Handheld spatial field imager and method for chronic wound identification

Abstract

The invention discloses a handheld space field imager and method for chronic wound identification, and relates to the technical field of chronic wound treatment. According to the invention, highly structured biological tissues can be imaged at the terminal, and optical parameters and physiological parameters of each layer of a double-layer tissue structure are obtained, so that medical care and researchers can more conveniently and quickly know the states of the skin and mucous membranes of a patient; by monitoring the thickness of the upper layer, the content of oxygen-containing and hypoxia hemoglobin and the content of melanin in real time, the method has excellent effects on early diagnosis, conventional monitoring and treatment effect monitoring of the chronic wound, and the quantification and visualization capability of tissue physiological information has great potential in non-invasive large-view evaluation of the chronic wound. A valuable unmarked and noninvasive solution is provided for early diagnosis, conventional monitoring and treatment effect monitoring of chronic wounds.

Description

Technical Field

[0001] The present invention relates to the technical field of chronic wound treatment, and in particular to a handheld space field imager and method for chronic wound identification. Background Art

[0002] Chronic wounds, also known as chronic refractory wounds, refer to wounds caused by skin tissue damage due to various reasons, and the healing process takes more than 4-8 weeks. Among them, the more common chronic wounds include pressure injuries, diabetic ulcers, arteriovenous ulcers, and poor healing after surgery. According to some studies, 1% to 2% of the world's population suffers from chronic wounds, and the cost of chronic wound treatment accounts for 2% to 4% of total medical expenses. In areas where the proportion of diabetic patients is high, the treatment expenses of chronic wounds can account for 10% of total medical expenses. In my country, due to the rising incidence of chronic diseases, the problem of chronic wounds has become increasingly prominent. The incidence of chronic wounds is between 1% and 2%, and it is increasing at a rate of 10% per year. Among them, the incidence of diabetic foot ulcers is about 5% to 15% in diabetic patients, and the incidence of pressure sores is also high in patients who are bedridden for a long time or have difficulty in moving. Therefore, chronic wounds have become a major challenge to global public health.

[0003] At present, in the clinical treatment of chronic wounds, it is difficult to heal due to a variety of reasons such as insufficient blood flow, infection, and tissue necrosis. Moreover, its surface appearance cannot fully reflect the deep pathological changes of the wound. Traditional wound assessment methods usually rely on naked eye observation or tissue biopsy, which lacks accuracy and objectivity and may cause discomfort to patients. This makes it impossible to detect obstacles to wound healing (such as infection, ischemia, and malnutrition) early, delaying appropriate treatment intervention and causing further deterioration of the wound. Therefore, we propose a handheld spatial domain imager and method for chronic wound identification. Summary of the invention

[0004] The purpose of the present invention is to solve the problems mentioned in the above background technology. The present invention provides a handheld space field imager and method for chronic wound identification.

[0005] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0006] A handheld space field imager for chronic wound identification, comprising an imaging system, wherein the imaging system comprises:

[0007] A light source unit, which serves as a light emission starting point for the entire imaging system;

[0008] A collimating and splitting unit, the collimating and splitting unit comprising a collimating lens, a grating and a projection lens, wherein the collimating lens is used to collimate the divergent light emitted by the light source, the grating is used to modulate the collimated light to generate a sinusoidal fringe pattern, the projection lens is installed on the light path after the grating modulation, and the projection lens is used to amplify the generated sinusoidal fringe pattern structured light and project it onto the sample;

[0009] The beam splitting imaging unit is used to realize beam splitting and imaging of the backscattered light of the sample.

[0010] Furthermore, the light source unit uses an RGBLED light source.

[0011] Furthermore, the light source unit is divided into an illumination light path and a collection light path, wherein the illumination light path includes: a collimating lens for collimating the light, converting the point light source form of the light source into parallel light, and irradiating the grating with a sinusoidal stripe pattern, and then passing the structured light through a projection lens, and finally projecting it into the biological tissue; the collection light path includes a beam splitter and a CCD lens for imaging, which is used to collect light after diffuse reflection from the tissue.

[0012] Furthermore, the beam splitting imaging unit includes: a thin film beam splitter and a CMOS camera, wherein the thin film beam splitter is arranged on the path of the sample backscattered light and is used to split the backscattered light according to a certain ratio, and a part of the light can be reflected or transmitted for subsequent imaging, and the CMOS camera is used to receive the light processed by the thin film beam splitter and perform imaging.

[0013] A method for using the handheld spatial domain imager for chronic wound identification according to any one of the above items comprises the following steps:

[0014] S1, light from the RGBLED light source passes through the collimating lens and grating to produce a sinusoidal fringe pattern;

[0015] S2, the structured light is magnified by the projection objective and projected onto the sample;

[0016] S3, backscattered light from the sample is imaged into the CMOS camera through a pellicle beam splitter and imaged by the SSMD-SFD I system after resting in the supine position for 5 to 10 minutes;

[0017] S4, select the region of interest with a size of 8.5 mm x 8.5 mm in the center of the image window for data analysis;

[0018] S5, using a single snapshot multi-frequency demodulation system (SSMD) and further inverting the optical parameters of the double-layer tissue according to the calculated modulation function;

[0019] S6. Get organizational information;

[0020] S7. Use support vector machine (SVM) to classify different chronic wounds such as healthy subjects, pressure injuries, and diabetic foot according to structural and physiological parameters, determine the characteristics of tissue damage and physiological structural changes in different wounds, and use interpretable neural networks to evaluate different chronic wounds.

[0021] Furthermore, the tissue information obtained in S6 includes: optical parameters such as reduced scattering coefficient and absorption coefficient of the upper and lower layers, as well as information such as the thickness of the upper layer, oxygenated and deoxygenated hemoglobin content, and melanin content.

[0022] Further, the S7 comprises the following steps:

[0023] S71, data collection, used for collecting data;

[0024] S72, data processing, used to process the collected data;

[0025] S73. Risk analysis is used to make risk assessments on the processed data.

[0026] Furthermore, the S71 data collection includes the following steps:

[0027] S711, Instrument positioning, accurately place the probe of the handheld spatial frequency domain imager at an appropriate distance above the chronic wound, in a darkroom, and according to the instrument's operating instructions to ensure that clear and accurate images can be obtained;

[0028] S712, light emission, the instrument emits light of specific frequency and intensity distribution to the wound surface. The light contains different spatial frequency components. For example, there may be a combination of multiple wavelengths of light to obtain optical property information such as absorption and scattering of different wavelengths of light by the wound tissue;

[0029] S713, reflected light collection. The wound tissue reflects the incident light. The optical sensor in the imager collects the reflected light. The reflected light carries information about the structure and physiological characteristics of the wound tissue, such as the oxygenation status of the tissue, microvascular density, melanin content and other related information.

[0030] Furthermore, the S72 data processing includes the following steps:

[0031] S721, image reconstruction, reconstructing an optical characteristic image of the wound surface using a specific convolutional neural network algorithm based on the collected reflected light data, and the optical characteristic image shows the distribution of optical parameters of different areas of the wound, such as an absorption coefficient image, a scattering coefficient image, etc.;

[0032] S722, feature extraction, extracting features related to chronic wound risk from the reconstructed image, wherein the absorption coefficient image is analyzed to determine whether there is abnormal blood absorption in the wound area, the microstructure information of the tissue is extracted from the scattering coefficient image, and the features of the wound edge are extracted, such as edge clarity, irregularity, etc.;

[0033] S723, parameter calculation; calculate quantitative parameters, such as average oxygen saturation of tissues, estimated microvascular blood flow velocity, etc.

[0034] Furthermore, the S73 risk analysis includes the following steps:

[0035] S731, comparing with normal range, comparing the extracted features and calculated parameters with the reference value range in the normal wound healing process;

[0036] S732, multi-factor comprehensive analysis, comprehensively considering multiple extracted features and parameters for risk assessment;

[0037] S733. Determine the risk classification and classify the risk of chronic wounds based on the analysis results.

[0038] The beneficial effects of the present invention are as follows:

[0039] 1. The present invention can image highly structured biological tissues at the terminal and obtain the optical parameters and physiological parameters of each layer of the double-layer tissue structure. At this time, medical staff and researchers can understand the status of the patient's skin and mucous membranes more conveniently and quickly.

[0040] 2. The present invention has excellent effects on the early diagnosis, routine monitoring and treatment effect monitoring of chronic wounds through real-time monitoring of the upper layer thickness, oxygenated and hypoxic hemoglobin content and melanin content. The quantification of tissue physiological information and the visualization capability have great potential in the non-invasive large-field evaluation of chronic wounds, providing a valuable label-free and non-invasive solution for the early diagnosis, routine monitoring and treatment effect monitoring of chronic wounds. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a working block diagram of the imaging system in the present invention;

[0042] Figure 2 It is a work flow chart of the present invention;

[0043] Figure 3 is a workflow diagram for monitoring and analyzing the risk process of chronic wounds in the present invention;

[0044] Figure 4 It is a workflow diagram of data collection in the present invention;

[0045] Figure 5 It is a work flow chart of the data processing card in the present invention;

[0046] Figure 6 It is a workflow diagram of risk analysis in the present invention. DETAILED DESCRIPTION

[0047] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0048] See also Figure 1 - Figure 6 The present invention provides a handheld space field imager for chronic wound identification, including an imaging system, the imaging system comprising:

[0049] A light source unit, which serves as a light emission starting point for the entire imaging system;

[0050] A collimating and splitting unit, the collimating and splitting unit comprises a collimating lens, a grating and a projection lens, wherein the collimating lens is used to collimate the divergent light emitted by the light source, the grating is used to modulate the collimated light to generate a sinusoidal fringe pattern, the projection lens is installed on the light path after the grating modulation, and the projection lens is used to amplify the generated sinusoidal fringe pattern structured light and project it onto the sample;

[0051] The beam splitting imaging unit is used to realize the beam splitting and imaging of the backscattered light of the sample.

[0052] In the present invention, preferably, the light source unit uses an RGBLED light source; the RGBLED light source can produce a variety of colors to meet the lighting needs of different scenes. Compared with traditional light sources, the RGBLED light source has high energy efficiency and low power consumption, which is conducive to energy conservation and emission reduction. The brightness of LEDs of different colors can be adjusted to achieve light brightness adjustment to adapt to different environments and occasions. The LED light source has a fast response speed and can be lit when the power is turned on. The RGB LED light source is small in size and light in weight, and is suitable for various display backlight and lighting applications. LED technology is already quite mature, the manufacturing process is simple, and the cost is relatively low.

[0053] In the present invention, preferably, the light source unit is divided into an illumination light path and a collection light path, wherein the illumination light path includes: a collimating lens for collimating the light, converting the point light source form of the light source into parallel light, and irradiating the grating with a sinusoidal stripe pattern, and then passing the structured light through a projection lens, and finally projecting it into the biological tissue; the collection light path includes a beam splitter and a CCD lens for imaging, which is used to collect light after diffuse reflection from the tissue; it is capable of imaging highly structured biological tissues at the terminal, and obtaining the optical parameters and physiological parameters of each layer of the double-layer tissue structure.

[0054] In the present invention, preferably, the beam splitting imaging unit includes: a thin film beam splitter and a CMOS camera, wherein the thin film beam splitter is arranged on the path of the sample backscattered light, and is used to split the backscattered light according to a certain ratio, a part of the light can be reflected or transmitted for subsequent imaging, and the CMOS camera is used to receive the light processed by the thin film beam splitter and perform imaging.

[0055] The working principle and use process of the present invention, when the device is used, it is assembled according to the following steps:

[0056] First, install the RGBLED light source, which serves as the light emission starting point of the entire imager. Its installation position needs to be relatively fixed and ensure that the light can be smoothly emitted and enter the optical path of the subsequent optical components.

[0057] A collimating lens and a grating are placed in sequence on the path of light emitted by the light source. The function of the collimating lens is to collimate the divergent light emitted by the light source to make it parallel light so as to better interact with the grating. The grating modulates the collimated light to produce a sinusoidal fringe pattern. The collimating lens and the grating must be precisely aligned during installation to ensure that the light can be correctly modulated to produce the required pattern.

[0058] A projection lens is installed on the light path after grating modulation. The projection lens amplifies the generated sinusoidal fringe pattern structured light and projects it onto the sample. The projection lens needs to maintain a suitable distance and position relationship with the collimating lens and grating in front to achieve the best magnification effect and accurate projection position.

[0059] A thin film beam splitter is set on the path of the backscattered light from the sample. The thin film beam splitter is used to split the backscattered light in a certain proportion. A part of the light can be reflected or transmitted for subsequent imaging. Then the CMOS camera is placed in a suitable position to receive the light processed by the beam splitter and perform imaging.

[0060] A method for using a handheld spatial domain imager for chronic wound identification according to any one of the above items comprises the following steps:

[0061] S1, light from the RGBLED light source passes through the collimating lens and grating to produce a sinusoidal fringe pattern;

[0062] S2, the structured light is magnified by the projection objective and projected onto the sample;

[0063] S3, backscattered light from the sample is imaged into the CMOS camera through a pellicle beam splitter and imaged by the SSMD-SFD I system after resting in the supine position for 5 to 10 minutes;

[0064] S4, select the region of interest with a size of 8.5 mm x 8.5 mm in the center of the image window for data analysis;

[0065] S5, using a single snapshot multi-frequency demodulation system (SSMD) and further inverting the optical parameters of the double-layer tissue according to the calculated modulation function;

[0066] S6. Get organizational information;

[0067] S7. Use support vector machine (SVM) to classify different chronic wounds such as healthy subjects, pressure injuries, and diabetic foot according to structural and physiological parameters, determine the characteristics of tissue damage and physiological structural changes in different wounds, and use interpretable neural networks to evaluate different chronic wounds.

[0068] In the present invention, preferably, the tissue information obtained in S6 includes: optical parameters such as the reduced scattering coefficient and absorption coefficient of the upper and lower layers, as well as information such as the thickness of the upper layer, the content of oxygenated and hypoxic hemoglobin, and the content of melanin; the hypoxic environment will hinder wound healing, affect angiogenesis, and cause the proliferation rate of various cells that promote wound healing to slow down, or even stop dividing. Secondly, hypoxia will also damage the process of wound re-epithelialization, aggravate wound infection, destroy angiogenesis, and reduce the differentiation and proliferation of fibroblasts. These changes will affect the elasticity of the skin, making the skin fragile and difficult to restore normal elasticity; chronic wounds often have melanin deposition after healing, and there is a significant color difference between the wound healing site and the normal site. This melanin deposition is to some extent a self-protection mechanism of the body, and the skin repair state can be reflected by monitoring the melanin content of the skin; through real-time monitoring of the thickness of the upper layer, the content of oxygenated and hypoxic hemoglobin, and the content of melanin, it has excellent effects on the early diagnosis, routine monitoring, and treatment effect monitoring of chronic wounds.

[0069] In the present invention, preferably, S7 comprises the following steps:

[0070] S71, data collection, used for collecting data;

[0071] S72, data processing, used to process the collected data;

[0072] S73. Risk analysis is used to make risk assessments on the processed data.

[0073] In the present invention, preferably, S71 data collection includes the following steps:

[0074] S711, Instrument positioning, accurately place the probe of the handheld spatial frequency domain imager at an appropriate distance above the chronic wound, in a darkroom, and according to the instrument's operating instructions to ensure that clear and accurate images can be obtained;

[0075] S712, light emission, the instrument emits light of specific frequency and intensity distribution to the wound surface. The light contains different spatial frequency components. For example, there may be a combination of multiple wavelengths of light to obtain optical property information such as absorption and scattering of different wavelengths of light by the wound tissue;

[0076] S713, reflected light collection. The wound tissue reflects the incident light. The optical sensor in the imager collects the reflected light. The reflected light carries information about the structure and physiological characteristics of the wound tissue, such as the oxygenation status of the tissue, microvascular density, melanin content and other related information.

[0077] In the present invention, preferably, S72 data processing includes the following steps:

[0078] S721, image reconstruction, reconstructing an optical characteristic image of the wound surface using a specific convolutional neural network algorithm based on the collected reflected light data, and the optical characteristic image shows the distribution of optical parameters of different areas of the wound, such as an absorption coefficient image, a scattering coefficient image, etc.;

[0079] S722, feature extraction, extracting features related to chronic wound risk from the reconstructed image, wherein the absorption coefficient image is analyzed to determine whether there is abnormal blood absorption in the wound area, to confirm whether there is inflammation or infection, to extract tissue microstructure information from the scattering coefficient image, such as changes in cell density, etc., to extract features of the wound edge, such as edge clarity, irregularity, etc., unclear or very irregular edges indicate that the wound is difficult to heal and there is a higher risk;

[0080] S723. Parameter calculation; calculate quantitative parameters, such as average oxygen saturation of tissues, estimated microvascular blood flow velocity, etc. Lower oxygen saturation indicates tissue hypoxia, which is not conducive to wound healing and increases the risk of chronic wounds; abnormal microvascular blood flow velocity reflects vascular dysfunction and is also a risk factor for chronic wounds.

[0081] In the present invention, preferably, S73 risk analysis includes the following steps:

[0082] S731. Compare the extracted features and calculated parameters with the reference value range in the normal wound healing process. The oxygen saturation of a normal wound should gradually increase within a certain range during the healing process. If the measured value is far below this range, it indicates that there is a risk of wound healing.

[0083] S732. Multi-factor comprehensive analysis. Risk assessment is performed by comprehensively considering multiple extracted features and parameters. The risk of chronic wounds is not determined by a single factor, but may be the result of multiple factors such as tissue oxygenation, microvascular function, and infection. For example, if there are low tissue oxygen saturation, abnormal microvascular blood flow velocity, and irregular wound edges, the risk of chronic wounds not healing or worsening will be higher.

[0084] S733. Determine the risk classification, and combine it with the comparison of normal range and multi-factor comprehensive analysis to classify the risk, which will help medical staff to formulate corresponding treatment and care strategies. For high-risk chronic wounds, more active intervention measures may be required, such as strengthening anti-infection treatment, improving local blood circulation, etc.

[0085] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A handheld spatial domain imager for chronic wound identification, comprising an imaging system, characterized in that: The imaging system comprises: A light source unit, which serves as a light emission starting point for the entire imaging system; A collimating and splitting unit, the collimating and splitting unit comprising a collimating lens, a grating and a projection lens, wherein the collimating lens is used to collimate the divergent light emitted by the light source, the grating is used to modulate the collimated light to generate a sinusoidal fringe pattern, the projection lens is installed on the light path after the grating modulation, and the projection lens is used to amplify the generated sinusoidal fringe pattern structured light and project it onto the sample; The beam splitting imaging unit is used to realize beam splitting and imaging of the backscattered light of the sample.

2. A handheld spatial domain imager and method for chronic wound identification according to claim 1, characterized in that: The light source unit uses an RGBLED light source.

3. A handheld spatial domain imager for chronic wound identification according to claim 1, characterized in that: The light source unit is divided into an illumination light path and a collection light path, wherein the illumination light path includes: a collimating lens for collimating the light, converting the point light source form of the light source into parallel light, and irradiating the grating with a sinusoidal stripe pattern, and then passing the structured light through a projection lens, and finally projecting it into the biological tissue; the collection light path includes a beam splitter and a CCD lens for imaging, which is used to collect the light after diffuse reflection by the tissue.

4. A handheld spatial domain imager for chronic wound identification according to claim 1, characterized in that: The beam splitting imaging unit includes: a thin film beam splitter and a CMOS camera, wherein the thin film beam splitter is arranged on the path of the sample backscattered light and is used to split the backscattered light according to a certain ratio, and a part of the light can be reflected or transmitted for subsequent imaging, and the CMOS camera is used to receive the light processed by the thin film beam splitter and perform imaging.

5. A method for using the handheld spatial domain imager for chronic wound identification according to any one of claims 1 to 4, characterized in that: The steps include: S1, light from the RGBLED light source passes through the collimating lens and grating to produce a sinusoidal fringe pattern; S2, the structured light is magnified by the projection objective and projected onto the sample; S3, backscattered light from the sample is imaged into the CMOS camera through a pellicle beam splitter and imaged by the SSMD-SFDI system after resting in the supine position for 5 to 10 minutes; S4, select the region of interest with a size of 8.5 mm x 8.5 mm in the center of the image window for data analysis; S5, using a single snapshot multi-frequency demodulation system (SSMD) and further inverting the optical parameters of the double-layer tissue according to the calculated modulation function; S6. Get organizational information; S7. Use support vector machine (SVM) to classify different chronic wounds such as healthy subjects, pressure injuries, and diabetic foot according to structural and physiological parameters, determine the characteristics of tissue damage and physiological structural changes in different wounds, and use interpretable neural networks to evaluate different chronic wounds.

6. The method for using a handheld spatial domain imager for chronic wound identification according to claim 5, characterized in that: The tissue information obtained in S6 includes: optical parameters such as the reduced scattering coefficient and absorption coefficient of the upper and lower layers, as well as information such as the thickness of the upper layer, the content of oxygenated and deoxygenated hemoglobin, and the content of melanin.

7. The method for using a handheld spatial domain imager for chronic wound identification according to claim 5, characterized in that: The S7 comprises the following steps: S71, data collection, used for collecting data; S72, data processing, for processing the collected data; S73. Risk analysis is used to make risk assessments on the processed data.

8. The method for using a handheld spatial domain imager for chronic wound identification according to claim 7, characterized in that: The S71 data collection includes the following steps: S711, Instrument positioning, accurately place the probe of the handheld spatial frequency domain imager at an appropriate distance above the chronic wound, in a darkroom, and according to the instrument's operating instructions to ensure that clear and accurate images can be obtained; S712, light emission, the instrument emits light of specific frequency and intensity distribution to the wound surface. The light contains different spatial frequency components. For example, there may be a combination of multiple wavelengths of light to obtain optical property information such as absorption and scattering of different wavelengths of light by the wound tissue; S713, reflected light collection. The wound tissue reflects the incident light. The optical sensor in the imager collects the reflected light. The reflected light carries information about the structure and physiological characteristics of the wound tissue, such as the oxygenation status of the tissue, microvascular density, melanin content and other related information.

9. The method for using a handheld spatial domain imager for chronic wound identification according to claim 7, characterized in that: The S72 data processing includes the following steps: S721, image reconstruction, reconstructing an optical characteristic image of the wound surface using a specific convolutional neural network algorithm based on the collected reflected light data, and the optical characteristic image shows the distribution of optical parameters of different areas of the wound, such as an absorption coefficient image, a scattering coefficient image, etc.; S722, feature extraction, extracting features related to chronic wound risk from the reconstructed image, wherein the absorption coefficient image is analyzed to determine whether there is abnormal blood absorption in the wound area, the microstructure information of the tissue is extracted from the scattering coefficient image, and the features of the wound edge are extracted, such as edge clarity, irregularity, etc.; S723, parameter calculation; calculate quantitative parameters, such as average oxygen saturation of tissues, estimated microvascular blood flow velocity, etc.

10. The method for using a handheld spatial domain imager for chronic wound identification according to claim 7, characterized in that: The S73 risk analysis includes the following steps: S731, comparing with normal range, comparing the extracted features and calculated parameters with the reference value range in the normal wound healing process; S732, multi-factor comprehensive analysis, comprehensively considering multiple extracted features and parameters for risk assessment; S733. Determine the risk classification and classify the risk of chronic wounds based on the analysis results.

Images