Home monitoring method for cognitive impairment of breast cancer patient

Abstract

The invention discloses a household monitoring method for cognitive impairment of a breast cancer patient, and relates to the technical field of health monitoring, and the method comprises the following steps: S1, constructing a behavior record sequence based on an intelligent interaction device; s2, recognizing a cognitive abnormal association path by using a daily behavior micro-mode group; s3, acquiring a short-time cognitive feedback log, and performing fusion analysis by using the dynamic abnormal track group and the short-time cognitive feedback log; s4, extracting multi-dimensional cognitive features by using the comprehensive cognitive state sequence; s5, matching a medical cognitive index rule set by using the layered cognitive feature matrix; and S6, performing fitting mapping on a cognitive function grading scale result by using a time sequence scoring sequence. By setting a graph embedding function and embedding a historical health behavior graph into a constructed high-dimensional structure graph space, introducing a structure distance function and combining a structure offset threshold value, the method can realize dynamic comparison and abnormal trajectory recognition of a daily behavior graph structure and generate a structure offset.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring, and in particular to a method for home monitoring of cognitive impairment in breast cancer patients. Background Art

[0002] Breast cancer patients often experience varying degrees of cognitive impairment after undergoing comprehensive interventions such as surgery, chemotherapy, and endocrine therapy, manifesting as abnormal fluctuations in attention, memory, and executive function. Numerous clinical studies have shown that cognitive impairment not only significantly impacts the quality of life of breast cancer survivors but can also be a potential precursor to later depression and social dysfunction. Therefore, cognitive function monitoring for breast cancer patients is gradually shifting from traditional centralized medical assessments to remote, continuous home monitoring.

[0003] In existing technologies, cognitive impairment monitoring in home scenarios mostly uses rule-matching or behavior counting-based methods to analyze individual behavior sequences. In addition, existing methods often use static threshold judgment and lack the ability to model temporal evolution, making it difficult to identify atypical perturbation behaviors manifested in the early stages of cognitive impairment. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides a method for home monitoring of cognitive impairment in breast cancer patients to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: In a first aspect, an embodiment of the present invention provides a method for home monitoring of cognitive impairment in breast cancer patients, comprising the following steps: S1. Construct a behavior record sequence based on intelligent interactive devices to obtain a daily behavior micro-pattern group; S2. Use the daily behavior micro-pattern group to identify cognitive abnormality association paths and obtain the dynamic abnormal trajectory group; S3. Obtain short-term cognitive feedback logs, perform fusion analysis using the dynamic abnormal trajectory group and the short-term cognitive feedback logs to obtain a comprehensive cognitive state sequence; S4. Extract multidimensional cognitive features using the comprehensive cognitive state sequence to obtain a hierarchical cognitive feature matrix; S5. Use the hierarchical cognitive feature matrix to match the medical cognitive indicator rule set to obtain a time series score sequence; S6. Use the time series scoring sequence to fit and map the cognitive function grading scale results to obtain the cognitive function grading scale output.

[0006] To further optimize this technical solution, in step S1, the intelligent interactive device deployed at home is used to synchronously collect behavior-related sensor signals, construct a multi-dimensional behavior original sequence, segment the behavior sequence using a fixed-width sliding window, and use the transfer frequency matrix between behavioral micro-patterns to determine the rationality of its timing. Step S1 finally outputs the daily behavior micro-pattern group. , is the time window span; in, Indicates that at the current moment Sliding window The micro-patterns of daily behavior formed within Indicates time After collecting and processing behavioral micro-pattern vectors; Each micro-pattern is a dimensional normalized feature vector; is a discrete time point within the time sliding window.

[0007] To further optimize this technical solution, the step S2 uses a mapping model to group daily behavior micro-patterns The micro-pattern evolution of behavior is represented as a directed graph sequence. The relative structural offset relationship between the offset mapping and the historical health behavior graph is established on the graph. The mapping model is: ; in, Indicates that at the current moment forward The set of dynamic anomaly trajectories identified within the time window; Indicates the behavioral trajectory subsegments within the observation time segment; Represents a sequence of historical health behavior micro-patterns that matches the current user's behavior characteristics; represents the graph embedding function; Embedding function representing the historical health behavior graph; Represents the high-dimensional structural distance function between graph structures; is the structural deviation threshold, which represents the judgment boundary of cognitive abnormality; Indicates the current time The time slices sampled within the sliding time window of .

[0008] Further optimize this technical solution, the graph embedding function in step S2 The graph structure constructed by behavioral micro-patterns is mapped into a vector space representation, and its function formula is: ; Indicates the A graph structure feature function, corresponding to the one constructed in Structural description indicators of the behavior graph on Indicates time slice The behavior embedding vector is used for subsequent comparison with the reference image.

[0009] Further optimize this technical solution, the function embedded in step S2 Used to construct a reference structure, its function formula is: ; Indicates time In the time window, the historical health behavior samples have the following characteristics: The statistical mean on : , represents the number of historical health behavior trajectory samples, Indicates the Historical health behavior samples in time slices Micro-mode input.

[0010] Further optimizing this technical solution, the distance function in step S2 is The formula is: .

[0011] Further optimize this technical solution, the structural offset threshold in step S2 The calculation formula is: ; Represents the structural offset sensitivity factor, which needs to be tuned through the validation set.

[0012] To further optimize this technical solution, step S3 adopts a cognitive state modulation model, and the model formula is as follows: ; in, : Current time The fused cognitive state expression of represents the aggregate structure of cognitive shift responses; : Multidimensional response variation mapping function composed of heterogeneous data pairs; :Represents the trajectory path Synchronous alignment feedback items from the cognitive feedback log In the time point Short-term language / behavioral cognitive abnormality feedback encoding in the form of vectors or label sets; ; :by A sliding time window centered on is half the width of the window; : A fusion transformation function used to map raw log entries into feedback vectors with a unified structure.

[0013] To further optimize this technical solution, step S4 performs structural deconstruction and feature extraction on the multi-channel cognitive information contained in the time series expression obtained in step S3, generating a cognitive feature matrix with a hierarchical structure for subsequent time series recognition or state reasoning, and finally outputting a three-dimensional structure tensor. , each For in time The hierarchical cognitive feature matrix extracted under different time periods.

[0014] Further optimize this technical solution, in step S5, a cognitive indicator rule set is constructed based on the quantitative indicators formulated by cognitive medicine literature or experts. , for each time period matrix Read rows and columns and standardize the dimensions in the rule set With the actual eigenvalue Compare and form a local score vector ,in To identify the number of feature dimensions, a rule tree structure is used to evaluate the local scoring vector Aggregate into final score , and finally get the complete temporal score sequence , each of which , indicating the Cognitive scores for each period.

[0015] In a second aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of a method for home monitoring cognitive impairment in breast cancer patients as described in the first aspect of the present invention are implemented.

[0016] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of a method for home monitoring cognitive impairment in breast cancer patients as described in the first aspect of the present invention are implemented.

[0017] Compared with the existing technology, the present invention provides a method for home monitoring of cognitive impairment in breast cancer patients, which has the following beneficial effects: This home monitoring method for cognitive impairment in breast cancer patients uses a graph embedding function to construct a high-dimensional structured graph space, introduces a structural distance function, and combines it with a structural offset threshold. This method dynamically compares the structure of daily behavior graphs, identifies abnormal trajectories, and generates structural offsets. Compared to traditional behavioral feature counting methods, this method offers the ability to quantitatively analyze microstructural offsets in behavioral sequences, improving the sensitivity of detecting abnormal pathways in the early stages of cognitive impairment and the foresight of system responses. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only 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.

[0019] Figure 1 This is a flow chart of a method for home monitoring of cognitive impairment in breast cancer patients proposed by the present invention; Figure 2 This is a schematic diagram of the process of identifying cognitive abnormality-related pathways in a method for home monitoring of cognitive impairment in breast cancer patients proposed by the present invention; Figure 3 This is a schematic diagram of the use flow of the cognitive state modulation model of the home monitoring method for cognitive impairment in breast cancer patients proposed by the present invention. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive with other embodiments.

[0023] Example 1: Reference Figures 1 to 3 , which is the first embodiment of the present invention, provides a method for home monitoring of cognitive impairment in breast cancer patients, comprising the following steps: S1. Construct a behavior record sequence based on intelligent interactive devices to obtain a daily behavior micro-pattern group; In step S1, the intelligent interactive devices deployed at home (including smart bracelets, cameras, kitchen scales, and door sensors) are used to synchronously collect behavior-related sensor signals. The mature feature channel normalization technology is used to uniformly convert the individual channel signals into a [0,1] interval time series to construct a multi-dimensional behavior original sequence. Use a fixed-width sliding window to segment the behavior sequence. In each sliding window segment, use a clustering-based pattern extraction algorithm to cluster the multi-dimensional behavior vector segments. Encode the clustering results into pattern labels to form a behavioral micro-pattern sequence under the time index. The transition frequency matrix between behavioral micro-patterns is used to determine the rationality of their timing, and the mature Markov smoothing correction algorithm is applied to smooth the trajectory of the behavioral micro-pattern sequence. Step S1 finally outputs the daily behavior micro-pattern group , is the time window span; in, Indicates that at the current moment Sliding window The micro-patterns of daily behavior formed within Indicates time After collecting and processing behavioral micro-pattern vectors; Each micro-pattern is a dimensional standardized feature vector, including action type, position status, interaction duration, and physiological indicator interaction value. These indicator data are obtained through smart bracelets, cameras, kitchen scales, and door magnetic devices; is a discrete time point within the time sliding window, used to construct a sequence.

[0024] S2. Use the daily behavior micro-pattern group to identify cognitive abnormality association paths and obtain the dynamic abnormal trajectory group; Step S2 uses a mapping model to group daily behavior micro-patterns The micro-pattern evolution of behavior is represented as a directed graph sequence. The relative structural offset relationship between the offset mapping and the historical health behavior graph is established on the graph. The mapping model is: ; in, Indicates that at the current moment forward The set of dynamic anomaly trajectories identified within the time window; Indicates the behavioral trajectory subsegments within the observation time segment; Represents a sequence of historical health behavior micro-patterns that matches the current user's behavior characteristics; Represents the graph embedding function, which maps the graph structure constructed by behavioral micro-patterns into a vector space representation. Its function formula is: ; Indicates the A graph structure feature function, corresponding to the one constructed in Structural description indicators of the behavior graph on Indicates time slice The behavior embedding vector is used for subsequent comparison with the reference graph; The embedding function representing the historical health behavior graph is used to construct the reference structure. Its function formula is: ; Indicates time In the time window, the historical health behavior samples have the following characteristics: The statistical mean on : , represents the number of historical health behavior trajectory samples, Indicates the Historical health behavior samples in time slices Micro-mode input; Represents the high-dimensional structural distance function between graph structures, and its function formula is: ; is the structural deviation threshold, which represents the judgment boundary of cognitive abnormality; ; Represents the structural offset sensitivity factor, which needs to be tuned through the validation set; Indicates the current time The time slices sampled within the sliding time window of .

[0025] The mapping model includes: Behavior diagram construction: The continuous micro-pattern sequence in is segmented into sub-sequences in a fixed time window, and a micro-pattern transition directed graph corresponding to each segment is constructed; Graph embedding calculation: call the function for each subgraph , generating vectorized representations through a structure embedding method based on path compression; Structural offset calculation: The current graph embedding of each segment is calculated with the corresponding graph embedding of historical healthy users, and the relative structural offset is measured using the distance function. Abnormal trajectory screening: Any segment with a deviation greater than the threshold is considered a potential cognitive abnormality trajectory segment, and its original trajectory segment is collected as one of the items.

[0026] Step S2 includes the following process in the process of identifying cognitive abnormality association pathways: Sub-trajectory generation; right Use fixed window sliding to obtain multiple subsequences , each segment represents a local behavior evolution process; Micro-pattern construction; Each subsequence is constructed into a directed behavior graph: nodes are micro-patterns, edges are temporal transitions, and edge weights are transition frequencies; Graph embedding representation generation; Using graph embedding functions Convert each behavior graph into a fixed-length embedding vector to reflect its structural information; Construction and embedding of historical health comparison charts; Match the corresponding window cycle behavior graph from the historical healthy population database , calculate its embedding ; Structural deviation determination and screening; Compare the embedding distance between the current and historical graphs using the distance function , if the difference exceeds the threshold , then keep the sub-trajectory segment until , which is the dynamic abnormal trajectory group finally output by this step.

[0027] In step S2, we construct an embedding function for the health behavior graph. , combined with historical graph embedding Distance function with high-dimensional structure , enabling monitoring of the structural evolution of behavior sequence graphs. This method significantly differs from existing pattern recognition methods that rely solely on behavioral event frequency or anomaly score calculations. The latter are often based on statistical feature extraction or time series similarity (such as DTW and HMM). This step quantifies graph pattern deviations in the structural space, providing stronger anomalous pattern structure recognition capabilities and is particularly suitable for scenarios where behavioral topology changes are sensitive.

[0028] S3. Obtain short-term cognitive feedback logs, perform fusion analysis using the dynamic abnormal trajectory group and the short-term cognitive feedback logs to obtain a comprehensive cognitive state sequence; Step S3 uses the cognitive state modulation model to fuse the dynamic abnormal trajectory group obtained in step S2 with the feedback of the actual cognitive response. ,establish anomaly trigger response mutation mapping mechanism; Feedback on actual cognitive responses It mainly comes from the interaction records of breast cancer patients in their home environment oriented towards cognitive status. These records are obtained through the smart bracelets and cameras involved in step S1. These data are generated at the home end through the technical means of edge device inference and local log packaging. The existing real-time event stream analysis technology, fuzzy semantic recognition rule library and Android mobile data collection SDK toolkit are used to generate log data grouped by timestamp and source channel to form a data structure. A collection of constant organization : The cognitive state modulation model formula is as follows: ; in, : Current time The fused cognitive state expression of represents the aggregate structure of cognitive shift responses; : Multidimensional response variation mapping function composed of heterogeneous data pairs; :Represents the trajectory path Synchronous alignment feedback items from the cognitive feedback log In the time point Short-term language / behavioral cognitive abnormality feedback encoding in the form of vectors or label sets; ; :by A sliding time window centered on is half the width of the window; : A fusion transformation function used to map raw log entries into feedback vectors with a unified structure.

[0029] The cognitive state modulation model in step S3, when used, includes: Locating abnormal trajectory time segments : Output from S2 Each of represents a time segment identified as abnormal in behavioral structure; Extracting structural offset strength indicators : Calculated in step S2 , which is the embedding distance between the current behavior graph and the historical health graph; this value represents the deviation amplitude of the abnormal behavior; Synchronous alignment short-term feedback log :For each time slice ,exist Extracting synchronous or temporally adjacent feedback data; Combining input pairs : The structural deviation intensity and cognitive response are matched one by one, and the cognitive deviation response associated with each type of abnormality is integrated; Aggregate inference function , the final output .

[0030] Step S3 constructs the fusion dynamic trajectory group Short-term cognitive feedback log This cognitive state modulation model introduces a fusion transformation function to achieve continuous expression of state representation. Unlike existing multimodal fusion methods such as principal component analysis (PCA) or feature splicing, this method incorporates the dynamic evolution of cognitive trajectories and the synergistic constraints of cognitive language feedback into its model structure. It preserves temporal relationships and semantic associations, forming a cognitive state representation with greater semantic depth and time sensitivity, making it advantageous for detecting subtle fluctuations in cognitive function.

[0031] S4. Extract multidimensional cognitive features using the comprehensive cognitive state sequence to obtain a hierarchical cognitive feature matrix; Step S4 converts the time series expression obtained in step S3 into The multi-channel cognitive information contained in the structured deconstruction and feature extraction is performed to generate a cognitive feature matrix with a hierarchical structure for subsequent time series recognition or state reasoning. The extraction process includes: Time dimension segmentation; Using fixed sliding window right The sequence is segmented by sliding; In each time segment Classify them into the same time domain feature processing unit to ensure time consistency; Channel mode separation; right The different types of information fused in the channel are processed separately, typically using a modality separation network decoding structure or multi-modal tensor expansion technology; The separated channel information includes: behavioral trajectory dimension features , physiological feedback dimension characteristics , visual feedback dimension features ; Use mature methods such as tensor multimodal decomposition to map its structure; Cognitive semantic extraction; On the basis of modality separation, the modality features are further mapped to standard cognitive features such as "attention intensity", "reaction time delay", and "behavior interruption probability" through structural semantic extraction algorithm; The output of this stage is a set of standard cognitive indicators under each modal channel, forming a basic cognitive vector unit.

[0032] Hierarchical matrix construction; The above standard cognitive features are stacked in chronological order to form a two-dimensional structure; Then, the column vector regions are divided according to the cognitive dimensions to form the final hierarchical cognitive feature matrix; The hierarchical structure adopts a matrix arrangement in which rows represent time series slices and columns represent standard feature sets under cognitive dimensions. The matrix size is related to the window step size and the number of modalities.

[0033] Step S4 finally outputs the three-dimensional structure tensor , each For in time The hierarchical cognitive feature matrix extracted under different time periods.

[0034] S5. Use the hierarchical cognitive feature matrix to match the medical cognitive indicator rule set to obtain a time series score sequence; In step S5, a cognitive indicator rule set is constructed based on the quantitative indicators developed by cognitive medicine literature or experts. , including thresholds, scoring criteria, and grading intervals under different dimensions. The rule set format is a table or JSON structure, and is normalized using existing literature data; For each time period matrix Read rows and columns, extract each standard feature value according to the cognitive dimension, and set the dimension standard in the rule set With the actual eigenvalue For comparison, the interval mapping scoring method is used to find the scoring interval of each dimension feature value and map it to the corresponding scoring value. Each feature dimension is scored separately, and the result forms a local scoring vector ,in is the number of cognitive feature dimensions; Use a regular tree structure to evaluate the local score vector Aggregate into final score , and finally get the complete temporal score sequence , each of which , indicating the Cognitive scores for each period.

[0035] S6. fitting and mapping the cognitive function grading scale results using the time series scoring sequence to obtain the cognitive function grading scale output; Step S6 constructs a high-order transformation mapping function to achieve structural fitting between the scoring sequence and the predicted scale score; Its function model is: ; in, : the change in adjacent scores; : Memory enhancement term, the weighted history sum of the ratings within the sliding window; : Model structure coefficients and nonlinear mapping strength parameters were calibrated by clinical experts; : Hypertangent activation function, used for nonlinear compression and mapping to psychological scale score intervals.

[0036] In this formula: is a scoring base mapping item, which is used to take the basic scoring value as the main input for cognitive level prediction; The score change rate term is introduced to measure the sensitivity of cognitive score changes in the time dimension, which is used to fit the responsiveness of clinical cognitive scales to the "degradation trend"; As a historical weighted influence term, the long-term cognitive score trend is introduced as a steady-state criterion; is the offset, which controls the starting score of the baseline scale and aligns it with the reference level of cognitive health.

[0037] The fitting mapping process in step S6 includes: Based on the output of step S5 , using sliding window and first-order difference technology to construct historical change characteristics ; Adjust model parameters based on historical samples or expert experience ; The prediction results are calculated by high-order transformation mapping functions , for all After calculation, the output sequence is obtained .

[0038] Step S6 proposes a cognitive function prediction model based on the combination of score change rate, historical feature compression term and nonlinear hypertangent mapping function, and introduces a dynamic change rate term to the score sequence. and historical impact items , and constructs an asymmetric compression mapping through adjustable hyperparameters. Unlike traditional linear regression or neural network regression methods, this model offers structural interpretability and can control the nonlinear trends of the prediction curve through parameter adjustment. This makes it more suitable for modeling the gradual evolution of scale results in medical scenarios, improving sensitivity to abnormal score changes.

[0039] Example 2:

[0040] This embodiment also provides a computer device suitable for a method for home monitoring of cognitive impairment in breast cancer patients, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a method for home monitoring of cognitive impairment in breast cancer patients as proposed in the above embodiment.

[0041] This embodiment also provides a storage medium having a computer program stored thereon. When the program is executed by a processor, the method for home monitoring of cognitive impairment in breast cancer patients proposed in the above embodiment is implemented.

[0042] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0043] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0044] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0045] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0046] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0047] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for home monitoring of cognitive impairment in breast cancer patients, characterized in that: The following steps are involved: S1. Construct a behavior record sequence based on intelligent interactive devices to obtain a daily behavior micro-pattern group; S2. Use the daily behavior micro-pattern group to identify cognitive abnormality association paths and obtain the dynamic abnormal trajectory group; S3. Obtain short-term cognitive feedback logs, perform fusion analysis using the dynamic abnormal trajectory group and the short-term cognitive feedback logs to obtain a comprehensive cognitive state sequence; S4. Extract multidimensional cognitive features using the comprehensive cognitive state sequence to obtain a hierarchical cognitive feature matrix; S5. Use the hierarchical cognitive feature matrix to match the medical cognitive indicator rule set to obtain a time series score sequence; S6. Use the time series scoring sequence to fit and map the cognitive function grading scale results to obtain the cognitive function grading scale output.

2. A method for home monitoring of cognitive impairment in breast cancer patients according to claim 1, characterized in that: In step S1, the intelligent interactive devices deployed at home are used to synchronously collect behavior-related sensor signals, construct a multi-dimensional behavior original sequence, segment the behavior sequence using a fixed-width sliding window, and use the transfer frequency matrix between behavior micro-patterns to determine the rationality of its timing. Step S1 finally outputs the daily behavior micro-pattern group. , is the time window span; in, Indicates that at the current moment Sliding window The micro-patterns of daily behavior formed within Indicates time After collecting and processing behavioral micro-pattern vectors; Each micro-pattern is a dimensional normalized feature vector; is a discrete time point within the time sliding window.

3. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 1, characterized in that: The step S2 uses a mapping model to group daily behavior micro-patterns The micro-pattern evolution of behavior is represented as a directed graph sequence. The relative structural offset relationship between the offset mapping and the historical health behavior graph is established on the graph. The mapping model is: ； in, Indicates that at the current moment forward The set of dynamic anomaly trajectories identified within the time window; Indicates the behavioral trajectory subsegments within the observation time segment; Represents a sequence of historical health behavior micro-patterns that matches the current user's behavior characteristics; represents the graph embedding function; Embedding function representing the historical health behavior graph; Represents the high-dimensional structural distance function between graph structures; is the structural deviation threshold, which represents the judgment boundary of cognitive abnormality; Indicates the current time The time slices sampled within the sliding time window of .

4. A method for home monitoring of cognitive impairment in breast cancer patients according to claim 3, characterized in that: In step S2, the graph embedding function The graph structure constructed by behavioral micro-patterns is mapped into a vector space representation, and its function formula is: ； Indicates the A graph structure feature function, corresponding to the one constructed in Structural description indicators of the behavior graph on Indicates time slice The behavior embedding vector is used for subsequent comparison with the reference image.

5. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 3, characterized in that: The function embedded in step S2 Used to construct a reference structure, its function formula is: ； in, Indicates time In the time window, the historical health behavior samples have the following characteristics: The statistical mean on : , represents the number of historical health behavior trajectory samples, Indicates the Historical health behavior samples in time slices Micro-mode input.

6. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 3, characterized in that: The distance function in step S2 The formula is: 。 7. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 3, characterized in that: The structural offset threshold in step S2 The calculation formula is: ； Represents the structural offset sensitivity factor, which needs to be tuned through the validation set.

8. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 1, characterized in that: The step S3 adopts the cognitive state modulation model, and the model formula is as follows: ； in, : Current time The fused cognitive state expression of represents the aggregate structure of cognitive shift responses; : Multidimensional response variation mapping function composed of heterogeneous data pairs; :Represents the trajectory path Synchronous alignment feedback items from the cognitive feedback log In the time point Short-term language / behavioral cognitive abnormality feedback encoding in the form of vectors or label sets; ； :by A sliding time window centered on is half the width of the window; : A fusion transformation function used to map raw log entries into feedback vectors with a unified structure.

9. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 1, characterized in that: Step S4 performs structural deconstruction and feature extraction on the multi-channel cognitive information contained in the time series expression obtained in step S3 to generate a cognitive feature matrix with a hierarchical structure for subsequent time series recognition or state reasoning, and finally outputs a three-dimensional structure tensor. , each For in time The hierarchical cognitive feature matrix extracted under different time periods.

10. The method for home monitoring of cognitive impairment in breast cancer patients according to claim 1, characterized in that: In step S5, a cognitive indicator rule set is constructed based on the quantitative indicators formulated by cognitive medicine literature or experts. , for each time period matrix Read rows and columns and standardize the dimensions in the rule set With the actual eigenvalue Compare and form a local score vector ,in To identify the number of feature dimensions, a rule tree structure is used to evaluate the local scoring vector Aggregate into final score , and finally get the complete temporal score sequence , each of which , indicating the Cognitive scores for each period.

Images