Vulva VIN prediction method and device based on infrared thermal imaging

Abstract

The invention belongs to the technical field of medical instruments, and particularly discloses a vulva VIN prediction method and device based on infrared thermal imaging, and the method comprises the following steps: collecting infrared thermal imaging images of the vulva parts of a plurality of VIN samples; based on an FLIR system, a diseased region and a temperature value corresponding to the diseased region are segmented and extracted from the infrared thermal imaging image; calculating a plurality of temperature characteristic values based on the extracted temperature values, and performing principal component analysis to obtain contribution coefficients of the temperature characteristic values; building a prediction model by using a support vector machine, and training the prediction model by using the temperature characteristic value with the contribution coefficient greater than a preset value; and collecting a thermal imaging image of the vulva of a patient to be predicted, inputting the thermal imaging image into the prediction model, and predicting whether the patient suffers from the VIN of the vulva. According to the technical scheme, infrared thermal imaging and machine learning are combined, and the diagnosis accuracy is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical devices and relates to a vulva VIN prediction method and device based on infrared thermal imaging. Background Art

[0002] Vulvar intraepithelial neoplasia is one of the common diseases of the external genitalia. In recent years, the incidence of vulvar precancerous lesions has shown a significant upward trend. According to the pathogenesis, vulvar intraepithelial neoplasia can be divided into two categories: related to HPV infection and unrelated to HPV infection.

[0003] HPV infection-related vulvar squamous intraepithelial lesions (VSIL) can be divided into low-grade vulvar squamous extraepithelial lesions (VLSIL) and high-grade vulvar squamous intraepithelial lesions (VHSIL). Vulvar intraepithelial neoplasia not related to HPV infection may be formed on the basis of non-tumor-like lesions and chronic skin diseases of the vulvar epithelium, which is called differentiated vulvar intraepithelial neoplasia (dVIN).

[0004] VLSIL is a transient infection with simple HPV, most of which resolve spontaneously. VHSIL and dVIN carry a significant risk of progression to invasive cancer and are considered vulvar precancerous lesions that require immediate treatment to avoid disease progression. The clinical manifestations of VHSIL are often nonspecific, and some patients may experience symptoms such as vulvar itching, burning, and dyspareunia. Signs vary, mainly including flat plaques, papules or warts, and slightly elevated pigmented lesions on the skin.

[0005] According to relevant research reports, the 10-year cancer risk of HPV-related HSIL is 8.0%. Therefore, early diagnosis of VIN (vulvar intraepithelial neoplasia) is crucial. The diagnosis of vulvar intraepithelial neoplasia can be initially clinically diagnosed based on medical history and typical symptoms and signs. Colposcopy is a commonly used diagnostic tool for lower reproductive system diseases, especially for the diagnosis of precancerous lesions and early cancers of the lower reproductive tract. However, the accuracy, sensitivity, and specificity of colposcopy are largely limited by the experience level of clinicians. Therefore, finding a new auxiliary diagnostic method is of great significance for the diagnosis of vulvar diseases. Summary of the invention

[0006] The purpose of the present invention is to provide a vulvar VIN prediction method and device based on infrared thermal imaging, combining infrared thermal imaging with machine learning to improve diagnostic accuracy.

[0007] In order to achieve the above object, the basic scheme of the present invention is: a vulva VIN prediction method based on infrared thermal imaging, comprising the following steps:

[0008] Collect infrared thermal imaging images of the vulva area of ​​multiple VIN samples;

[0009] Based on the FLIR system, the lesion area and the corresponding temperature value of the lesion area are segmented and extracted from the infrared thermal imaging image;

[0010] Based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis is performed to obtain the contribution coefficient of each temperature characteristic value;

[0011] A prediction model is constructed using a support vector machine, and the prediction model is trained using temperature feature values ​​whose contribution coefficients are greater than preset values;

[0012] A thermal imaging image of the vulva of the patient to be predicted is collected and input into the prediction model to predict whether the patient has vulvar VIN.

[0013] The working principle and beneficial effects of this basic solution are as follows: This technical solution uses infrared thermal imaging to assist in the diagnosis of VIN, collects thermal imaging images of patients with VIN, combines infrared thermal imaging with machine learning, and establishes a support vector machine model to distinguish between normal vulva and VIN, thereby improving the accuracy of diagnosis.

[0014] Furthermore, based on the FLIR system, the method for segmenting and extracting the lesion part and the temperature value corresponding to the lesion part from the infrared thermal imaging image is as follows:

[0015] Divide the vulva into five parts: mons pubis, clitoris, labia minora, labia majora and posterior symphysis, obtain their normal temperature values, and establish a normal vulva temperature baseline;

[0016] The lesion area of ​​the infrared thermal imaging image of each sample is segmented as follows:

[0017] The FLIR system was used to extract three uniform 3x3 pixel points from the lesion area of ​​each sample, and the average value was taken as the temperature value of the lesion site. Three points on the normal vulva area 2 cm away from the lesion area of ​​the sample were extracted, and the average value was taken as the control group, representing the temperature value of the normal area.

[0018] The operation is simple and convenient for subsequent use.

[0019] Furthermore, based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis is performed to obtain the contribution coefficient of each temperature characteristic value. The specific steps are as follows:

[0020] Use the temperature values ​​to calculate seven temperature characteristics: entropy, maximum-minimum, mean, standard deviation, coefficient of variation, skewness, and kurtosis:

[0021]

[0022] Maximum–MinimumT max -T min

[0023] Mean

[0024] Standard Deviation

[0025] Coefficient of Variation

[0026] Skewness

[0027] Kurtosis

[0028] Among them, Entropy H(T) is entropy; Maximum–Minimum Tmax-Tmin is the maximum-minimum value; MeanTmean is the average value; Standard Deviationσ T is the standard deviation; Coefficient of Variation cv is the coefficient of variation; Skewness Sk is the skewness; Kurtosis k is the kurtosis; N represents the number of temperature values, T i represents the temperature value of the i-th point, p(T i ) refers to the probability of each temperature value occurring.

[0029] Use the temperature eigenvalues ​​with larger contribution coefficients to train the model and optimize the model.

[0030] Furthermore, the sensitivity, specificity, accuracy, AUC, positive predictive value (PPV) and negative predictive value (NPV) of the prediction model are collected, and the collected data values ​​are compared with the preset corresponding value ranges to obtain the comparison difference, and the prediction performance of the prediction model is evaluated based on the comparison difference.

[0031] Evaluate the various data of the prediction model to determine the calculation performance of the prediction model so as to optimize the prediction model in the future and to determine the reliability of the prediction model.

[0032] The present invention also provides a vulva VIN prediction device, comprising a bed, an infrared thermal imager, an auxiliary mechanism, a human-computer interaction module and a processing module;

[0033] The bed is provided with a depression, and when the patient lies flat on the bed, his buttocks are located in the depression;

[0034] The infrared thermal imager is arranged above the bed and faces the recessed position;

[0035] The auxiliary mechanism includes a pedal and an angle sensor. The pedal is tilted and located on both sides of the upper surface of the bed. When the patient lies flat on the bed, the soles of his feet contact the pedal. The side of the pedal away from the patient is connected to a telescopic mechanism that controls its horizontal linear movement.

[0036] The angle sensor is arranged on the side of the bed, and is used to collect the bending angle of the patient's legs. The output end of the angle sensor is connected to the first input end of the angle comparator, the second input end of the angle comparator is connected to the angle threshold memory, and the output end of the angle comparator is connected to the control end of the telescopic mechanism.

[0037] The output end of the angle comparator is connected to the start end of the infrared thermal imager after passing through a NOT gate. The output end of the infrared thermal imager is respectively connected to the human-computer interaction module and the processing module. The processing module executes the method described in the present invention to predict whether the patient suffers from vulvar VIN. The processing module is connected to the human-computer interaction module. The human-computer interaction module and the processing module are installed on the side of the bed.

[0038] When the patient lies flat on the bed, his buttocks are located in the depression, and the infrared thermal imager is facing the depression for easy positioning, which is conducive to the infrared thermal imager to collect accurate data.

[0039] Since the infrared thermal imager collects thermal images of the patient's vulva, the patient needs to expose the perineum, so the patient's legs need to be bent at a certain angle so as not to block the perineum. Pedals are set on both sides of the bed. When the patient lies flat on the bed, his feet are in contact with one end surface of the pedals. The telescopic mechanism controls the movement of the pedals to guide the patient's legs to bend. At the same time, the pedals can also support the patient's feet to prevent the patient from changing the angle by mistake when bending his legs.

[0040] The angle sensor collects the patient's leg bending angle information and transmits it to the angle comparator. The angle comparator compares the collected angle signal with the angle threshold in the angle threshold memory. If the collected angle signal is greater than the angle threshold, it means that the patient's leg bending degree is not enough. At this time, the angle comparator outputs a control signal to the control end of the telescopic mechanism, and the telescopic mechanism starts to control the pedal to move to the side close to the patient's buttocks. The telescopic mechanism stops until the collected angle signal is less than or equal to the angle threshold.

[0041] At the same time, the output end of the angle comparator is connected to the start end of the infrared thermal imager after passing through a NOT gate. That is, when the collected angle signal is less than or equal to the angle threshold, it means that the patient's leg curvature is appropriate. At this time, the output end of the angle comparator outputs a control signal to the infrared thermal imager after passing through a NOT gate, starting the infrared thermal imager for thermal imaging collection.

[0042] After the infrared thermal imager collects data, it is input into the human-computer interaction module and the processing module, displayed in the human-computer interaction module, and analyzed and processed by the processing module to determine whether the patient has vulvar VIN, and the prediction result is input into the human-computer interaction module for display, which is convenient for viewing.

[0043] Furthermore, it also includes a plurality of pressure sensors and pressure comparators corresponding to the pressure sensors one by one;

[0044] The pressure sensor is arranged in the depression of the bed and on the side of the bed where the depression is away from the pedal. When the patient lies flat on the bed, pressure is applied to the pressure sensor.

[0045] The first input end of the pressure comparator is connected to the output end of the pressure sensor, the second input end of the pressure comparator is connected to the pressure threshold memory, and the output end of the pressure comparator is connected to the start end of the angle sensor through an AND gate.

[0046] A pressure sensor is set to detect the pressure signal between the patient and the bed, and transmit it to the corresponding pressure comparator to determine whether the patient is lying correctly on the bed. The pressure comparator compares the collected pressure signal with the pressure threshold in the pressure threshold memory. If the pressure signal value collected by all pressure sensors is greater than or equal to the pressure threshold in the corresponding pressure threshold memory, it means that the patient is lying properly. At this time, the AND gate outputs a control signal to the angle sensor to start the angle sensor, thereby starting the telescopic mechanism to adjust the angle of the patient's legs. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic flow chart of the vulva VIN prediction method based on infrared thermal imaging of the present invention;

[0048] Figure 2 It is a schematic diagram of the structure of the vulva VIN prediction device of the present invention.

[0049] The reference numerals in the drawings of the specification include: bed 1, infrared thermal imager 2, pedal 3, angle sensor 4, telescopic mechanism 5, depression 6, pressure sensor 7. DETAILED DESCRIPTION

[0050] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0051] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0052] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0053] The present invention discloses a method for predicting vulva VIN based on infrared thermal imaging, such as Figure 1 As shown, the following steps are included:

[0054] Collect infrared thermal imaging images of the vulva area of ​​multiple VIN (vulvar intraepithelial neoplasia) samples;

[0055] Based on the FLIR system, the lesion area and the corresponding temperature value of the lesion area are segmented and extracted from the infrared thermal imaging image;

[0056] Based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis is performed to obtain the contribution coefficient of each temperature characteristic value;

[0057] A prediction model is constructed using a support vector machine, and the prediction model is trained using temperature feature values ​​whose contribution coefficients are greater than preset values;

[0058] A thermal imaging image of the vulva of the patient to be predicted is collected and input into the prediction model to predict whether the patient has vulvar VIN.

[0059] In a preferred embodiment of the present invention, based on the FLIR system, a method for segmenting and extracting the lesion part and the temperature value corresponding to the lesion part from the infrared thermal imaging image is as follows:

[0060] The temperature distribution diagram was drawn using the extracted normal vulva temperature values, and the vulva was divided into five parts: mons pubis, clitoris, labia minora, labia majora, and posterior commissure, and their normal temperature values ​​were obtained, which helped to clarify the distribution of normal vulva temperature and establish a normal vulva temperature baseline. The temperatures of normal vulva parts from low to high were: clitoris (33.14±1.15℃), labia minora (33.47±1.56℃), mons pubis (33.51±1.86℃), and labia majora (33.37±1.08℃), with the posterior commissure having the highest temperature (34.70±1.44℃).

[0061] The lesion area of ​​the infrared thermal imaging image of each sample is segmented as follows:

[0062] The FLIR system was used to extract three uniform 3x3 pixel points from the lesion area of ​​each sample, and the average value was taken as the temperature value of the lesion site. Three points on the normal vulva area 2 cm away from the lesion area of ​​the sample were extracted, and the average value was taken as the control group, representing the temperature value of the normal area.

[0063] In a preferred embodiment of the present invention, based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis (PCA) is performed to obtain the contribution coefficient of each temperature characteristic value. The specific steps are:

[0064] Use the temperature values ​​to calculate seven temperature characteristics: entropy, maximum-minimum, mean, standard deviation, coefficient of variation, skewness, and kurtosis:

[0065] Entropy

[0066] Maximum–Minimum T max -T min

[0067] Mean

[0068] Standard Deviation

[0069] Coefficient of Variation

[0070] Skewness

[0071] Kurtosis

[0072] Among them, Entropy H(T) is entropy; Maximum–Minimum Tmax-Tmin is the maximum-minimum value; MeanTmean is the average value; Standard Deviationσ T is the standard deviation; Coefficient of Variation cv is the coefficient of variation; Skewness Sk is the skewness; Kurtosis k is the kurtosis; N represents the number of temperature values, T i represents the temperature value of the i-th point, p(T i ) refers to the probability of each temperature value occurring.

[0073] In a preferred embodiment of the present invention, the sensitivity, specificity, accuracy, AUC (area under the receiver operating characteristic curve), positive predictive value (PPV) and negative predictive value (NPV) of the prediction model are collected, the collected data values ​​are compared with the preset corresponding numerical ranges, the comparison difference is obtained, and the prediction performance of the prediction model is evaluated according to the comparison difference.

[0074] Evaluate the various data of the prediction model to determine the calculation performance of the prediction model so as to optimize the prediction model in the future and to determine the reliability of the prediction model.

[0075] The present invention also provides a vulva VIN prediction device, such as Figure 2 As shown, it includes a bed 1, an infrared thermal imager 2, an auxiliary mechanism, a human-computer interaction module and a processing module. A depression 6 is provided on the bed 1. When the patient lies flat on the bed 1, his buttocks are located in the depression 6. The infrared thermal imager 2 is arranged above the bed 1 through a structure such as a rack and faces the depression 6. The infrared thermal imager 2 can also be installed on the rack through a universal pan-tilt head to facilitate the adjustment of the orientation.

[0076] The auxiliary mechanism includes a pedal 3 and an angle sensor 4. The pedal 3 is tilted and located on both sides of the upper surface of the bed 1. When the patient lies flat on the bed 1, the soles of his feet contact the pedal 3. The side of the pedal 3 away from the patient is connected to a telescopic mechanism 5 for controlling its horizontal linear movement, and the pedal 3 moves linearly along the edge of the bed 1. The telescopic mechanism 5 can be a cylinder, a hydraulic cylinder or the like. The telescopic mechanism 5 can be installed on the bed 1. The telescopic end of the telescopic mechanism 5 is fixedly connected to the end of the pedal 3 that is not in contact with the patient's feet (such as welding, bonding, riveting, etc.).

[0077] An angle sensor 4 (such as an optical fiber angle sensor 4 designed to capture the human body joint angle, through structural design and theoretical analysis, a mathematical model of optical power at the optical fiber output end and joint bending angle is established) is arranged on the side of the bed 1 to collect the patient's leg bending angle. The output end of the angle sensor 4 is electrically connected to the first input end of the angle comparator, the second input end of the angle comparator is electrically connected to the angle threshold memory, and the output end of the angle comparator is electrically connected to the control end of the telescopic mechanism 5.

[0078] The output end of the angle comparator is electrically connected to the start end of the infrared thermal imager 2 after passing through a NOT gate. The output end of the infrared thermal imager 2 is electrically connected to a human-computer interaction module (such as a touch screen, a computer, etc.) and a processing module respectively. The processing module executes the method described in the present invention to predict whether the patient suffers from vulvar VIN. The processing module is electrically connected to the human-computer interaction module. The human-computer interaction module and the processing module are installed on the side of the bed 1.

[0079] When the patient lies flat on the bed 1 , his buttocks are located in the depression 6 , and the infrared thermal imager 2 faces the depression 6 for easy positioning, which is beneficial for the infrared thermal imager 2 to collect accurate data.

[0080] Since the infrared thermal imager 2 collects thermal imaging images of the patient's vulva, the patient needs to expose the perineum, so the patient's legs need to be bent at a certain angle so as not to block the perineum. Pedals 3 are set on both sides of the bed 1. When the patient lies flat on the bed 1, his feet are in contact with one end surface of the pedals 3. The telescopic mechanism 5 controls the movement of the pedals 3 to guide the patient's legs to bend. At the same time, the pedals 3 can also support the patient's feet to prevent the patient from changing the angle by mistake when bending his legs.

[0081] The angle sensor 4 collects the patient's leg bending angle information and transmits it to the angle comparator. The angle comparator compares the collected angle signal with the angle threshold in the angle threshold memory. If the collected angle signal is greater than the angle threshold, it means that the patient's leg bending degree is insufficient. At this time, the angle comparator outputs a control signal to the control end of the telescopic mechanism 5, and the telescopic mechanism 5 starts to control the pedal 3 to move to the side close to the patient's buttocks. The telescopic mechanism 5 stops until the collected angle signal is less than or equal to the angle threshold.

[0082] At the same time, the output end of the angle comparator is connected to the start end of the infrared thermal imager 2 after passing through the NOT gate. That is, when the collected angle signal is less than or equal to the angle threshold, it means that the patient's leg curvature is appropriate. At this time, the output end of the angle comparator outputs a control signal to the infrared thermal imager 2 after passing through the NOT gate, starts the infrared thermal imager 2, and performs thermal imaging collection.

[0083] After the infrared thermal imager 2 collects data, it is input into the human-computer interaction module and the processing module, displayed in the human-computer interaction module, and analyzed and processed by the processing module to determine whether the patient has vulvar VIN, and the prediction result is input into the human-computer interaction module for display, which is easy to view.

[0084] In a preferred embodiment of the present invention, the vulva VIN prediction device further includes a plurality of pressure sensors 7 and pressure comparators corresponding to the pressure sensors 7. The pressure sensors 7 are arranged (such as embedded, etc.) in the depression 6 of the bed body 1 and on the bed body 1 on the side of the depression 6 away from the pedal 3. When the patient lies flat on the bed body 1, pressure is applied to the pressure sensors 7.

[0085] The first input end of the pressure comparator is electrically connected to the output end of the pressure sensor 7, the second input end of the pressure comparator is electrically connected to the pressure threshold memory, and the output end of the pressure comparator is electrically connected to the start end of the angle sensor 4 through the AND gate.

[0086] A pressure sensor 7 is provided to detect the pressure signal between the patient and the bed 1, and transmit it to the corresponding pressure comparator, so as to judge whether the patient lies correctly on the bed 1. The pressure comparator compares the collected pressure signal with the pressure threshold in the pressure threshold memory. If the pressure signal values ​​collected by all pressure sensors 7 are greater than or equal to the pressure threshold in the corresponding pressure threshold memory, it means that the patient lies properly. At this time, the AND gate outputs a control signal to the angle sensor 4, starts the angle sensor 4, and thus starts the telescopic mechanism 5 to adjust the angle of the patient's legs.

[0087] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0088] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A vulva VIN prediction method based on infrared thermal imaging, characterized in that: The steps include: Collect infrared thermal imaging images of the vulva area of ​​multiple VIN samples; Based on the FLIR system, the lesion area and the corresponding temperature value of the lesion area are segmented and extracted from the infrared thermal imaging image; Based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis is performed to obtain the contribution coefficient of each temperature characteristic value; A prediction model is constructed using a support vector machine, and the prediction model is trained using temperature feature values ​​whose contribution coefficients are greater than preset values; A thermal imaging image of the vulva of the patient to be predicted is collected and input into the prediction model to predict whether the patient has vulvar VIN.

2. The method for predicting vulva VIN based on infrared thermal imaging according to claim 1, characterized in that: Based on the FLIR system, the method for segmenting and extracting the lesion part and the temperature value corresponding to the lesion part from the infrared thermal imaging image is as follows: Divide the vulva into five parts: mons pubis, clitoris, labia minora, labia majora and posterior commissure, obtain their normal temperature values, and establish a normal vulva temperature baseline; The lesion area of ​​the infrared thermal imaging image of each sample is segmented as follows: The FLIR system was used to extract three uniform 3x3 pixel points from the lesion area of ​​each sample, and the average value was taken as the temperature value of the lesion site. Three points on the normal vulva area 2 cm away from the lesion area of ​​the sample were extracted, and the average value was taken as the control group, representing the temperature value of the normal area.

3. The method for predicting vulva VIN based on infrared thermal imaging according to claim 1, characterized in that: Based on the extracted temperature values, multiple temperature characteristic values ​​are calculated, and principal component analysis is performed to obtain the contribution coefficient of each temperature characteristic value. The specific steps are as follows: Use the temperature values ​​to calculate seven temperature characteristics: entropy, maximum-minimum, mean, standard deviation, coefficient of variation, skewness, and kurtosis: Maximum–Minimum T max -T min Among them, Entropy H(T) is entropy; Maximum–Minimum Tmax-Tmin is the maximum-minimum value; Mean Tmean is the average value; Standard Deviationσ T is the standard deviation; Coefficient of Variation cv is the coefficient of variation; Skewness Sk is the skewness; Kurtosis k is the kurtosis; N represents the number of temperature values, T i represents the temperature value of the i-th point, p(T i ) refers to the probability of each temperature value occurring.

4. The method for predicting vulva VIN based on infrared thermal imaging according to claim 1, characterized in that: The sensitivity, specificity, accuracy, AUC, positive predictive rate and negative predictive rate of the prediction model were collected, and the collected data values ​​were compared with the preset corresponding value ranges to obtain the comparison difference, and the prediction performance of the prediction model was evaluated based on the comparison difference.

5. A vulva VIN prediction device, characterized in that: It includes a bed, an infrared thermal imager, an auxiliary mechanism, a human-computer interaction module and a processing module; The bed is provided with a depression, and when the patient lies flat on the bed, his buttocks are located in the depression; The infrared thermal imager is arranged above the bed and faces the recessed position; The auxiliary mechanism includes a pedal and an angle sensor. The pedal is tilted and located on both sides of the upper surface of the bed. When the patient lies flat on the bed, the soles of his feet contact the pedal. The side of the pedal away from the patient is connected to a telescopic mechanism that controls its horizontal linear movement. The angle sensor is arranged on the side of the bed, and is used to collect the bending angle of the patient's legs. The output end of the angle sensor is connected to the first input end of the angle comparator, the second input end of the angle comparator is connected to the angle threshold memory, and the output end of the angle comparator is connected to the control end of the telescopic mechanism. The output end of the angle comparator is connected to the start end of the infrared thermal imager after passing through a NOT gate. The output end of the infrared thermal imager is respectively connected to the human-computer interaction module and the processing module. The processing module executes the method described in one of claims 1-4 to predict whether the patient suffers from vulvar VIN. The processing module is connected to the human-computer interaction module, and the human-computer interaction module and the processing module are installed on the side of the bed.

6. The vulva VIN prediction device according to claim 5, characterized in that: Also includes a plurality of pressure sensors and pressure comparators corresponding one to one with the pressure sensors; The pressure sensor is arranged in the depression of the bed and on the side of the bed where the depression is away from the pedal. When the patient lies flat on the bed, pressure is applied to the pressure sensor. The first input end of the pressure comparator is connected to the output end of the pressure sensor, the second input end of the pressure comparator is connected to the pressure threshold memory, and the output end of the pressure comparator is connected to the start end of the angle sensor through an AND gate.

Images