Middle-aged and elderly mild cognitive impairment rapid auxiliary screening system based on machine learning

Abstract

The invention belongs to the technical field of auxiliary medical treatment, and particularly relates to a rapid auxiliary screening system for mild cognitive impairment of middle-aged and elderly people based on machine learning. The system includes: an input module configured to input examination data of a subject; the prediction module is configured to input the examination data into a machine learning model to obtain a prediction result whether the middle-aged and elderly mild cognitive impairment risk exists or not; and the output module is configured to output the prediction result. The examination data is selected from at least one of the following indexes: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, upper arm middle circumference and delayed recall. The MCI screening model for accurately predicting mild cognitive impairment of middle-aged and elderly people based on indexes easy to collect is successfully constructed. The model ensures the prediction performance, simplifies the collection of indexes, has strong generalization ability, promotes the popularization of the technology, and shows huge practical potential.

Description

Technical Field

[0001] The present invention belongs to the field of auxiliary medical technology, and specifically relates to a rapid auxiliary screening system for mild cognitive impairment in middle-aged and elderly people based on machine learning. Background Art

[0002] Identifying mild cognitive impairment (MCI), a precursor to dementia, is crucial for early detection and intervention of dementia. Identifying early mild symptoms of MCI in seemingly healthy community dwellers is the key to preventing and controlling dementia. Although large-scale population screening for MCI has not yet been possible due to various factors, the development and application of new screening tools to achieve rapid, accurate, and intelligent widespread screening of MCI in community dwellers has become a common goal of researchers around the world.

[0003] Among the tools for assessing overall cognitive function, the 2018 Chinese Guidelines for the Diagnosis and Treatment of Dementia and Cognitive Impairment (V): Diagnosis and Treatment of Mild Cognitive Impairment recommends the use of the Montreal Cognitive Assessment (MoCA) scale as the preferred tool for early screening of mild cognitive impairment, which is considered the benchmark for MCI screening. The scale was developed by Nasreddine et al. in 2005 and is specifically used to screen for mild cognitive impairment. The assessment includes eight aspects: visual-spatial and executive function, naming, memory, attention, language, abstraction, delayed recall, and orientation. It also takes into account the educational background of the subjects and shows high sensitivity (80-100%) and specificity (50-76%).

[0004] In the research on the interdisciplinary field of cognitive impairment and machine learning, the research on MCI screening models is still in its infancy, mainly focusing on the late prediction of diseases such as Alzheimer's disease, while there are relatively few studies on MCI auxiliary screening and auxiliary diagnosis in the early stages of the disease. In the study of applying traditional regression models and other methods to MCI screening models, it was found that the effects of these traditional methods were not ideal. However, with the rapid development of artificial intelligence technology, machine learning methods may become one of the future directions for solving this problem.

[0005] In current research, the use of machine learning methods to assist in the diagnosis of MCI usually relies on some medical diagnostic indicators, such as biomarkers in cerebrospinal fluid (such as tau protein, Aβ protein, etc.), imaging features (such as magnetic resonance imaging MRI, positron emission tomography PET, etc.), and bioelectric signals (such as EEG signals, etc.). There are certain challenges in the collection of these indicators, such as difficulty in collection, high cost, expensive equipment or the need to use high-value consumables, as well as invasive operations and infection risks that patients may face. These challenges limit the widespread application of machine learning technology in the auxiliary diagnosis of MCI.

[0006] Therefore, developing a technology that can use easily accessible indicators for auxiliary diagnosis of MCI while ensuring diagnostic performance is an urgent problem to be solved in the current field. Summary of the invention

[0007] In response to the problems of the prior art, the present invention provides a rapid auxiliary screening system for mild cognitive impairment in middle-aged and elderly people based on machine learning.

[0008] A rapid auxiliary screening system for mild cognitive impairment in the middle-aged and elderly based on machine learning, comprising: An input module is configured to input the examination data of the subject, wherein the examination data is selected from at least one of the following indicators: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, mid-upper arm circumference, and delayed recall; A prediction module is configured to input the inspection data into a machine learning model to obtain a prediction result of whether the patient has a risk of mild cognitive impairment in the middle-aged and elderly; The output module is configured to output the prediction result.

[0009] Preferably, the examination data is selected from delayed recall and at least one of the following indicators: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, and mid-upper arm circumference.

[0010] Preferably, the examination data is selected from: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, mid-upper arm circumference and delayed recall.

[0011] Preferably, the algorithm of the machine learning model is selected from the extreme gradient boosting algorithm.

[0012] Preferably, the parameter settings of the machine learning model include: sampling rate = 1, learning rate = 0.01, maximum depth = 3, subsample = 0.5.

[0013] The present invention also provides a computer-readable storage medium, on which is stored: a computer program for implementing the above-mentioned simple multimodal index rapid auxiliary screening system for mild cognitive impairment in the elderly.

[0014] In the present invention, the symbol " / " represents a division symbol, for example: "calf circumference / neck circumference" refers to the ratio of the calf circumference divided by the neck circumference.

[0015] The present invention has developed a machine learning system (MCI screening model) based on easily accessible indicators for assisting in the diagnosis of mild cognitive impairment. By carefully selecting indicators and algorithms, a high prediction performance is achieved, meeting the actual needs of middle-aged and elderly people for rapid assistance in screening mild cognitive impairment. Specifically, the application of the technical solution of the present invention can achieve the following beneficial technical effects: 1. The input indicators used in the present invention are easy to obtain, low-cost, and do not require the purchase of expensive large-scale equipment, making it easy for medical institutions at different levels, from community health service centers to township health centers, to collect data. This solution effectively overcomes the challenges faced by the existing technology, such as the collection of data such as cerebrospinal fluid indicators, imaging features, and bioelectric signals, and contributes to the widespread popularization and application of machine learning methods.

[0016] 2. The MCI screening model of the present invention performed well in the cross-validation of the data set used when constructing the model (AUC>0.8). In addition, the screening model was independently externally validated in different years and for populations different from the model development data set, and its AUC value remained above 0.8, indicating that the MCI screening model has good generalization ability.

[0017] In general, the MCI screening model of the present invention reduces the difficulty of inputting indicators while ensuring the prediction performance, which is convenient for wider application. The present invention is suitable for early community screening of mild cognitive impairment, and potential application sites include primary medical and health service stations, biological sample analysis institutions, physical examination institutions, and health management institutions, showing broad application prospects.

[0018] Obviously, according to the above contents of the present invention, in accordance with common technical knowledge and customary means in the art, without departing from the above basic technical ideas of the present invention, other various forms of modification, replacement or change may be made.

[0019] The above contents of the present invention are further described in detail below through specific implementation methods in the form of embodiments. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following examples. All technologies realized based on the above contents of the present invention belong to the scope of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Establish a technical roadmap for the MCI screening model in Experimental Example 1; Figure 2 A path diagram for the establishment and external validation of the MCI screening model in Experimental Example 1; Figure 3 The ROC curve diagram of the best MCI screening model obtained in Experimental Example 1 in the model establishment data set; Figure 4 The figure is the ROC curve diagram of the evaluation of the optimal MCI screening model obtained in Experimental Example 1 in an external independent validation data set. DETAILED DESCRIPTION

[0021] It should be noted that the algorithms of data collection, transmission, storage and processing steps not specifically described in the embodiments, as well as the hardware structure, circuit connection, etc. not specifically described can all be implemented through the contents disclosed in the prior art.

[0022] Example 1 Rapid auxiliary screening system for mild cognitive impairment in the middle-aged and elderly based on machine learning The system of this embodiment includes: An input module is configured to input examination data of a subject; A prediction module is configured to input the inspection data into a machine learning model (hereinafter referred to as an MCI screening model) to obtain a prediction result of whether the patient has a risk of mild cognitive impairment in the middle-aged and elderly; The output module is configured to output the prediction result.

[0023] Wherein, the inspection data is selected from: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, mid-upper arm circumference, and delayed recall; the quantitative measurement method of delayed recall is: before starting the body dimension measurement, select 5 words, record and repeat reading twice, for example: "face", "velvet", "church", "chrysanthemum", "red", and then start the body dimension measurement. These 5 words are not prompted during the measurement process. After all measurements are completed (more than 5 minutes), ask whether you remember the 5 words. No prompts are given, and there is no need to recall in order. Each time you recall one, record 1 point, with a full score of 5 points, and just enter the score.

[0024] In this embodiment, the algorithm of the machine learning model is selected from the extreme gradient boosting algorithm. The parameter settings of the machine learning model include: sampling rate=1, learning rate=0.01, maximum depth=3, subsample=0.5.

[0025] The technical solution of the present invention is further illustrated below through experimental data.

[0026] Experimental Example 1 Establishment and external validation of MCI screening model 1. Experimental Methods In this experimental case, we created a cognitive health cohort of a natural population in the community and used a snowball multi-stage sample recruitment strategy. First, we used the baseline sample population data recruited in the first stage to build and evaluate the community MCI screening model. Subsequently, we used the baseline sample population data newly added in the second stage through the snowball method to conduct independent external validation and evaluation of the community MCI screening model to test the generalization ability of the model.

[0027] The overall technical roadmap and model technical roadmap of the present invention are shown in Figure 1 and Figure 2 Specifically, the technical path of model construction mainly includes the following three parts: 1. Model construction and external validation data collection technology based on community cohort studies: Create a cognitive health cohort of natural populations in the community, collect cognitive-related assessment data of cohort study samples, and provide real-world data support for the development and external validation of rapid screening models for mild cognitive impairment. This study recruited cohort study population samples from natural population communities in Chengdu, adopted a cohort study strategy, conducted a comprehensive cognitive health assessment of all participants, systematically collected basic demographic information of the sample population, assessment data on cognition and influencing factors, physical examinations, and biological sample testing information, and established a high-quality cognitive health cohort of natural populations in the community. The establishment process of the community cohort adopted a "three-level quality control" strategy, with quality control carried out in three stages: preparation, execution, and maintenance, providing a solid foundation for accurately collecting real-world data, conducting cognitive-related scientific research, and developing sustainable cohorts.

[0028] 2. Construction and internal evaluation technology of community MCI screening model: Using the biological samples and evaluation data of the cohort study sample population, through different model training methods, a rapid screening model for mild cognitive impairment suitable for middle-aged and elderly residents in the community is constructed and internally evaluated. Using the baseline data of the first phase of the community natural population cognitive health cohort, try to use a variety of methods to screen variables such as biological sample indicators and demographic information, and combine professional knowledge injection and other methods to select highly relevant variables; by comparing the prediction performance of traditional models and machine learning models, select models or model combinations with good generalization ability for screening and prediction of mild cognitive impairment in the community, and complete the internal evaluation of model prediction.

[0029] 3. External validation and feature interpretability analysis technology of community MCI screening model: Using the newly added sample data included in the cohort study in a snowballing manner, the established rapid screening model for mild cognitive impairment was externally validated, and an interpretable description of the variables was performed. Using the design idea of ​​diagnostic tests, the evaluation data of the second phase of the snowballing population baseline sample of the community natural population cognitive health cohort was used to evaluate the prediction effect of the established model, and the difference between the prediction results and the traditional gold standard scale screening results was compared, and the consistency was evaluated.

[0030] 2. Experimental Results In this experimental example, in terms of community MCI screening model establishment, effect evaluation and screening results, focusing on the problem of building a screening model suitable for mild cognitive impairment in the community, a total of 274 alternative MCI screening models were established; during the internal evaluation process within the MCI screening model establishment data set, it was found that according to the screening model optimization rules (condition one: AUC ≥ 0.8; condition two 0.75 ≤ AUC < 0.8, and sensitivity is greater than specificity), a total of 38 preferred MCI screening models entered the external validation stage; in the context of an external validation independent data set, based on the expert consultation opinion of the optimal screening model of "AUC ≥ 0.8 and sensitivity is greater than specificity", while taking into account the comprehensive effectiveness of the screening model in the model establishment data set and the model external validation data set, after expert consultation, it was finally determined that the MCI screening model described in Example 1 had the best predictive performance.

[0031] In the cross-validation evaluation of the model establishment data set, the sensitivity, specificity, accuracy and AUC of the above-mentioned optimal MCI screening model were 70.59%, 90.36%, 81.46% and 80.47%, respectively; the performance of the relevant evaluation indicators of this method in the external independent validation data set included sensitivity, specificity, accuracy and AUC of 88.24%, 73.33%, 82.72% and 85.13%, respectively; the relevant ROC curve Figure 3 and Figure 4 shown.

[0032] From the above implementation cases and experimental examples, it can be seen that the present invention successfully constructs an MCI screening model based on easily collected indicators to accurately predict mild cognitive impairment in middle-aged and elderly people. This model not only ensures the prediction performance, but also simplifies the collection of indicators, has strong generalization ability, is conducive to the large-scale promotion of technology, and shows great practical application potential.

Claims

1. A rapid auxiliary screening system for mild cognitive impairment in the middle-aged and elderly based on machine learning, characterized in that: include: An input module is configured to input the examination data of the subject, wherein the examination data is selected from at least one of the following indicators: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, mid-upper arm circumference, and delayed recall; A prediction module is configured to input the inspection data into a machine learning model to obtain a prediction result of whether the patient has a risk of mild cognitive impairment in the middle-aged and elderly; The output module is configured to output the prediction result.

2. The simplified multimodal index rapid auxiliary screening system for mild cognitive impairment in the elderly according to claim 1, characterized in that: The examination data is selected from delayed recall and at least one of the following indicators: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, and mid-upper arm circumference.

3. The simplified multimodal index rapid auxiliary screening system for mild cognitive impairment in the elderly according to claim 2, characterized in that: The examination data are selected from: calf circumference / neck circumference, height / calf circumference, height / neck circumference, calf circumference, mid-upper arm circumference and delayed recall.

4. The simplified multimodal index rapid auxiliary screening system for mild cognitive impairment in the elderly according to claim 1, characterized in that: The algorithm of the machine learning model is selected from the extreme gradient boosting algorithm.

5. The simplified multimodal index rapid auxiliary screening system for mild cognitive impairment in the elderly according to claim 1 or 2, characterized in that: The parameter settings of the machine learning model include: sampling rate = 1, learning rate = 0.01, maximum depth = 3, subsample = 0.

5.

6. A computer-readable storage medium, characterized in that: The computer program is stored thereon for implementing the system for rapid auxiliary screening of mild cognitive impairment in the elderly using the simple multimodal index as described in any one of claims 1 to 4.

Images