Alzheimer's disease prediction method and system, device, and medium

Abstract

The present invention relates to the technical field of medical data processing, and discloses a method for constructing an Alzheimer's disease prediction model, comprising: obtaining spectral data of a target group by using a non-invasive mode; and on the basis of spectral data, performing training to obtain an Alzheimer's disease prediction model, the target group comprising a healthy group and a group suffering from Alzheimer's disease. According to the present invention, samples of the target group are obtained by using a non-invasive mode, by using the advantages of being capable of detecting all elements at a same time and having a fast analyzing speed of a laser-induced breakdown spectroscopy technology, qualitative or quantitative analysis of a chemical composition of a material surface is implemented by measuring an emission spectrum of micro-plasmas, so that the spectral data of the target group is obtained accurately and quickly, and the Alzheimer's disease prediction model is obtained by means of training on the basis of the spectral data; therefore, a doctor is effectively assisted to carry out early screening for a patient suffering from early Alzheimer's disease. The present invention has a broad application prospect.

Description

Alzheimer's disease prediction method, system, device and medium

[0001] Cross-reference to related applications

[0002] This application claims priority to the Chinese patent application (application number 202311188399.6) filed on September 14, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present invention relates to the field of medical data processing technology, and in particular to an Alzheimer's disease prediction method, system, device and medium. Background Art

[0004] In 2020, 55 million people worldwide were suffering from dementia, and the number of patients is still doubling every 20 years. As one of the main types of dementia, Alzheimer's disease (AD) is a neurodegenerative disease with no obvious lesions. To date, there is no effective cure for AD. Early diagnosis and timely intervention are the only effective measures to delay the progression of the disease. AD is a continuous process that includes three stages: asymptomatic period (preclinical AD), pre-dementia (mild cognitive impairment caused by AD, MCI), and dementia (dementia caused by AD). The pathophysiological changes of AD have already begun 15 to 20 years before the onset of clinical symptoms. Early prediction of AD can continuously advance the diagnosis stage, buying time for timely intervention.

[0005] Currently, the primary diagnostic method for Alzheimer's disease is behavioral cognitive testing, which is often delayed and uncertain. While biomarker testing can provide a more accurate diagnosis at an earlier stage, it is expensive and requires invasive sampling, which can be harmful. Therefore, the medical field urgently needs to develop non-invasive and rapid complementary diagnostic technologies for early screening of diseases like Alzheimer's.

[0006] SUMMARY OF THE INVENTION

[0007] The present invention provides an Alzheimer's disease prediction method, system, device and medium to overcome the defects of the existing technology of Alzheimer's disease detection, such as lag, uncertainty and high detection cost, and to achieve non-invasive and quick assistance to doctors in early screening of Alzheimer's disease.

[0008] The present invention provides a method for constructing an Alzheimer's disease prediction model, comprising:

[0009] Obtain spectral data of target groups in a non-invasive way;

[0010] Based on spectral data, an Alzheimer's disease prediction model was trained;

[0011] The target groups include healthy people and people with Alzheimer's disease.

[0012] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, obtaining spectral data of a target group in a non-invasive manner includes:

[0013] Collecting fecal samples from target groups;

[0014] Based on fecal samples, laser-induced breakdown spectroscopy technology was used to obtain spectral data of the target group.

[0015] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, the Alzheimer's disease prediction model is trained based on spectral data, comprising:

[0016] Preprocessing the spectral data, wherein the preprocessing includes abnormal data processing and spectral area normalization processing;

[0017] Based on the preprocessed spectral data, an Alzheimer's disease prediction model was trained.

[0018] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, the abnormal data processing of spectral data includes:

[0019] According to the spectra of the healthy group and the group with Alzheimer's disease, the spectral difference is obtained;

[0020] According to the spectrum difference, the abnormal spectrum peak is determined;

[0021] According to the abnormal spectral peaks, the spectral lines are identified and assigned, and the abnormal data are filtered out.

[0022] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, the spectral area normalization processing of the spectral data is specifically as follows:

[0023] The spectral data is normalized by the spectral area through the normalization formula;

[0024] Wherein, the normalization formula is:

[0025] In the normalized formula, represents the spectral data after spectral area normalization, x ij Represents the original spectral data of the i-th row and j-th column in the spectral matrix with m rows and n columns.

[0026] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, the Alzheimer's disease prediction model is trained based on spectral data, comprising:

[0027] Mapping spectral data through kernel function;

[0028] Based on the mapped spectral data, a linear boundary is obtained to train an Alzheimer's disease prediction model.

[0029] According to a method for constructing an Alzheimer's disease prediction model provided by the present invention, the expression of the kernel function is:

[0030] In the expression of kernel function, K represents kernel function, x i represents the spectral vector of row i, x j represents the j-th column spectral vector, Represents a mapping;

[0031] The linear boundary expression is:

[0032] In the expression of linear boundary, x test represents the vector to be measured, b represents the intercept, α i Lagrange multiplier corresponding to the row vector, α j Lagrange multipliers corresponding to column vectors, y i Represents the label vector of row i, y j represents the label vector of the jth column, X represents the overall spectral matrix, x i represents the spectral vector of row i, x j represents the j-th column spectral vector.

[0033] The present invention also provides a system for constructing an Alzheimer's disease prediction model, comprising:

[0034] A data receiving module, wherein the data receiving module is configured to: receive spectral data of a target group obtained in a non-invasive manner;

[0035] A model training module, wherein the model training module is configured to: train an Alzheimer's disease prediction model based on the spectral data received by the data receiving module;

[0036] The target groups include healthy groups and groups suffering from Alzheimer's disease.

[0037] It will be appreciated by those skilled in the art that the construction system of the Alzheimer's disease prediction model of the present invention can be implemented by software or hardware or a combination of software and hardware. When implemented by hardware, the data receiving module is electrically connected to the model training module. When implemented by a combination of software and hardware, the model training module can be implemented by executing logic instructions or calling program instructions stored in a memory by a processor of a hardware device such as a host computer, an industrial computer or a personal computer. In this case, the data receiving module can be built into the hardware device or can be a separate hardware device that transmits data by communicating with the hardware device (for example, wireless or wired communication). When implemented by software, data can be manually input by an input device or obtained from a data acquisition device by an I / O interface, and the process of data reception and model training is implemented by a processor.

[0038] The present invention also provides an Alzheimer's disease early screening system, comprising:

[0039] A data receiving module, wherein the data receiving module is configured to: receive spectral data of a subject obtained in a non-invasive manner;

[0040] A prediction module configured to: obtain an Alzheimer's disease prediction result for the subject based on the spectral data of the subject received by the data receiving module and the Alzheimer's disease prediction model obtained by any of the above-mentioned methods for constructing an Alzheimer's disease prediction model;

[0041] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0042] It will be appreciated by those skilled in the art that the Alzheimer's disease prediction system of the present invention can be implemented by software or hardware or a combination of software and hardware. When implemented by hardware, the data receiving module is electrically connected to the prediction module. When implemented by a combination of software and hardware, the prediction module can be implemented by executing logic instructions stored in a memory or calling program instructions by a processor of a hardware device such as a host computer, an industrial computer or a personal computer. In this case, the data receiving module can be built into the hardware device or can be a separate hardware device that transmits data by communicating with the hardware device (for example, wireless or wired communication). When implemented by software, data can be manually input by an input device or obtained from a data acquisition device through an I / O interface, and the processor implements the process of data reception and prediction.

[0043] The present invention further provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, the following steps are implemented:

[0044] Receiving spectral data of a subject obtained in a non-invasive manner;

[0045] Obtaining an Alzheimer's disease prediction result for the subject by using the Alzheimer's disease prediction model obtained by any of the above-mentioned methods for constructing an Alzheimer's disease prediction model based on the received spectral data of the subject;

[0046] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0047] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0048] Receiving spectral data of a subject obtained in a non-invasive manner;

[0049] Obtaining an Alzheimer's disease prediction result for the subject by using the Alzheimer's disease prediction model obtained by any of the above-mentioned methods for constructing an Alzheimer's disease prediction model based on the received spectral data of the subject;

[0050] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0051] The present invention further provides a computer program product, comprising a computer program, which may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can implement the following steps:

[0052] Receiving spectral data of a subject obtained in a non-invasive manner;

[0053] Obtaining an Alzheimer's disease prediction result for the subject by using the Alzheimer's disease prediction model obtained by any of the above-mentioned methods for constructing an Alzheimer's disease prediction model based on the received spectral data of the subject;

[0054] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0055] In the present invention, Alzheimer's disease may be Alzheimer's disease associated with or caused by brain-gut axis disorder.

[0056] The present invention provides a method, system, device, and medium for predicting Alzheimer's disease. These methods utilize laser-induced breakdown spectroscopy (LIBS), which allows for simultaneous detection of all elements and rapid analysis, to accurately and rapidly analyze the surface chemical composition of materials by measuring the emission spectrum of microplasmas. This method, along with an Alzheimer's disease prediction model trained based on this spectral data, can effectively assist doctors and patients in early screening, enabling effective intervention for Alzheimer's patients at an early stage and slowing their progression. This invention has broad application prospects in the medical field and can bring significant economic and social value. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] FIG1 is a flow chart of a method for constructing an Alzheimer's disease prediction model provided by the present invention.

[0059] FIG2 shows the difference in LIBS spectra between feces of healthy mice and feces of mice with Alzheimer's disease, where wavelength represents wavelength.

[0060] FIG3 shows the performance evaluation of the Alzheimer's disease prediction model trained by the present invention, wherein confusion matrix represents a confusion matrix.

[0061] FIG4 is a schematic diagram of the structure of a system for constructing an Alzheimer's disease prediction model provided by the present invention.

[0062] FIG5 is a schematic structural diagram of an electronic device provided by the present invention. Modes for Carrying Out the Invention

[0063] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0064] The following describes the Alzheimer's disease prediction method, system, device and medium provided by the present invention in conjunction with Figures 1 to 5.

[0065] FIG1 is a flow chart of a method for constructing a prediction model for Alzheimer's disease provided by the present invention. Referring to FIG1 , a method for constructing a prediction model for Alzheimer's disease provided by the present invention may include:

[0066] Step S110, obtaining spectral data of the target group in a non-invasive manner;

[0067] Step S120: training an Alzheimer's disease prediction model based on the spectral data;

[0068] It should be noted that the target population includes a healthy population and a population with Alzheimer's disease. The target population can be animals, preferably mammals, including experimental animals such as rodents (e.g., mice, rats, etc.), canines, equines, bovines, felines, primates, and humans, more preferably rodents, primates, and humans, and most preferably commonly used experimental rodents (e.g., mice, rats, etc.) and humans. When using the Alzheimer's disease prediction model, the corresponding prediction object should be consistent with the object used to train the Alzheimer's disease prediction model.

[0069] In one embodiment, step S110 may include:

[0070] Collecting fecal samples from target groups;

[0071] Based on fecal samples, laser-induced breakdown spectroscopy technology was used to obtain spectral data of the target group.

[0072] It should be noted that the excrement samples can be feces, urine, saliva or sweat, etc., and there is no need to invade the body of the target group to obtain them. Obtaining samples in a non-invasive way can protect the physical integrity of the target group, avoid causing damage to the target group's body, and at the same time improve the accuracy of the spectral data.

[0073] It should be noted that laser-induced breakdown spectroscopy utilizes a focused pulsed laser to ablate a localized surface of a material to generate a microplasma. The emission spectrum of this microplasma is then measured to enable qualitative or quantitative analysis of the chemical composition of the material surface. As an emerging emission spectroscopy technique, LIBS offers unique advantages, including the lack of sample pretreatment, simultaneous detection of all elements, and rapid analysis. It holds broad application prospects in the biomedical field.

[0074] Laser-induced breakdown spectroscopy can be used to reflect the content and variation patterns of various chemical components in fecal samples. The types of blood elements in healthy groups and those with Alzheimer's disease are the same, but the content of elements such as Mg, Na, and Ca in the Alzheimer's disease group is relatively high, and the corresponding spectral intensity collected is also high (Figure 2 shows the difference between the average spectrum of fecal samples of mice with Alzheimer's disease and the average spectrum of fecal samples of healthy mice. It can be seen that the spectral intensity of the constant elements Na, Mg, and Ca has obvious differences). Using the target group attribute label and the obtained spectral data to train the Alzheimer's disease prediction model can improve the prediction accuracy. In this embodiment, the spectral data can be direct spectral peaks or other forms of spectral data, such as the fitted peak height / area of ​​the spectral peak or the spectral peak ratio.

[0075] Specifically, when collecting excrement samples from the target group, fecal samples from the target group can be taken, the target group attribute labels can be recorded, each fecal sample can be compacted to the same thickness (for example, 1 mm thick), and left to dry at room temperature (or oven-dried) so that the water content of each fecal sample is roughly maintained at a uniform level (for example, a preset numerical range). Then trigger and adjust the laser until the spectral signal is good and there is no continuous background (adjustment parameters include laser energy, optimization integration time, delay time, etc.), use a focused pulse laser to burn the surface of the fecal sample to generate plasma, and scan each fecal sample at the same time to collect spectral data at different positions. After obtaining the spectral data, the spectral data and the corresponding target group attribute labels can be divided into a training set and a test set. The training set is used for training the Alzheimer's disease prediction model, and the test set is used for testing the Alzheimer's disease prediction model. When training the Alzheimer's disease prediction model, an Alzheimer's disease prediction model for different stages of the disease can be established by tracking and sampling the entire onset cycle, assisting doctors in diagnosing Alzheimer's disease in the early, middle and late stages.

[0076] In one embodiment, step S120 may include:

[0077] Preprocessing the spectral data, wherein the preprocessing includes abnormal data processing and spectral area normalization processing;

[0078] Map the preprocessed spectral data through the kernel function;

[0079] Based on the mapped spectral data, a linear boundary is obtained to train an Alzheimer's disease prediction model.

[0080] In one embodiment, abnormal data processing of spectral data can be achieved by the following steps:

[0081] According to the spectra of the healthy group and the group with Alzheimer's disease, the spectral difference is obtained;

[0082] According to the spectrum difference, the abnormal spectrum peak is determined;

[0083] According to the abnormal spectral peaks, the spectral lines are identified and assigned, and the abnormal data are filtered out.

[0084] Abnormal data can occur in both healthy and diseased groups, primarily due to large data fluctuations, such as noise instead of normal data. Specifically, the spectra of the healthy group and the spectra of the Alzheimer's group are compared as a whole to obtain a spectral difference. If the spectral difference reaches an intensity of more than 10% of the stronger original spectral peak (the original spectral peak can be either the spectrum of the healthy group or the spectrum of the Alzheimer's group, the difference between the two is small), it can be determined to be an abnormal spectral peak. Comparing the spectral difference with the original spectral peak helps identify the location of the abnormal spectral peak and filter out abnormal data.

[0085] In one embodiment, the spectral area normalization process is performed on the spectral data, specifically:

[0086] The spectral data is normalized by the spectral area through the normalization formula;

[0087] Wherein, the normalization formula is:

[0088] In the normalized formula, represents the spectral data after spectral area normalization, x ij Represents the original spectral data (e.g., original spectral intensity) at the i-th row and j-th column in the spectral matrix with m rows and n columns.

[0089] In one embodiment, the kernel function is expressed as:

[0090] In the expression of kernel function, K represents kernel function, x i represents the spectral vector of row i, x j represents the j-th column spectral vector, Represents a mapping;

[0091] When calling the sklearn.svm.SVC() class, the linear boundary expression can be:

[0092] In the expression of linear boundary, x test represents the vector to be measured, b represents the intercept, α i Lagrange multiplier corresponding to the row vector, α j Lagrange multipliers corresponding to column vectors, y i Represents the label vector of row i, y j represents the label vector of the jth column, X represents the overall spectral matrix, x i represents the spectral vector of row i, x j Represents the j-th column spectral vector, with no other hyperparameter adjustments.

[0093] This embodiment uses a support vector machine (SVM) as a function to train the Alzheimer's disease prediction model. It uses a kernel function to map spectral data to a higher-dimensional feature space, making it easier to find classification boundaries to distinguish the characteristic differences between healthy groups and groups with Alzheimer's disease, so as to more quickly and accurately predict whether the subject is an Alzheimer's disease patient. When the Alzheimer's disease prediction model training is completed, the Alzheimer's disease prediction model is tested using a test set, and the predicted results are compared with the actual results. If there are prediction errors, the erroneous data is included in the training set, and the Alzheimer's disease prediction model is iteratively trained again until the prediction performance of the Alzheimer's disease prediction model is stable (for example, the prediction accuracy reaches a certain preset threshold, indicating that the prediction performance is stable).

[0094] Based on the method for constructing an Alzheimer's disease prediction model provided by the present invention, a specific example will be provided below.

[0095] Several experimental mice (10 healthy mice (HC) and 10 Alzheimer's disease mice (AD)) were purchased from Cavensburg (Suzhou) Model Animal Research Co., Ltd. (3×TgAD). The mice were housed two per cage under a 12-h light-dark cycle (lights on from 7:00 AM to 7:00 PM), with a relative humidity of 53% and a temperature of 24 ± 1°C. Water (RO purified water) and food (rat and mouse maintenance diet) were freely available. The mice were cleaned and their bedding changed weekly. After a period of housing, the mice were grasped by the scruff of the neck to stimulate defecation. Ten fecal pellets each from healthy and AD mice (one pellet per experimental mouse) were collected. Fresh fecal pellets were collected directly into 1.5 ml Eppendorf tubes. All work areas and experimental equipment were disinfected with 75% ethanol. After drying, the fecal pellets were placed on a three-dimensional translation stage, and the sample height was adjusted to the focal point of the LIBS system. The pulsed laser of the LIBS system was focused on the surface of the fecal pellet, ablating the surface to generate plasma. Simultaneously, the spectrometer parameters and laser energy were adjusted until the spectral signal was good and there was no continuous background. The optimal laser energy was 50 mJ (power density reached 45 GW / cm2), the spectrometer integration time was 5 μs, and the delay time was 0.3 μs. Under the optimized acquisition parameters, the fecal pellets were translated and scanned, acquiring 100 LIBS spectra at different positions, for a total of 2000 spectral data. The spectral data were then processed for abnormal data and area normalization. The preprocessed spectral data of 16 fecal pellets (8 normal and 8 diseased, for a total of 1600 LIBS spectra) were used as the training set and combined with the support state vector machine algorithm (linear kernel function) to develop an Alzheimer's disease prediction model. This model was then used to predict the preprocessed spectral data of the remaining four feces particles (a total of 400 LIBS spectra) to obtain a model performance evaluation. As shown in Figure 3, the Alzheimer's disease prediction model trained in this example achieved an accuracy of 0.93, a precision of 0.91, a sensitivity / recall of 0.96, and a specificity of 0.9, indicating excellent performance across all performance indicators.

[0096] The following describes a system for constructing an Alzheimer's disease prediction model provided by the present invention. The system for constructing an Alzheimer's disease prediction model described below and the method for constructing an Alzheimer's disease prediction model described above can refer to each other.

[0097] 4 , the present invention provides a system for constructing an Alzheimer's disease prediction model, which may include:

[0098] A data receiving module, wherein the data receiving module is configured to: receive spectral data of a target group obtained in a non-invasive manner;

[0099] A model training module, wherein the model training module is configured to: train an Alzheimer's disease prediction model based on the spectral data received by the data receiving module;

[0100] The target groups include healthy groups and groups suffering from Alzheimer's disease.

[0101] The present invention also provides an Alzheimer's disease prediction system, which may include:

[0102] A data receiving module, wherein the data receiving module is configured to: receive spectral data of a subject obtained in a non-invasive manner;

[0103] A prediction module configured to obtain an Alzheimer's disease prediction result for the subject based on the spectral data of the subject received by the data receiving module and the Alzheimer's disease prediction model obtained by the above-mentioned method for constructing an Alzheimer's disease prediction model;

[0104] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0105] It should be noted that the subject can be an animal, preferably a mammal, including rodents, canines, equines, bovines, felines, primates and humans, more preferably primates and humans.

[0106] FIG5 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG5 , the electronic device may include: a processor 810, a communications interface 820, a memory 830, and a communication bus 840. The processor 810, the communications interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to perform the following steps:

[0107] Receiving spectral data of a subject obtained in a non-invasive manner;

[0108] Obtaining an Alzheimer's disease prediction result for the subject based on the received spectral data of the subject and the Alzheimer's disease prediction model obtained by the above-mentioned method for constructing the Alzheimer's disease prediction model;

[0109] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0110] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0111] In another aspect, the present invention further provides a computer program product, comprising a computer program, which may be stored on a non-transitory computer-readable storage medium, and wherein when the computer program is executed by a processor, the computer is capable of performing the following steps:

[0112] Receiving spectral data of a subject obtained in a non-invasive manner;

[0113] Obtaining an Alzheimer's disease prediction result for the subject based on the received spectral data of the subject and the Alzheimer's disease prediction model obtained by the above-mentioned method for constructing the Alzheimer's disease prediction model;

[0114] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0115] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is configured to execute the following steps when executed by a processor:

[0116] Receiving spectral data of a subject obtained in a non-invasive manner;

[0117] Obtaining an Alzheimer's disease prediction result for the subject based on the received spectral data of the subject and the Alzheimer's disease prediction model obtained by the above-mentioned method for constructing the Alzheimer's disease prediction model;

[0118] The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or early stage of dementia, or patients whose Alzheimer's disease is to be excluded.

[0119] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0120] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention. Industrial Applicability

[0122] The present invention provides an Alzheimer's disease prediction method, system, device and medium, which non-invasively obtain samples from a target group (such as fecal samples from the target group), and utilizes the advantages of laser-induced breakdown spectroscopy (LIBS) technology to simultaneously detect all elements and analyze them at a fast speed. By measuring the emission spectrum of microplasma, qualitative or quantitative analysis of the chemical composition of the material surface is achieved, and spectral data of the target group is accurately and quickly obtained. The Alzheimer's disease prediction model obtained by training based on the spectral data can effectively assist doctors in early diagnosis and intervention of patients, so that Alzheimer's patients can receive timely and effective treatment. The present invention has broad application prospects in the medical field and can bring huge economic and social value.

Claims

1. A method for constructing an Alzheimer's disease prediction model, comprising: Obtain spectral data of target groups in a non-invasive manner; Based on spectral data, an Alzheimer's disease prediction model was trained; The target groups include a healthy group and a group suffering from Alzheimer's disease.

2. The method for constructing an Alzheimer's disease prediction model according to claim 1, wherein: The target population is animals, preferably mammals, which are selected from the group consisting of rodents, canines, equines, bovines, felines, primates and humans, more preferably rodents, primates and humans, and still more preferably, experimental animals and humans.

3. The method for constructing an Alzheimer's disease prediction model according to claim 1, wherein: The group with Alzheimer's disease in the target group includes the following groups with Alzheimer's disease: a group with Alzheimer's disease in the asymptomatic period, a group with Alzheimer's disease in the pre-dementia period, a group with Alzheimer's disease in the dementia period, or any combination thereof, preferably including a group with Alzheimer's disease in the asymptomatic period, a group with Alzheimer's disease in the pre-dementia period, and a group with Alzheimer's disease in the dementia period.

4. The method for constructing an Alzheimer's disease prediction model according to claim 1, wherein: The spectral data is selected from any one of the following: spectral peak data, fitted peak height data of spectral peaks, fitted area data of spectral peaks, spectral peak ratio data, preferably spectral peak data.

5. The method for constructing an Alzheimer's disease prediction model according to claim 1, wherein: The non-invasive method of obtaining the spectral data of the target group includes: Collecting fecal samples from target groups; Based on fecal samples, laser induced breakdown spectroscopy technology was used to obtain spectral data of the target group.

6. The method for constructing an Alzheimer's disease prediction model according to claim 5, wherein: The excrement sample is selected from any one of the following excrement samples obtained without invading the body of the target group: feces, urine, saliva and sweat, preferably feces.

7. The method for constructing an Alzheimer's disease prediction model according to any one of claims 1 to 6, wherein: The Alzheimer's disease prediction model is trained based on the spectral data, including: Preprocessing the spectral data, wherein the preprocessing includes abnormal data processing and spectral area normalization processing; Based on the preprocessed spectral data, an Alzheimer's disease prediction model was trained.

8. The method for constructing an Alzheimer's disease prediction model according to claim 7, wherein: The abnormal data processing of the spectral data includes: According to the spectra of the healthy group and the group with Alzheimer's disease, the spectral difference is obtained; According to the spectrum difference, the abnormal spectrum peak is determined; According to the abnormal spectral peaks, the spectral lines are identified and attributed, and the abnormal data are filtered out.

9. The method for constructing an Alzheimer's disease prediction model according to claim 7, wherein: The spectral area normalization process for the spectral data is specifically as follows: The spectral data is normalized by spectral area through the normalization formula; Wherein, the normalization formula is: In the normalized formula, represents the spectral data after spectral area normalization, x ij Represents the original spectral data of the i-th row and j-th column in the spectral matrix of m rows and n columns.

10. The method for constructing an Alzheimer's disease prediction model according to claim 7, wherein: The Alzheimer's disease prediction model is trained based on the spectral data, including: Map the spectral data through the kernel function; Based on the mapped spectral data, a linear boundary is obtained to train an Alzheimer's disease prediction model.

11. The method for constructing an Alzheimer's disease prediction model according to claim 10, wherein: The expression of the kernel function is: In the expression of kernel function, K represents kernel function, x i represents the spectral vector of the i-th row, x j represents the j-th column spectral vector, Representation mapping; Wherein, the expression of the linear boundary is: In the linear boundary expression, x test represents the vector to be measured, b represents the intercept, α i The Lagrange multiplier corresponding to the row vector, α j Lagrange multipliers corresponding to column vectors, y i Represents the label vector of the i-th row, y j represents the label vector of the jth column, X represents the overall spectral matrix, x i represents the spectral vector of the i-th row, x j represents the j-th column spectral vector.

12. A method for early screening of Alzheimer's disease, wherein: include: Receiving spectral data of a subject obtained by testing a sample obtained in a non-invasive manner; According to the received spectral data of the subject, a prediction result of Alzheimer's disease of the subject is obtained by using the Alzheimer's disease prediction model obtained by the method for constructing an Alzheimer's disease prediction model according to any one of claims 1 to 11; The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or pre-dementia stage, or patients whose Alzheimer's disease is to be excluded.

13. The method for early screening of Alzheimer's disease according to claim 12, wherein: The type of subjects is consistent with the type of target population used when training the Alzheimer's disease prediction model, wherein the subjects are animals, preferably mammals, and the mammals are selected from the following mammals: rodents, canines, equines, bovines, felines, primates and humans, more preferably rodents, primates and humans, and also preferably, experimental animals and humans.

14. The method for early screening of Alzheimer's disease according to claim 12, wherein: The type of the spectral data of the subject is consistent with the type of spectral data of the target group used in training the Alzheimer's disease prediction model, wherein the spectral data of the subject is selected from any one of the following: spectral peak data, fitted peak height data of the spectral peak, fitted area data of the spectral peak, spectral peak ratio data, preferably spectral peak data.

15. The method for early screening of Alzheimer's disease according to claim 12, wherein: The predicted result of Alzheimer's disease of the subject is any of the following: the subject is healthy, the subject is in the asymptomatic stage of Alzheimer's disease, the subject is in the pre-dementia stage of Alzheimer's disease, and the subject is in the dementia stage of Alzheimer's disease.

16. The method for early screening of Alzheimer's disease according to claim 12, wherein: The receiving and detecting of spectral data of a subject obtained by a sample obtained in a non-invasive manner comprises: Collect fecal samples from the subjects; Based on the fecal samples, the spectral data of the subject is obtained using laser induced breakdown spectroscopy technology.

17. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor performs the following steps: Receiving spectral data of a subject obtained by testing a sample obtained in a non-invasive manner; According to the received spectral data of the subject, a prediction result of Alzheimer's disease of the subject is obtained by using the Alzheimer's disease prediction model obtained by the method for constructing an Alzheimer's disease prediction model according to any one of claims 1 to 11; The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or pre-dementia stage, or patients whose Alzheimer's disease is to be excluded.

18. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, a method is implemented comprising the following steps: Receiving spectral data of a subject obtained by testing a sample obtained in a non-invasive manner; According to the received spectral data of the subject, a prediction result of Alzheimer's disease of the subject is obtained by using the Alzheimer's disease prediction model obtained by the method for constructing an Alzheimer's disease prediction model according to any one of claims 1 to 11; The subjects are healthy people, Alzheimer's patients suspected to be in the asymptomatic stage or pre-dementia stage, or patients whose Alzheimer's disease is to be excluded.

Images