An intelligent physical evaluation system for autologous arteriovenous fistula and an evaluation method thereof

Abstract

The present application relates to the technical field of medical apparatus, in particular to a kind of autologous arteriovenous fistula intelligent physical evaluation system and evaluation method thereof;The present application includes electronic stethoscope, the left side of the electronic stethoscope is for the department of treatment, the right side of the electronic stethoscope is handle portion, the handle portion is equipped with visual examination module, auscultation module and palpation module in sequence, the visual examination module includes data collector, the auscultation module includes sound collector, and the palpation module includes blood vessel tremor collector;It also includes data integration analysis system, and the data integration analysis system includes data receiving end;The lower part of the data receiving end is diagnosis result display area;The present application can objectively evaluate the function of hemodialysis patient autologous arteriovenous fistula through simple positioning;It can provide a stable, reliable and convenient diagnosis and treatment means for patients, improve the cure rate of patients, and also timely feedback the sick area of patients needing treatment for medical staff.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, specifically to an intelligent physical assessment system and assessment method for autologous arteriovenous fistulas. Background Technology

[0002] The number of hemodialysis patients is increasing year by year. Autogenous arteriovenous fistula is the preferred vascular access for hemodialysis patients. At the same time, physical assessment plays an important role in assessing fistula maturity and function of mature fistulas, and is more cost-effective than ultrasound.

[0003] While existing handheld ultrasound devices are portable and lightweight, they have a high technical threshold and are related to the operator's technical experience, which hinders their widespread application. At the same time, current physical assessments of clinical dialysis patients rely on the experience of dialysis nurses, often resulting in significant individual variations and making it impossible to establish a standardized system.

[0004] Therefore, the existing market needs a standardized, safe, reliable and stable fistula assessment system to provide patients and medical staff with intuitive diagnostic and treatment measures. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent physical assessment system and assessment method for autogenous arteriovenous fistulas that can solve the above-mentioned problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An intelligent physical assessment system for autogenous arteriovenous fistulas includes an electronic stethoscope, with the left side of the electronic stethoscope being the examination section and the right side being the handle section, the examination section and the handle section being integrally formed.

[0008] The handle is sequentially equipped with a visual inspection module, a stethoscope module, and a palpation module. The visual inspection module includes a data collector, the stethoscope module includes a sound collector, and the palpation module includes a vasomotor collector. The treatment unit includes an infrared camera, a stethoscope end, and a palpation end corresponding to the visual inspection module, stethoscope module, and palpation module on the handle. The infrared camera, stethoscope end, and palpation end are electrically connected to the visual inspection module, stethoscope module, and palpation module, respectively. The lower part of the visual inspection module, stethoscope module, and palpation module is provided with a data transmission end. A switch button is provided on the side of the handle away from the treatment unit.

[0009] It also includes a data integration and analysis system, which comprises a data receiving end and a data output end; the lower part of the data receiving end is a diagnostic result display area, which displays visual examination results, auscultation results, palpation results, and comprehensive results; the data integration and analysis system has a built-in deep learning model, which analyzes the data collected by the visual examination module, auscultation module, and palpation module; a charging port is provided on the left side of the data integration and analysis system for charging.

[0010] Preferably, the infrared camera can be replaced by a high-resolution digital camera or mobile phone, and the hand and forearm can be photographed under standard lighting conditions using a high-resolution digital camera or mobile phone to identify whether the patient's arteriovenous fistula is clearly visible and whether the blood vessel course is natural and straight.

[0011] Preferably, the palpation module is equipped with an AI algorithm. After the infrared camera, high-resolution digital camera or mobile phone captures images under natural light, the AI ​​algorithm can identify and delineate the veins and arteries of the autologous arteriovenous fistula.

[0012] An assessment method for an intelligent physical assessment system for autogenous arteriovenous fistulas includes the following steps:

[0013] S1. The infrared camera acquires image data of the patient's forearm fistula vascular area, the auscultation end collects the fistula sound audio signals from the fistula opening, arterial puncture site, and venous puncture site, and the palpation end collects the vibration and fluctuation sensation of the fistula area.

[0014] S2. The data collector receives image data of the patient's forearm fistula vascular region acquired by an infrared camera, and processes the image data through an AI recognition algorithm; the sound collector receives audio signals of the fistula from three locations—the fistula opening, the arterial puncture site, and the venous puncture site—acquired by the auscultation end, and converts the signals into a Mel spectrogram using Fourier transform through software, which is then used to build a deep learning model; the vascular thrill collector collects the thrill and fluctuation sensations of the fistula opening acquired by the palpation end;

[0015] S3. Input the patient data information and characteristics obtained in S2 into the data integration and analysis system's built-in deep learning model. The deep learning model analyzes and obtains the visual examination results, auscultation results, palpation results, and comprehensive results. The palpation assessment estimation method of scholars in Taiwan Province only uses the palpation scoring standard.

[0016] S4. Display the results of the deep learning model analysis in the diagnostic results display area.

[0017] Preferably, the visual examination results, auscultation results, palpation results and comprehensive results are obtained by using a deep learning model, wherein the palpation results are combined with the palpation assessment estimation method of scholars from Taiwan Province; The results are then scored on a scale of 0-5, where 0 indicates complete blockage of the arteriovenous fistula and 5 indicates good function. A score of 5 is given if palpable vascular thrill is good and the fistula vessels are soft. A score of 3 is given if vascular pulsation is felt in 1 / 4 of the palpation and vascular thrill is felt in 3 / 4 of the palpation. A score of 2 is given if vascular pulsation is felt in 1 / 2 of the palpation and thrill is felt in 1 / 2 of the palpation. A score of 1 is given if there is only pulsation and no thrill. A score of 0 is given if there is no pulsation and no thrill. When the palpation result is lower than 3, a reminder is given in time.

[0019] Preferably, the vascular thrill collector can also be combined with PPG to assess vascular thrill and pulsation.

[0020] Preferably, the data integration and analysis system can also display a 3D schematic diagram of the blood vessels in the forearm of a hemodialysis patient, and can mark the abnormal areas found during visual inspection, palpation, and auscultation.

[0021] Preferably, the sounds from the auscultation module are divided into five categories: normal sounds that can be heard by the human ear, loud noises, high-frequency sounds, intermittent sounds, and whistles.

[0022] Beneficial effects

[0023] The assessment system of this invention provides objective results for the assessment of autogenous arteriovenous fistulas; moreover, the electronic stethoscope of this invention combines visual, auscultatory, and tactile functions, and can objectively assess the function of autogenous arteriovenous fistulas in hemodialysis patients through simple position assessment, providing patients with a stable, reliable, and convenient diagnostic and treatment method, improving the cure rate of patients, and also providing timely feedback to medical staff on the diseased areas that patients need treatment for.

[0024] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the electronic stethoscope structure of the present invention;

[0026] Figure 2 This is a schematic diagram of the data integration and analysis system of the present invention;

[0027] Figure 3 This is a schematic diagram of the palpation assessment estimation method of the present invention;

[0028] In the diagram: 1-Electronic stethoscope; 2-Handle; 3-Diagnosis section; 4-Infrared camera; 5-Visual inspection module; 6-Auscultation module; 7-Touch module; 8-Data transmission end; 9-Touch end; 10-Auscultation end. Detailed Implementation

[0029] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0031] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0032] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0033] like Figures 1-3 As shown, an intelligent physical assessment system for autogenous arteriovenous fistula includes an electronic stethoscope 1, with a consultation section 3 on the left and a handle section 2 on the right. The consultation section 3 and the handle section 2 are integrally formed.

[0034] The handle 2 is provided with a visual inspection module 5, an auscultation module 6, and a palpation module 7 in sequence. The visual inspection module 5 includes a data collector, the auscultation module 6 includes a sound collector, and the palpation module 7 includes a vascular thrill collector. The treatment unit 3 includes an infrared camera 4, an auscultation end 10, and a palpation end 9 corresponding to the visual inspection module 5, the auscultation module 6, and the palpation module 7 of the handle 2. The infrared camera 4, the auscultation end 10, and the palpation end 9 are electrically connected to the visual inspection module 5, the auscultation module 6, and the palpation module 7, respectively. A data transmission end 8 is provided at the lower part of the visual inspection module 5, the auscultation module 6, and the palpation module 7 for transmitting the data collected by the visual inspection module 5, the auscultation module 6, and the palpation module 7 to the data integration and analysis system. A switch button is provided on the side of the handle 2 away from the treatment unit 3. Pressing the switch will start the operation.

[0035] It also includes a data integration and analysis system, which comprises a data receiving end and a data output end. The data receiving end receives and outputs data from the visual inspection module 5, the auscultation module 6, and the palpation module 7 emitted from the data transmitting end 8 on the electronic stethoscope 1. The lower part of the data receiving end is a diagnostic result display area, which displays the visual inspection results, auscultation results, palpation results, and a comprehensive result. The data integration and analysis system has a built-in deep learning model that analyzes the data collected by the visual inspection module 5, the auscultation module 6, and the palpation module 7, and then displays the analysis results in the visual inspection result, auscultation result, palpation result, and comprehensive result areas, respectively. The data integration and analysis system also includes a charging port for charging the system.

[0036] The development of the deep learning model employed a three-fold cross-validation method. The performance of the deep learning model was evaluated through a comprehensive analysis of prediction results from an external dataset and additional test set data. The prediction results on the test set were calculated by averaging the outputs of each model. The deep learning model also included a convolutional neural network (CNN) model, which consisted of convolutional layers, pooling layers, and fully connected layers. Convolutional layers extracted feature maps from the Mel-ray spectrogram; pooling layers reduced the spatial size of the feature maps. The extracted feature maps were then flattened and passed to fully connected layers for processing; finally, the probability value was calculated using the sigmoid activation function as the model's output.

[0037] The visual inspection module 5 captures images using the infrared camera 4 and identifies whether the patient's arteriovenous fistula vessels are clearly visible and whether the vessels run naturally and straight. The captured data is then transmitted to the data collector. The module automatically measures the number and size of pseudoaneurysms, i.e., protruding aneurysms. The data is automatically uploaded to the healthcare personnel's autogenous arteriovenous fistula management platform for centralized analysis and evaluation results.

[0038] The auscultation module 6 collects blood flow sounds at three points in the AVF via the auscultation end 10 of the electronic stethoscope 1 and transmits the collected information to the sound acquisition unit. Then, it extracts the audio signal using a Mel-spectrum graph and analyzes it using a convolutional neural network to determine if the blood flow sounds are normal. In a normal autogenous arteriovenous fistula, the sound amplitude is highest near the anastomosis, gradually decreasing along the proximal direction. Stenosis of the arteriovenous fistula causes changes in the frequency domain characteristics of the acoustic signal; the greater the degree of stenosis, the greater the proportion of high-frequency sound segments in the spectrum.

[0039] Mel spectrograms are two-dimensional images used to represent audio signals and are widely used in audio recognition tasks. They can extract features from audio signals and perform frequency domain filtering on audio signals processed by time windows. By filtering out background noise, Mel spectrograms help deep learning extract features related to physiological or pathological states from audio signals, thereby improving classification accuracy.

[0040] The palpation module 7 collects sounds from three points on the arteriovenous fistula—the fistula opening, the arterial puncture site, and the venous puncture site—via the palpation end 9 of the electronic stethoscope 1, while simultaneously sensing the fistula's pulsation, thrill, and fluctuation. Pulsation refers to the vasodilation caused by the cardiac cycle, and it is directly related to the degree of stenosis in the arteriovenous fistula. When the blood supply to the fistula is insufficient, that is, when there is stenosis in the inflow tract, both the thrill and pulsation are weakened; when there is stenosis in the outflow tract, the thrill is weakened, and the pulsation is significantly enhanced.

[0041] Specifically, the infrared camera 4 can be replaced with a high-resolution digital camera or mobile phone, and take photos of the hand and forearm under standard lighting conditions using the high-resolution digital camera or mobile phone to identify whether the patient's arteriovenous fistula is clearly visible (yes=1, no=0), whether the course of the blood vessel (artery / vein) is natural and straight (yes=1, no=0), whether there is tortuosity and aneurysm-like dilation (yes=1, no=0), whether the skin is red and swollen (yes=1, no=0), induration (yes=1, no=0), ulceration (yes=1, no=0), whether the hands are pale (yes=1, no=0), and whether there is swelling (yes=1, no=0).

[0042] Specifically, the palpation module 7 incorporates an AI algorithm. After images are captured by an infrared camera, a high-resolution digital camera, or a mobile phone under natural light, the AI ​​algorithm can identify and delineate the veins and arteries of the autogenous arteriovenous fistula. The AI ​​algorithm's identification indicators are as follows:

[0043] (1) Check for local dilation or narrowing of blood vessels, and whether there are obvious bulges or collapses. If there are bulges or collapses, measure their number and size;

[0044] (2) Whether there are one or more collateral circulations, and whether there are vessels long enough for puncture;

[0045] (3) Whether there is redness, swelling, hardening, ulceration, etc. on the local skin.

[0046] An evaluation method for an autogenous arteriovenous fistula intelligent physical system includes the following steps:

[0047] S1. The infrared camera 4 collects image data of the fistula area in the patient's forearm, the auscultation end 10 collects the sound audio signals of the fistula at the fistula opening, arterial puncture site and venous puncture site, and the palpation end 9 collects the vibration and fluctuation sensation of the fistula area.

[0048] S2. The data collector receives image data of the patient's forearm fistula vascular area collected by the infrared camera 4, and processes the image data through an AI recognition algorithm; the sound collector receives audio signals of the fistula from three locations—the fistula opening, the arterial puncture site, and the venous puncture site—collected by the auscultation end 10, and converts the signals into a Mel spectrogram using Fourier transform through software, which is then used to build a deep learning model; the vascular thrill collector collects the thrill and fluctuation sensation of the fistula opening collected by the palpation end 9.

[0049] S3. Input the patient data and features obtained in S2 into the deep learning model built into the data integration and analysis system, and analyze the data through the deep learning model to obtain the visual examination results, auscultation results, palpation results and comprehensive results;

[0050] S4. Display the results of the deep learning model analysis in the diagnostic results display area.

[0051] Specifically, deep learning models are used to analyze and obtain visual examination results, auscultation results, palpation results, and comprehensive results. Among them, the palpation results are combined with the palpation assessment estimation method of scholars from Taiwan Province, which only uses the palpation scoring criteria.

[0052] Then the results are scored, such as Figure 3 The scoring system is 0-5 points, where 0 points indicates that the arteriovenous fistula is completely blocked and 5 points indicates that the arteriovenous fistula is functioning well.

[0053] If palpation reveals a good vascular thrill and the fistula vessels are soft, the score is 5. If vascular pulsation is felt in 1 / 4 of the palpation and vascular thrill is felt in 3 / 4 of the palpation, the score is 3. If vascular pulsation is felt in 1 / 2 of the palpation and thrill is felt in 1 / 2 of the palpation, the score is 2. If only pulsation is felt and there is no thrill, the score is 1. If there is no pulsation or thrill, the fistula is blocked, and the score is 0. When the palpation result is lower than 3, a reminder should be given promptly.

[0054] Specifically, the vascular thrill collector can also be combined with PPG to assess vascular thrill and pulsation.

[0055] Specifically, the data integration and analysis system can also display a 3D schematic diagram of the blood vessels in the forearm of a hemodialysis patient, marking any abnormalities detected by visual inspection, palpation, and auscultation. The system then aggregates the visual inspection, palpation, and auscultation data to obtain a comprehensive result, which is then used for a holistic evaluation.

[0056] Specifically, the auscultation criteria are as follows: Using a deep learning classifier, sounds identified as originating from arteriovenous fistulas (AVCs) with a probability exceeding 50% are considered to represent the pulsation of an AVC. These sounds are categorized into five types: normal sounds audible to the human ear, loud noises, high-frequency sounds, intermittent sounds, and whistles. After human auditory identification and classification, almost no abnormal sounds, including whistles or intermittent sounds, are found among the sounds from these AVCs. Whistles are actually caused by turbulence in blood flow due to a sudden narrowing of the blood vessel diameter; intermittent sounds indicate discontinuity in the noise generated by the blood vessel during diastole, a phenomenon usually indicating severe obstruction leading to a complete interruption of blood flow during diastole. High-frequency sounds, loud noises, and normal sounds are commonly seen in routine medical examinations: high-frequency sounds indicate that the noise generated by the blood vessel is in the high-frequency range, which usually means that the diameter of the shunt vessel is relatively small over a long distance; loud noises are caused by a significant increase in the peak value of the noise generated by the blood vessel.

[0057] In summary

[0058] This invention transmits data between an electronic stethoscope 1 and a data integration and analysis system, and then analyzes the collected data through a deep learning model. By collecting data from the patient's autologous arteriovenous fistula through palpation, inspection, and auscultation, it comprehensively and objectively assesses the fistula function, providing medical staff with intuitive scoring results for the patient. This allows medical staff to understand the patient's current fistula status, make appropriate medical plans based on the patient's fistula condition, reduce medical risks, and improve the patient's cure rate.

[0059] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0060] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A smart physical assessment system for autogenous arteriovenous fistulas, characterized in that, The device includes an electronic stethoscope (1), with a consultation section (3) on the left side and a handle section (2) on the right side. The consultation section (3) and the handle section (2) are integrally formed. The handle (2) is provided with a visual inspection module (5), an auscultation module (6) and a palpation module (7) in sequence. The visual inspection module (5) includes a data collector, the auscultation module (6) includes a sound collector, and the palpation module (7) includes a vascular thrill collector. The consultation unit (3) includes an infrared camera (4), an auscultation end (10) and a palpation end (9) corresponding to the visual inspection module (5), the auscultation module (6) and the palpation module (7) of the handle (2). The infrared camera (4), the auscultation end (10) and the palpation end (9) are electrically connected to the visual inspection module (5), the auscultation module (6) and the palpation module (7) respectively. The lower part of the visual inspection module (5), the auscultation module (6) and the palpation module (7) is provided with a data transmission end (8). The handle (2) is provided with a switch button on the side away from the consultation unit (3). It also includes a data integration and analysis system, which includes a data receiving end and a data output end; the lower part of the data receiving end is a diagnostic result display area, which displays visual examination results, auscultation results, palpation results and comprehensive results; the data integration and analysis system has a built-in deep learning model, which analyzes the data collected by the visual examination module (5), auscultation module (6) and palpation module (7); the data integration and analysis system has a charging port on the left side.

2. The intelligent physical assessment system for autogenous arteriovenous fistulas as described in claim 1, characterized in that, The infrared camera (4) can be replaced by a high-resolution digital camera or mobile phone, and the hand and forearm photos can be taken under standard lighting conditions using a high-resolution digital camera or mobile phone.

3. The intelligent physical assessment system for autogenous arteriovenous fistulas as described in claim 2, characterized in that, The palpation module (7) is equipped with an AI algorithm. After the infrared camera (4), high-resolution digital camera or mobile phone takes an image under natural light, the AI ​​algorithm can identify and delineate the veins and arteries of the autologous arteriovenous fistula.

4. An evaluation method for an intelligent physical evaluation system for autogenous arteriovenous fistulas, characterized in that, Includes the following steps: S1. The infrared camera (4) collects image data of the patient's forearm fistula vascular area, the auscultation end (10) collects the fistula sound audio signals at the fistula opening, arterial puncture site and venous puncture site, and the palpation end (9) collects the tremor and fluctuation sensation of the fistula area. S2. The data collector receives image data of the patient's forearm fistula vascular area collected by the infrared camera (4) and processes the image data through an AI recognition algorithm; the sound collector receives the sound audio signals of the fistula at the patient's fistula, arterial puncture site, and venous puncture site collected by the auscultation end (10), and converts the signals into a Mel spectrogram using Fourier transform through software, which is then used to build a deep learning model; the vascular thrill collector collects the thrill and fluctuation sensation of the fistula collected by the palpation end (9); S3. Input the patient data information and characteristics obtained in S2 into the data integration and analysis system's built-in deep learning model. The deep learning model analyzes and obtains the visual examination results, auscultation results, palpation results, and comprehensive results. The palpation assessment estimation method of scholars in Taiwan Province only uses the palpation scoring standard. S4. Display the results of the deep learning model analysis in the diagnostic results display area.

5. The assessment method of the intelligent physical assessment system for autogenous arteriovenous fistula as described in claim 4, characterized in that, The results of visual examination, auscultation, palpation and comprehensive examination were obtained by using a deep learning model. The palpation results were combined with the palpation assessment and estimation method of scholars from Taiwan Province. The results were then scored on a scale of 0-5, with 0 indicating complete blockage of the arteriovenous fistula and 5 indicating good fistula function. A score of 5 was given if palpable vascular thrill was good and the fistula vessels were soft; a score of 3 was given if vascular pulsation was felt in 1 / 4 of the palpation and vascular thrill was felt in 3 / 4 of the palpation; a score of 2 was given if vascular pulsation was felt in 1 / 2 of the palpation and thrill was felt in 1 / 2 of the palpation; and a score of 1 was given if there was only pulsation and no thrill. If there is no pulsation or thrill upon palpation, the fistula blockage score is 0; if the palpation result is below 3, a timely reminder should be given.

6. The assessment method of the intelligent physical assessment system for autogenous arteriovenous fistula as described in claim 4, characterized in that, The vascular thrill collector can also be combined with PPG to assess vascular thrill and pulsation.

7. The assessment method of the intelligent physical assessment system for autogenous arteriovenous fistula as described in claim 4, characterized in that, The data integration and analysis system can also display a 3D schematic diagram of the blood vessels in the forearm of a hemodialysis patient, and can mark the abnormal areas found during visual inspection, palpation, and auscultation.

8. The assessment method of the intelligent physical assessment system for autologous arteriovenous fistula as described in claim 4, characterized in that, The auscultation module's sounds are categorized into five types: normal sounds that the human ear can hear, loud noises, high-frequency sounds, intermittent sounds, and whistles.

Images