow pattern can be represented as a second vector or matrix, where each element includes a value indicating the estimated blood flow through the corresponding portion of the unknown person's anatomical region in its deformed state. Simultaneously, the third flow pattern can be represented as a third vector or matrix, where each element includes a value indicating the estimated blood flow through the corresponding portion of the given individual's anatomical region in its deformed state. In this scenario, the authentication platform can apply an algorithm to the second and third vectors or matrices to generate a score indicating the similarity between the second and third flow patterns. Thus, the authentication platform can establish the probability that the unknown person is a given individual based on this score. Note that the term "matrix" as used in this article can refer to a series of row vectors or column vectors.

[0115] Other steps may also be included. As an example, the authentication platform can generate a signal (e.g., in the form of a message or notification) indicating whether an unknown person has been authenticated as a given individual. The authentication platform can send this signal to the source from which it received the authentication request. For example, if a request to authenticate an unknown person is received from a computer program running on a mobile phone, the authentication platform can provide the signal to the mobile phone so that the computer program can determine whether to allow the unknown person to perform any task requiring authentication. Similarly, if a request to authenticate an unknown person is received from a POS system during a transaction, the authentication platform can provide a signal to the POS system so that the transaction can be completed.

[0116] As referenced above Figure 8 As discussed, performing authentication may require the authentication platform to apply a flow prediction algorithm to (i) measured flow patterns associated with an unknown person, (ii) a vein map associated with a given individual, and (iii) a deformed vein map associated with a given individual, to generate a predicted deformed flow pattern. This predicted deformed flow pattern represents the authentication platform's prediction about how blood will flow through the vascular system if a given individual were to perform a certain posture. In some embodiments, the flow prediction algorithm is part of a set of algorithms that collectively define the flow prediction model. Generally, the flow prediction model is a machine learning (ML) or artificial intelligence (AI) model that is "trained" using examples to make predictions, i.e., how blood will flow through the vascular system when deformed.

[0117] Figure 11 The flowchart includes a process 1100 for creating a model that is trained to predict blood flow through the vascular system of an anatomical region during deformation. As described above, anatomical regions can deform by the execution of a posture, or by the application of an external force (e.g., tactile feedback generated by a haptic actuator). The nature of the deformation may depend on the anatomical region. For example, the vascular system of the face may be more easily deformed by instructing a person to perform a posture (e.g., smiling or frowning), while the vascular system of the fingers may be more easily deformed by instructing a person to place their fingers on an electronic device and then applying an external force (e.g., via tactile feedback).

[0118] Initially, the authentication platform can identify the model to be trained to predict blood flow through venous networks in anatomical regions during deformation (step 1101). Note that the terms "venous network" and "vascular system" may be used interchangeably. Thus, the term "venous network" can refer to a portion of the vascular system located within an anatomical region. While an anatomical region can be any part of the body from which vascular dynamics can be monitored via imaging, common anatomical regions include the fingers, palms and backs of the hands, and the face.

[0119] Then, the authentication platform can obtain (i) a first series of vascular patterns corresponding to the anatomical region in its natural state, (ii) a second series of vascular patterns corresponding to the anatomical region in its deformed state, (iii) a series of flow patterns, for each vascular pattern in the first series, conveying how blood flows through that vascular pattern when the anatomical region is in its natural state, and (iv) a series of deformed flow patterns, for each vascular pattern in the second series, conveying how blood flows through that vascular pattern when the anatomical region is in its deformed state (step 1102). Each vascular pattern in the first series can indicate the spatial relationship between subcutaneous vessels when the anatomical region is in its natural state. Simultaneously, each vascular pattern in the second series can indicate the spatial relationship between subcutaneous vessels when the anatomical region is in its deformed state.

[0120] Furthermore, each vascular pattern in the first series can be associated with a corresponding vascular pattern in the second series, and the corresponding vascular patterns in the first and second series can be associated with the same individual. Therefore, a single individual can be associated with one of the vascular patterns in the first series, one of the vascular patterns in the second series, one of the flow patterns, and one of the deformation flow patterns. Typically, each vascular pattern in the first series is associated with a different individual, although the same individual may be associated with multiple vascular patterns in the first series. For example, a single individual may be associated with vascular patterns corresponding to the same anatomical region, but these patterns were generated using image data produced by different electronic devices. Similarly, each vascular pattern in the second series is typically associated with a different individual. However, as mentioned above, each vascular pattern in the first series may be associated with the same individual as the corresponding vascular pattern in the second series.

[0121] The authentication platform can then provide the model with (i) a first series of vascular patterns, (ii) a second series of vascular patterns, (iii) a series of flow patterns, and (iv) a series of deformed flow patterns as training data (step 1103). This method trains the model to predict blood flow through venous networks in anatomical regions of a person when applied to vascular patterns associated with that person. In other words, the authentication platform can provide this information as training data to generate a trained model capable of predicting blood flow. For example, if the authentication platform's task is to predict blood flow through anatomical regions of a given individual, the platform can apply the trained model to a pair of vascular patterns associated with that individual. This pair of vascular patterns may include one vascular pattern corresponding to an anatomical region in its natural state and another corresponding to an anatomical region in its deformed state. After training is complete, the authentication platform can store the trained model in a biometric database (step 1104).

[0122] Other considerations and implementation methods

[0123] A. A personalized attitude

[0124] As mentioned above, vein mapping can play a crucial role in determining whether an unknown person should be identified as a given individual. To customize the identification process, identification platforms can design or select deformations based on these vein maps.

[0125] For example, suppose an unknown person wishes to authenticate themselves as a given individual. As part of the authentication process, the authentication platform can obtain vein maps associated with the given individual (e.g., a first vein map of an anatomical region in its natural state and a second vein map of an anatomical region in a deformed state). In this scenario, the authentication platform can design or select a deformation that will better reveal or highlight the unique aspects of these vein maps. For example, the authentication platform can analyze some or all vein maps included in a biometric database to identify sufficiently unique features. These features may relate to the spatial relationships between different blood vessels (e.g., unusual branching locations or unusual sizes), or they may relate to the vascular properties of the vessels (e.g., if the velocity, volume, or pressure of blood flowing through the venous network changes more or less than the average after deformation).

[0126] Alternatively or additionally, the authentication platform may utilize a system to deliver a request for an execution posture to an unknown person in a manner that reveals unique characteristics (or otherwise prompts, provokes, or induces deformation). For example, the authentication platform may instruct the unknown person to apply pressure along a specific location on the palm of their hand to inhibit or block blood flow to a particular blood vessel, thereby uniquely affecting the venous resistance of other vessels flowing into that particular vessel. The effect of deformation on pressure pulses emitted through anatomical regions (and thus on image data visually capturing pressure pulses) may be located in the same place as the deformation, or the effect may be at a distance from the location of the deformation.

[0127] As described above, as part of the authentication process, the authentication platform can generate measured flow patterns and measured deformation flow patterns. In some embodiments, the authentication platform determines the differences between the measured flow patterns and the measured deformation flow patterns, and then compares these differences with other examples to ensure that the changes in image data generated by deformation are also sufficiently unique.

[0128] B. Measurement of flow pattern matching

[0129] Once the measured flow patterns and deformation flow patterns of unknown individuals are known, the authentication platform can authenticate them solely based on the matching of the measured flow patterns, since a prior association with the authentication factors of the original vein map has been established. This "lightweight" authentication process may only be suitable for certain situations, such as those involving minimal sensitive information or actions. However, if time or computational resources are limited, for example, this "lightweight" authentication process can be useful for quickly authenticating unknown individuals.

[0130] C. Multiple deformations

[0131] As described above, a given individual may be prompted to perform multiple poses during the registration phase. This approach offers a significant security advantage because there are multiple authentication options. If an unknown person seeks to be authenticated as a given individual, the authentication platform can request that unknown person to perform any combination of the poses performed by the given individual during the registration phase. Thus, the authentication platform can request an unknown person to perform several different poses during the authentication process, and the platform can only authenticate the unknown person as the given individual if a predetermined percentage (e.g., more than 50%, exactly 100%) of these poses results in a match with the given individual.

[0132] The authentication platform can also request an unknown person to perform the same pose more than once. For example, the authentication platform can request an unknown person to perform a single pose more than once during the authentication process, and the authentication platform can only authenticate the unknown person as a given individual if a predetermined percentage of these performances (e.g., more than 50%, exactly 100%) results in a match with the given individual.

[0133] In embodiments where the authentication platform allows individuals to perform multiple poses during the registration phase, the platform can manage separate biometric databases for these different poses. For example, the platform could manage a first biometric database including information about a first pose (e.g., vein maps and deformed vein maps), a second biometric database including information about a second pose, and so on. Alternatively, the platform could store information associated with different poses in different portions of a single biometric database.

[0134] Furthermore, entries in biometric databases can be associated not only with the names or identifiers that identify the corresponding individuals (e.g., email addresses or phone numbers), but also with tags that identify the corresponding postures. Therefore, different postures (e.g., smiling and frowning) can be associated with different tags, which can be attached to entries in the biometric database.

[0135] In some embodiments, appropriate labeling is identified based on analysis of image data used for authentication. For example, if the image data includes digital images of faces, the authentication platform can examine these digital images to determine which pose has been performed. Automated analysis of image data can be useful in several ways. First, the authentication platform may be able to infer which pose an unknown person performed, rather than explicitly instructing the unknown person to perform a pose. Second, the authentication platform may be able to establish appropriate vein maps to retrieve from a biometric database. For example, if the authentication platform determines that an unknown person performed a given pose in a digital image, then the authentication platform can retrieve a vein map associated with that given pose from a biometric database.

[0136] II. Benefits of Authentication of Unknown Individuals Through Remote Vascular Studies

[0137] A. Changes due to aging and the environment

[0138] While it's understandable that the shape of blood vessels generally doesn't change, vascular properties (such as flow rate) can be affected by factors such as age and disease. Environmental factors, such as temperature and humidity, can also influence vascular properties. For example, blood vessels may constrict due to low temperatures. Blood flow can also be affected by physiological factors (such as stress and tension) and physiological activities (such as exercise).

[0139] The design of the authentication platform can make it robust to these changes in vascular properties. An important aspect of the platform is its focus on the local spatial properties and directionality of blood flow due to deformation. Therefore, the effects of these factors (which often affect the entire body) are usually negligible or manageable (e.g., through modeling). For example, in states of tension and relaxation, the directional patterns of blood flow through venous networks in anatomical regions due to postural changes will be observable, although the absolute intensity of the signal (e.g., determined by analyzing image data) may differ.

[0140] Variations in global blood flow can introduce some noise into individual measurements. However, this effect has been well addressed in recent studies showing that blood flow remains readily observable even after physical activity (e.g., exercise).

[0141] B. Robustness in challenging scenarios

[0142] Ongoing research has improved the accuracy and robustness of establishing or monitoring vascular dynamics through the analysis of digital images. There is also growing interest in remote monitoring, particularly methods utilizing readily available electronic devices such as mobile phones and tablets.

[0143] In scenarios where changes in the health of an unknown user affect spatial properties within a localized area (e.g., due to injury, stroke, etc.), authentication platforms can employ modeling techniques to account for these changes. For example, if the authentication platform detects deformation based on analysis of image data related to an unknown person, it can apply an ML-based model designed to adjust the vein map accordingly. Thus, the platform may be able to intelligently manipulate the vein map to account for changes in the health of these individuals after they have completed the registration phase. As another example, if the authentication platform detects a high heart rate based on analysis of image data related to an unknown person, it can apply an ML-based model to determine appropriate adjustments to vascular properties, such as flow rate or pressure. However, these types of adjustments are not expected to be widely needed, as spatial information and vascular properties tend to remain fairly consistent over time.

[0144] Processing system

[0145] Figure 12 The block diagram illustrates an example of a processing system 1200, in which at least some of the operations described herein can be implemented. For example, components of the processing system 1200 may be housed in an electronic device including an image sensor. As another example, components of the processing system 1200 may be housed in an electronic device that includes an authentication platform responsible for inspecting image data generated by the image sensor.

[0146] Processing system 1200 may include processor 1202, main memory 1206, non-volatile memory 1210, network adapter 1212 (e.g., network interface), video display 1218, input / output device 1220, control device 1222 (e.g., keyboard, pointing device, or mechanical input such as buttons), drive unit 1224 including storage medium 1226, or signal generation device 1230, all communicatively connected to bus 1216. Bus 1216 is illustrated as an abstract concept, representing one or more physical buses and / or point-to-point connections connected by appropriate bridges, adapters, or controllers. Therefore, bus 1216 may include a system bus, a Peripheral Component Interconnect (PCI) bus, a High Speed ​​PCI bus, a HyperTransport bus, an Industry Standard Architecture (ISA) bus, a Small Computer System Interface (SCSI) bus, a Universal Serial Bus (USB), and an Inter-Integrated Circuit (I2C) bus. 2 C) Bus, or a bus conforming to Institute of Electrical and Electronics Engineers (IEEE) standard 1394.

[0147] Processing system 1200 may share a similar computer processor architecture with computer servers, routers, desktop computers, tablet computers, mobile phones, video game consoles, wearable electronic devices (e.g., watches or fitness trackers), network-connected (“smart”) devices (e.g., televisions or home assistant devices), augmented or virtual reality systems (e.g., head-mounted displays), or other electronic devices capable of executing a set of instructions (sequential or otherwise) specifying actions to be taken by processing system 1200.

[0148] Although main memory 1206, non-volatile memory 1210, and storage medium 1226 are shown as a single medium, the terms "storage medium" and "machine-readable medium" should be understood to include a single medium or multiple media storing one or more sets of instructions 1228. The terms "storage medium" and "machine-readable medium" should also be understood to include any medium capable of storing, encoding, or carrying a set of instructions for execution by processing system 1200.

[0149] Generally, routines executed to implement embodiments of the present disclosure can be implemented as part of an operating system or a particular application, component, program, object, module, or sequence of instructions (collectively, a "computer program"). A computer program typically includes one or more instructions (e.g., instructions 1204, 1208, 1228) that are set in various memories and storage devices of a computing device at various times. When read and executed by processor 1202, the instructions cause processing system 1200 to perform operations to execute various aspects of the present disclosure.

[0150] Although embodiments have been described in the context of a full-featured computing device, those skilled in the art will understand that various embodiments can be distributed as a program product in various forms. This disclosure applies regardless of the specific type of machine or computer-readable medium in which the distribution is actually caused. Further examples of machine and computer-readable media include recordable media such as volatile memory devices, non-volatile memory devices 1210, removable disks, hard disk drives, optical discs (e.g., compact disk read-only memory (CD-ROM) and digital versatile disks (DVD)), cloud-based storage, and transport media such as digital and analog communication links.

[0151] Network adapter 1212 enables processing system 1200 to perform data mediation with entities outside processing system 1200 in network 1214 via any communication protocol supported by processing system 1200 and external entities. Network adapter 1212 may include a network adapter card, wireless network interface card, switch, protocol converter, gateway, bridge, hub, receiver, repeater, or transceiver including integrated circuits (e.g., enabling communication via Bluetooth or Wi-Fi).

[0152] Notes

[0153] The foregoing description of various embodiments of the claimed subject matter is provided for illustrative and descriptive purposes. It is not intended to be exhaustive or to limit the claimed subject matter to the precise forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to best illustrate the principles of the invention and its practical application, thereby enabling those skilled in the art to understand the claimed subject matter, the various embodiments, and the various modifications suitable for the particular intended use.

[0154] While the “Detailed Description” section describes certain embodiments and contemplated best modes, the technology can be implemented in many ways, however detailed the “Detailed Description” section may appear. Embodiments can vary considerably in their implementation details, yet are still covered by this specification. Specific terms used in describing certain features or aspects of the various embodiments should not be construed as implying that such terms are hereby redefined as limited to any specific characteristic, feature, or aspect of the technology associated with that term. Generally, the terms used in the following claims should not be construed as limiting the technology to the specific embodiments disclosed in the specification, unless such terms are expressly defined herein. Therefore, the actual scope of this technology covers not only the disclosed embodiments but also all equivalent ways of implementing or carrying out the embodiments.

[0155] The language used in this specification has been chosen primarily for readability and guidance purposes. It may not have been chosen to be limiting or restrictive of the subject matter. Therefore, it is intended that the scope of this technology not be limited by this "Detailed Description" section, but rather by any claims granted under the application based thereon. Thus, the disclosure of various embodiments is intended to be illustrative, and not to limit the scope of the technology recited in the claims.

[0156] Practicality of work

[0157] This disclosure can be applied to biometric authentication in computer security.

Claims

1. An authentication method, comprising: Receive input indicating a request for authentication of a person possessing an electronic device, including a camera; Obtain a digital profile associated with a given individual as claimed by the person; While the camera of the electronic device is oriented toward the anatomical region, instructions for performing a posture are presented to the person, the posture causing deformation of the anatomical region; Based on the analysis of the digital images generated by the camera, (i) a first flow pattern of blood in the anatomical region of the person in a natural state and (ii) a second flow pattern of blood in the anatomical region of the person in a deformed state are estimated. Based on the digital profile and the first flow pattern, a third flow pattern is predicted to indicate blood flow through the anatomical region of the given individual if the given individual will perform the posture; and Based on a comparison between the second flow pattern and the third flow pattern, it is determined whether to authenticate the person as the given individual. The digital profile includes (i) a first vascular pattern that provides spatial information about the blood vessels in the anatomical region when the anatomical region is in a natural state, and (ii) a second vascular pattern that provides spatial information about the blood vessels in the anatomical region when the anatomical region is in a deformed state. The predictions include: An algorithm is applied to the first vascular pattern, the second vascular pattern, and the first flow pattern, which generates the third flow pattern as output by simulating blood flow through the blood vessels during deformation of the anatomical region caused by the posture.

2. The authentication method as described in claim 1, wherein, The estimation includes: The digital image is examined to identify color changes that indicate blood flow through blood vessels in the anatomical region.

3. The authentication method as described in claim 1, in, The first flow pattern is represented as a first vector, wherein each element includes a value indicating estimated blood flow through a corresponding portion of the anatomical region of the person in a natural state, and The second flow pattern is represented as a second vector, where each element includes a value indicating estimated blood flow through a corresponding portion of the anatomical region of the person when in a deformed state. The third flow pattern is represented as a third vector, wherein each element includes a value indicating the estimated blood flow through a corresponding portion of the anatomical region of the given individual when in a deformed state.

4. The authentication method as described in claim 3, wherein, The determination includes: An algorithm is applied to the second and third vectors to generate scores indicating the similarity between the second and third flow patterns, and The probability that the person is the given individual is determined based on the score.

5. The authentication method as described in claim 1, wherein, The acquisition includes: Access a biometric database that stores multiple digital profiles associated with different individuals, and Based on the input, the numerical profile is selected from the plurality of numerical profiles.

6. The authentication method as described in claim 5, wherein, The input identifies the given individual claimed by the person.

7. A non-transitory medium storing instructions thereon, the instructions, when executed by a processor of an electronic device, causing the electronic device to perform an operation, the operation comprising: A recognition model is to be trained to predict blood flow through the venous network in the anatomical region during deformation of the anatomical region; get (i) A first series of vascular patterns corresponding to the anatomical regions in their natural state. (ii) A second series of vascular patterns corresponding to the anatomical regions in the deformed state. (iii) A series of flow patterns, for each vascular pattern in the first series, conveying how blood flows through that vascular pattern when the anatomical region is in the natural state, and (iv) A series of deformation flow patterns, for each vascular pattern in the second series, conveying how blood flows through that vascular pattern when the anatomical region is in the deformed state; and (i) the first series of vascular patterns, (ii) the second series of vascular patterns, (iii) the series of flow patterns, and (iv) the series of deformable flow patterns are provided as training data to the model to generate a trained model that, when applied to vascular patterns associated with a person, is able to predict blood flow through venous networks in the anatomical regions of that person.

8. The non-transient medium as described in claim 7, wherein, Each vascular pattern in the first series is associated with a corresponding vascular pattern in the second series, and wherein the corresponding vascular patterns in the first series and the second series are associated with the same individual.

9. The non-transient medium as described in claim 7, wherein, Each vascular pattern in the first series indicates the spatial relationship between subcutaneous vessels when the anatomical region is in the natural state, and each vascular pattern in the second series indicates the spatial relationship between subcutaneous vessels when the anatomical region is in the deformed state.

10. The non-transient medium as described in claim 7, wherein, Each vascular pattern in the first series is associated with a different individual.

11. The non-transient medium as described in claim 7, wherein, The anatomical region in question is the fingers.

12. The non-transient medium as described in claim 7, wherein, The anatomical region is the palm or back of the hand.

13. The non-transient medium as described in claim 7, wherein, The anatomical region in question is the face.

14. The non-transient medium as described in claim 7, wherein, The model includes a neural network with parameters that are adjusted based on the provided results.

Images