An AI-based system for individualized parameter optimization in transcranial magnetic stimulation therapy

Abstract

This invention discloses an artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation (TMS) therapy, belonging to the technical field of TMS optimization systems. This AI-based individualized parameter optimization system for TMS therapy includes a data acquisition and preprocessing module, a feature extraction module, a treatment parameter prediction module, and a treatment effect evaluation module. The data acquisition module is used to collect the patient's brain function data in real time; the feature extraction module is used to extract features from the preprocessed brain function data and clinical data; the treatment parameter prediction module is used to determine a treatment plan suitable for the patient's current condition based on the predicted TMS treatment parameters; and the treatment effect evaluation module is used to continuously monitor the patient's brain function data during treatment. This system solves the problems of traditional TMS therapy, where parameter settings rely on experience and generalized guidelines, lack individualized optimization, and require parameter adjustments, leading to inaccurate treatment effects and long treatment times.

Description

Technical Field

[0001] This invention relates to the field of transcranial magnetic stimulation optimization system technology, specifically to an artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy. Background Technology

[0002] Transcranial magnetic stimulation (TMS) is a non-invasive neuromodulation technique that uses magnetic fields to induce electrical currents, activating neurons in specific areas of the brain. This technique is widely used to treat depression, anxiety, and other neuropsychiatric disorders. While TMS has achieved some success in clinical applications, its effectiveness largely depends on the precise setting of treatment parameters, such as the target site, frequency, and intensity of stimulation. Optimizing these parameters is crucial, as inappropriate settings can lead to poor treatment outcomes or a negative patient experience.

[0003] Traditional transcranial magnetic stimulation (TMS) treatment parameter settings largely rely on experience and generalized clinical guidelines, lacking customization for the specific needs of individual patients. Furthermore, parameter determination often requires multiple trials and adjustments, which is not only time-consuming but also increases the treatment burden on patients. Therefore, developing a system capable of automating, accurately predicting, and adjusting treatment parameters is of paramount importance.

[0004] Furthermore, although existing technologies employ various biofeedback mechanisms to monitor treatment effectiveness, such as electroencephalography (EEG) and functional near-infrared spectroscopy, which provide valuable brain activity data, translating this data into specific, optimized treatment parameter adjustments remains a technical challenge.

[0005] Therefore, in order to address the above problems, there is an urgent need for an artificial intelligence-based system for optimizing individualized parameters of transcranial magnetic stimulation therapy. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides an artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation (TMS) therapy. This system solves the problems of traditional TMS therapy, which relies on experience and general guidelines for parameter settings, lacks individualized optimization and parameter adjustment, resulting in inaccurate treatment effects and long treatment times.

[0008] Technical solution

[0009] To achieve the above objectives, the present invention provides the following technical solution: an artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation (TMS) therapy, comprising a data acquisition and preprocessing module, a feature extraction module, a treatment parameter prediction module, and a treatment effect evaluation module. The data acquisition module is used to collect the patient's brain function data in real time, obtain target weights, stimulation frequency, and stimulation intensity, acquire clinical data, and preprocess the brain function data and clinical data. The feature extraction module is used to extract features from the preprocessed brain function data and clinical data, and mark the extracted data as feature data. The treatment parameter prediction module is used to predict TMS treatment parameters based on the feature data, and determine a treatment plan suitable for the patient's current condition based on the predicted TMS treatment parameters. The treatment effect evaluation module is used to continuously monitor the patient's brain function data during treatment, obtain brain function effects, evaluate treatment effects, and make corresponding optimization and adjustment measures based on the treatment effects.

[0010] Furthermore, the real-time acquisition of the patient's brain function data, obtaining target weights, stimulation frequency, and stimulation intensity, and the specific process of acquiring clinical data are as follows: Brain function data includes electroencephalogram (EEG) data and brain region blood oxygen concentration change data. EEG data is acquired through an EEG sensor, and brain region blood oxygen concentration change data is acquired through an fNIRS sensor. The activity of relevant brain regions is detected through EEG and functional near-infrared spectroscopy. The left prefrontal cortex stimulation target, the right prefrontal cortex stimulation target, and the medial prefrontal cortex stimulation target are selected. The appropriate stimulation frequency and intensity are selected based on the patient's EEG analysis results. Clinical data is automatically acquired through user input.

[0011] Furthermore, the specific process of preprocessing the brain function data and clinical data is as follows: the brain function data and clinical data are denoised and standardized, and the brain function data, clinical data and clinical data are merged to generate the patient's brain function state characteristics.

[0012] Furthermore, the specific process of feature extraction from the preprocessed brain function data and clinical data is as follows: extracting the alpha wave frequency of the occipital lobe from the electroencephalogram (EEG) data; extracting the blood oxygen concentration of the left prefrontal lobe, the blood oxygen concentration of the right prefrontal lobe, the phase changes of the left prefrontal cortex, and the phase changes of the right prefrontal cortex from the brain region blood oxygen concentration change data; obtaining the patient's clinical feedback and disease history information from the clinical data; and recording treatment progress data from the clinical records.

[0013] Furthermore, the specific process of predicting transcranial magnetic stimulation (TMS) treatment parameters based on feature data and determining a suitable treatment plan for the patient's current condition based on the predicted TMS treatment parameters is as follows: The stimulation target points for the left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex are obtained, along with the stimulation frequency, stimulation intensity, blood oxygen concentration in the left and right prefrontal cortex, and the temporal changes in the left and right prefrontal cortex. The predicted treatment parameters are then calculated by comprehensively considering these parameters. Finally, a treatment plan suitable for the patient's current condition is determined based on the predicted treatment parameters.

[0014] Furthermore, the specific process of determining a treatment plan suitable for the patient's current condition based on the predicted values ​​of treatment parameters is as follows: determining a treatment plan suitable for the patient's current condition based on the predicted values ​​of treatment parameters; selecting a suitable stimulation target for the patient based on the predicted stimulation target; setting the frequency of the transcranial magnetic stimulation device based on the predicted stimulation frequency; and adjusting the intensity of the transcranial magnetic stimulation device based on the predicted stimulation intensity.

[0015] Furthermore, the specific calculation method for the predicted value of the treatment parameter is as follows: In the formula T p This represents the predicted value of the treatment parameter. This represents a weighted sum of the stimulation targets in the left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex, w. i α represents the weight coefficient of the i-th target point. i Let f represent the exponent of the i-th target, w4 represent the stimulation frequency, I represent the stimulation intensity, w5 represent the stimulation intensity weighting coefficient, and O represent the stimulation intensity weighting coefficient. left This indicates the blood oxygen concentration in the left prefrontal cortex, O. right The value represents the blood oxygen concentration in the right prefrontal cortex, w6 represents the weighting coefficient of the product of the blood oxygen concentrations in the left and right prefrontal cortexes, and ΔO left This represents the temporal changes in the left prefrontal cortex, ΔO. right w7 represents the temporal changes in the right prefrontal cortex, and w7 represents the weighting coefficient of the product of the temporal changes in cerebral blood oxygen concentration in the left and right prefrontal lobes.

[0016] Furthermore, the specific process of evaluating the treatment effect and making corresponding optimization and adjustment measures based on the treatment effect is as follows: obtaining clinical feedback, disease history information, treatment progress data and brain function effects; comprehensively calculating the treatment effect evaluation value through clinical feedback, disease history information, treatment progress data and brain function effects; and making corresponding optimization and adjustment measures based on the treatment effect evaluation value.

[0017] Furthermore, the specific process of making corresponding optimization and adjustment measures based on the treatment effect evaluation value is as follows: compare the treatment effect evaluation value with the threshold in real time. When the treatment effect evaluation value is greater than the threshold, no adjustment is made. When the treatment effect evaluation value is less than or equal to the threshold, adjust the specific location of the target point, adjust the stimulation frequency, and adjust the stimulation intensity until the treatment effect evaluation value is raised to above the threshold.

[0018] Furthermore, the specific calculation method for the treatment effect evaluation value is as follows: In the formula E t R represents the evaluation value of treatment effectiveness. c Indicates clinical feedback, w c R represents the weighting coefficient of clinical feedback. f Indicates disease history information, w f B represents the weighting coefficient for disease history information. e Indicates brain function effect, w b The weighting coefficient T represents the effect on brain function. prog Indicates treatment progress data, w T Weighting coefficients representing treatment progress data.

[0019] Beneficial effects

[0020] The present invention has the following beneficial effects:

[0021] (1) This invention achieves precise formulation of individualized treatment plans by collecting patients' brain function data and clinical data in real time and combining them with artificial intelligence models to predict the most suitable transcranial magnetic stimulation (TMS) treatment parameters. This enables the tailoring of treatment plans for each patient, effectively addressing the shortcomings of existing TMS treatments that rely on experience and lack individualized adjustments.

[0022] (2) This invention, by combining electroencephalogram (EEG) data, brain region oxygenation concentration change data, and patient clinical feedback, enables the system to accurately predict and adjust transcranial magnetic stimulation (TMS) treatment parameters, thereby achieving precise matching between treatment targets and stimulation frequencies. This improves treatment precision and effectively solves the problem that existing TMS treatments fail to adequately consider individual patient differences.

[0023] (3) This invention automatically provides adjustment suggestions through a treatment effect evaluation module. This achieves real-time evaluation and dynamic feedback of treatment effects, effectively solving the problems of delayed treatment effect evaluation and lack of timely adjustment mechanisms in existing technologies.

[0024] (4) This invention, by combining EEG data, brain region blood oxygen concentration change data, and patient clinical feedback, enables the system to accurately predict and adjust transcranial magnetic stimulation (TMS) treatment parameters, thereby achieving precise matching between treatment targets and stimulation frequencies. This improves treatment precision and effectively solves the problem in existing TMS treatments that fail to adequately consider individual patient differences.

[0025] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0026] Figure 1 This is a structural diagram of an artificial intelligence-based transcranial magnetic stimulation therapy individualized parameter optimization system according to the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Please see Figure 1 This invention provides a technical solution: an artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation (TMS) therapy, comprising a data acquisition and preprocessing module, a feature extraction module, a treatment parameter prediction module, and a treatment effect evaluation module. The data acquisition module collects the patient's brain function data in real time, obtains target weights, stimulation frequency, and stimulation intensity, acquires clinical data, and preprocesses the brain function data and clinical data. The feature extraction module extracts features from the preprocessed brain function data and clinical data, and marks the extracted data as feature data. The treatment parameter prediction module predicts TMS treatment parameters based on the feature data and determines a suitable treatment plan for the patient's current condition based on the predicted TMS treatment parameters. The treatment effect evaluation module continuously monitors the patient's brain function data during treatment, obtains brain function effects, evaluates the treatment effect, and makes corresponding optimization adjustments based on the treatment effect.

[0029] Specifically, the system collects real-time brain function data from patients to obtain target weights, stimulation frequency, and stimulation intensity. The clinical data acquisition process is as follows: First, brain function data consists of electroencephalogram (EEG) data and brain region oxygenation concentration change data. Using an EEG sensor, the system can collect the patient's EEG data in real time, reflecting the brain's electrical activity, particularly changes in brain waves in different brain regions, thus providing fundamental information for treatment. Simultaneously, an fN IRS sensor is used to collect brain region oxygenation concentration change data, enabling real-time monitoring of fluctuations in brain oxygenation levels and reflecting changes in local brain activity and oxygenation status. These data, through EEG and functional near-infrared spectroscopy, collaboratively monitor the activity of different brain regions, providing a scientific basis for subsequent treatment target selection and parameter setting. Regarding target selection, the system determines the most suitable stimulation targets based on the collected EEG and oxygenation concentration change data, typically including three key brain regions: the left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex, to ensure treatment accuracy and effectiveness. Stimulation frequency and intensity are selected based on a thorough interpretation of the patient's EEG analysis results and their brain response characteristics. In addition, the system automatically collects clinical data through user input, such as patient medical history, feedback information, and treatment progress. The comprehensive collection and analysis of this data provides a complete and accurate basis for subsequent optimization of individualized transcranial magnetic stimulation (TMS) treatment parameters, thereby enabling more refined and intelligent treatment plan design.

[0030] In this implementation plan, by collecting patients' brain function and clinical data in real time, including electroencephalogram (EEG) data, changes in brain region blood oxygen concentration, and relevant clinical feedback information, this step can comprehensively and accurately assess the patient's brain function status. Using EEG and fNIRS sensors to collect EEG and blood oxygenation data provides a reliable basis for subsequent target selection and stimulation parameter optimization. Through the interpretation of EEG analysis results, the system can accurately select stimulation targets and automatically determine the stimulation frequency and intensity based on the brain's electrical activity characteristics, ensuring individualized and effective treatment. Simultaneously, the system collects clinical data in real time through user input, further enriching the patient's treatment information. This process ensures the comprehensiveness and accuracy of data collection, laying the foundation for the development of personalized transcranial magnetic stimulation (TMS) treatment plans.

[0031] Specifically, the preprocessing of brain function data and clinical data involves the following steps: First, brain function data, such as data collected via EEG and fNIRS, is often affected by various interferences and noises, such as interference from electronic devices or noise generated by physiological factors. The denoising step effectively removes these non-specific signals by applying various digital filtering techniques, improving the signal-to-noise ratio of the data. Subsequently, standardization is used to adjust the brain function data and clinical data to a uniform scale, which is crucial for comparing data from different time points or different patients. Standardization ensures data consistency, making data from different patients and under different conditions comparable. Furthermore, the preprocessing process includes merging brain function data and clinical data, such as patient medical history and treatment response. This step involves integrating information from multiple sources to generate a comprehensive profile of the patient's brain function status.

[0032] In this implementation plan, the system significantly improves data quality and ensures the reliability and consistency of collected information through denoising and standardization during the preprocessing of brain function and clinical data. Denoising effectively removes interfering signals and reduces errors, thereby improving the signal-to-noise ratio of EEG and blood oxygenation concentration change data, making the data more accurately reflect the patient's true brain function state. Standardization enables data from different patients and at different time points to be compared and analyzed under the same standard, which is particularly important for cross-patient analysis and long-term treatment efficacy assessment. By integrating and merging processed brain function and clinical data, the system can generate comprehensive patient brain function state characteristics, providing a solid foundation for subsequent feature extraction and accurate prediction of treatment parameters. This step not only improves data processing efficiency but also provides crucial support for achieving personalized and precise treatment plans.

[0033] Specifically, the feature extraction process for preprocessed brain function and clinical data involves the following steps: During feature extraction, the system uses advanced data analysis techniques to extract key information from the electroencephalogram (EEG) data, such as the alpha wave frequency in the occipital lobe. This frequency is typically associated with the brain's relaxation and resting states and is crucial for understanding the patient's neurophysiological state. Furthermore, the system analyzes blood oxygenation changes in detail from functional near-infrared spectroscopy (FIR), particularly in the left and right prefrontal lobes, areas that play key roles in emotion regulation and cognitive function. By monitoring blood oxygenation concentration and its temporal changes in the left and right prefrontal lobes, the system can accurately assess activity changes in these brain regions during treatment, providing a scientific basis for the selection of treatment targets. Simultaneously, the feature extraction process includes in-depth analysis of the patient's clinical data, such as clinical feedback and disease history. This information is obtained directly from patient or medical records and transformed into useful treatment reference data through the system's intelligent analysis. Treatment progress information from the clinical data is also recorded and analyzed, which is crucial for evaluating treatment effectiveness and making further treatment decisions. This comprehensive feature extraction process not only optimizes the selection and adjustment of treatment parameters, but also greatly enhances the individualization and scientific nature of treatment plans by systematically analyzing and integrating multi-source data, ensuring that each patient can receive the treatment most suitable for their specific condition.

[0034] In this implementation plan, by extracting features from preprocessed brain function and clinical data, the system can deeply analyze and uncover key brain functions and clinical characteristics of patients, providing necessary information for precise adjustment and personalized treatment. During feature extraction, the alpha wave frequency of the occipital lobe is extracted from the electroencephalogram (EEG) data, which helps assess the patient's relaxation state and baseline brain activity level. Simultaneously, by analyzing the blood oxygen concentration and its temporal changes in the left and right prefrontal lobes, the system can gain a detailed understanding of the functional status of these key brain regions in cognitive and emotional regulation. This information is crucial for identifying the most effective transcranial magnetic stimulation (TMS) targets. Furthermore, patient clinical feedback, disease history information, and treatment progress data extracted from clinical data further enhance the clinical basis for treatment decisions, ensuring that treatment measures closely correspond to the patient's specific needs and responses. This process, through systematic data analysis and feature recognition, significantly improves the scientific rigor and personalization of treatment plans, thereby achieving more precise and effective treatment interventions for patients.

[0035] Specifically, the process of predicting transcranial magnetic stimulation (TMS) treatment parameters based on feature data and determining a suitable treatment plan for the patient's current condition based on the predicted TMS treatment parameters is as follows: First, key input data for treatment is acquired: stimulation targets in the left, right, and medial prefrontal lobes, stimulation frequency and intensity, as well as blood oxygen concentration and temporal changes in the cortex of the left and right prefrontal lobes. This data not only reflects the physiological and functional state of the patient's brain but also indicates the activity and responsiveness of potential treatment targets. An algorithm comprehensively considers these features, including blood oxygen concentration and temporal changes in the left and right prefrontal lobes, as well as EEG activity characteristics related to the treatment targets, to calculate predicted treatment parameter values. This calculation process utilizes a machine learning model, combined with historical data and treatment responses from similar cases, to optimize and personalize treatment parameters. By comprehensively analyzing data on stimulation targets in the left, right, and medial prefrontal lobes, as well as stimulation frequency, intensity, oxygen saturation in the left and right prefrontal lobes, and temporal changes in the left and right prefrontal cortex, the system can accurately predict the most effective treatment parameters. Finally, based on these predicted treatment parameter values, a treatment plan suitable for the patient's current condition is determined.

[0036] In this implementation plan, by comprehensively utilizing key features extracted from patient brain function data and clinical data, the system can accurately predict transcranial magnetic stimulation (TMS) treatment parameters suitable for the patient's current condition. The system's integrated machine learning model uses blood oxygen concentration and temporal changes in the left and right prefrontal lobes, as well as EEG activity data, to calculate the treatment parameters. This method allows the system to not only consider target selection but also adjust the stimulation frequency and intensity based on the patient's actual brain response, ensuring personalized and optimized treatment. Through this process, the system can customize a treatment plan that precisely matches the patient's pathological state and brain function characteristics, greatly improving the targeting and efficiency of treatment. This not only enhances the clinical efficacy of treatment but also reduces potential time delays and additional costs during trial and error, effectively addressing the shortcomings of traditional methods in setting individualized treatment parameters.

[0037] Specifically, the process of determining a suitable treatment plan based on predicted treatment parameters is as follows: First, the treatment plan is based on predicted treatment parameters generated by a machine learning model. These values ​​integrate the patient's brain function and clinical data, providing personalized treatment input. The system selects the most suitable stimulation target based on these predictions for the patient's current neurophysiological state. For example, if the prediction indicates that the left prefrontal cortex is a more suitable target, the system will automatically adjust the positioning of the transcranial magnetic stimulation (TMS) device to precisely align with that area. Simultaneously, the system adjusts the frequency setting of the TMS device according to the predicted stimulation frequency. Different stimulation frequencies have significantly different effects on the brain; the correct frequency setting can enhance the treatment effect or alleviate symptoms. For example, high-frequency stimulation is typically used to activate brain function, while low-frequency stimulation is used to inhibit overactive brain regions. Furthermore, the adjustment of stimulation intensity is also based on the predictions, ensuring that the stimulation intensity achieves the therapeutic effect without causing discomfort to the patient. Through meticulous parameter adjustments, the system ensures that each setting best meets the patient's specific needs, thereby improving the safety and effectiveness of the treatment. This process makes transcranial magnetic stimulation therapy more precise and can provide more effective treatment solutions for patients' specific conditions.

[0038] In this implementation plan, determining a treatment plan suitable for the patient's current condition based on predicted treatment parameters significantly improves the personalization and precision of transcranial magnetic stimulation (TMS) therapy. The system utilizes machine learning models to precisely adjust the stimulation target, frequency, and intensity, ensuring that each setting is specifically tailored to the patient's neurophysiological condition and treatment needs. For example, by selecting the optimal stimulation target, the system can more effectively reach the treatment area, thereby improving therapeutic efficacy. Simultaneously, precise frequency adjustments help optimize brain response, while appropriate stimulation intensity ensures both effectiveness and safety. These fine-tuning adjustments enable TMS therapy to more accurately address various neurological diseases, maximizing therapeutic effects while minimizing unnecessary side effects, providing patients with a safer and more effective treatment option.

[0039] Specifically, the calculation method for the predicted values ​​of treatment parameters is as follows: In the formula T p This represents the predicted value of the treatment parameter. This represents a weighted sum of the stimulation targets in the left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex, w. i This represents the weight coefficient for the i-th target, typically ranging from 0 to 1. The specific value is optimized based on historical data and treatment efficacy. α iThe index of the i-th target is adjusted according to the influence of different targets. It can be positive or negative to adjust the degree of influence of different targets. f represents the stimulation frequency, w4 represents the weighting coefficient of the stimulation frequency, which is adjusted according to the contribution of the stimulation frequency to the therapeutic effect, usually selected between 0 and 1. I represents the stimulation intensity, w5 represents the weighting coefficient of the stimulation intensity, which is also set based on its influence on the therapeutic effect. left This indicates the blood oxygen concentration in the left prefrontal cortex, O. right The value represents the blood oxygen concentration in the right prefrontal cortex. w6 represents the weighting coefficient of the product of the blood oxygen concentrations in the left and right prefrontal cortexes, used to assess the synergistic effect of blood oxygen concentration in the two cerebral hemispheres. ΔO left This represents the temporal changes in the left prefrontal cortex, ΔO. right The value represents the temporal changes in the right prefrontal cortex. w7 represents the weighting coefficient of the product of the temporal changes in cerebral blood oxygen concentration in the left and right prefrontal lobes. This coefficient is used to assess the impact of time changes on the treatment effect. The weighting coefficient is set based on historical data and stored in the database, and is dynamically adjusted according to real-time changes.

[0040] In this implementation scheme, optimal treatment parameters are calculated based on real-time data. This includes integrating information on blood oxygen concentration and temporal changes in the left and right prefrontal lobes using weighted and exponential values ​​of target location, as well as weighted values ​​of stimulation frequency and intensity. This method not only optimizes the selection of treatment targets and stimulation intensity but also considers the interactions between brain regions, thereby significantly improving the precision and effectiveness of treatment. This comprehensive calculation method allows for more personalized treatment plans, ensuring that each patient receives treatment tailored to their specific brain physiology and functional condition.

[0041] Specifically, the process of evaluating treatment effectiveness and making corresponding optimization adjustments based on that effectiveness is as follows: First, the system acquires clinical feedback, disease history information, treatment progress data, and brain function effects. This integration of data provides a multi-dimensional perspective for comprehensive evaluation of treatment effectiveness. Clinical feedback may include patient self-reported symptoms, observations by medical professionals, and records of any side effects; disease history information provides the patient's health background and response data to previous treatments; treatment progress data reflects the implementation of the current treatment plan and the patient's response trend; and brain function effects are measured using the latest electroencephalogram (EEG) or functional near-infrared spectroscopy (FIR) data, showing changes in brain activity and the direct impact of treatment. Through the fusion of this information and the application of advanced data analysis techniques, a comprehensive treatment effectiveness evaluation value is calculated. This evaluation value, based on a preset algorithm and weighting parameters, reflects the overall effectiveness of the treatment and any aspects requiring adjustment. Then, based on this treatment effectiveness evaluation value, the system automatically proposes optimization adjustments, such as adjusting the stimulation intensity, frequency, or specific target points of the transcranial magnetic stimulation (TMS) device.

[0042] In this implementation plan, during the evaluation and optimization of treatment effectiveness, the system comprehensively analyzes data from clinical feedback, disease history, treatment progress, and brain function effects to generate a comprehensive treatment effectiveness assessment value. This comprehensive evaluation mechanism allows the system to accurately quantify treatment effectiveness and adjust treatment parameters in real time based on specific assessment results, such as adjusting the target point, frequency, and intensity of transcranial magnetic stimulation. This process significantly improves the adaptability and responsiveness of treatment, enabling treatment plans to flexibly respond to specific patient reactions and changes, ensuring that treatment measures are both effective and safe. Through such dynamic adjustments, the system not only optimizes treatment effects but also enhances the patient's treatment experience, effectively addressing the lack of personalization and real-time feedback in traditional treatment plans.

[0043] Specifically, the process of making corresponding optimization adjustments based on the treatment effect evaluation value is as follows: First, the system compares the treatment effect evaluation value with a set threshold in real time. When the treatment effect evaluation value is greater than the threshold, it indicates that the current treatment plan is effective, so no adjustments are made and the existing treatment plan continues. When the treatment effect evaluation value is less than or equal to the threshold, it indicates that the treatment effect has not met expectations and adjustments are needed. In this case, the system will automatically make several key adjustments to optimize the treatment effect. These include adjusting the target location of transcranial magnetic stimulation (TMS) treatment, which is determined by analyzing brain function data and clinical feedback to identify more effective stimulation areas; adjusting the stimulation frequency to better suit the current frequency of the patient's brain response; and adjusting the stimulation intensity to ensure that the stimulation achieves sufficient therapeutic effect without causing discomfort to the patient. These adjustments are based on real-time data and advanced analysis algorithms to ensure that every adjustment is accurate and evidence-based.

[0044] This implementation scheme ensures the adaptability and maximizes the effectiveness of the treatment process by monitoring and dynamically adjusting treatment parameters in real time. During implementation, the system compares the treatment effect assessment value with preset thresholds in real time, using this as the basis for adjustments. When the assessment value is higher than the threshold, it indicates that the current treatment strategy is effective, and the status quo is maintained. If the assessment value is lower than or equal to the threshold, an adjustment mechanism is triggered, including optimizing the specific location of the target point, adjusting the stimulation frequency, and adjusting the stimulation intensity. These adjustments are based on precise data analysis and aim to rapidly improve the treatment effect to the ideal level. This responsive adjustment strategy not only improves the personalization and precision of treatment but also enhances the safety of the treatment process and patient comfort, effectively addressing the shortcomings of traditional treatment methods, such as the lack of immediate feedback and adjustment capabilities. Through this method, the system ensures that each patient can obtain effective treatment results under appropriate treatment conditions.

[0045] Specifically, the calculation method for the treatment effect evaluation value is as follows: In the formula E t R represents the treatment effectiveness assessment value, an indicator used to comprehensively consider various factors. c This refers to clinical feedback, which includes direct feedback from patients regarding the treatment's effectiveness as well as evaluations from therapists. c The weighting coefficient representing clinical feedback typically ranges from 0.1 to 0.5. This range allows the parameter to have a moderate weight in the overall assessment, ensuring that other variables can also influence the overall assessment result. R f This indicates the patient's medical history, including previous treatment responses and detailed medical history. f This represents the weighting coefficient for disease history information. It is generally set between 0.1 and 0.4, and adjusted based on the extent to which the disease history influences the expected outcome of the current treatment plan. Used to adjust R f The nonlinear effect on the overall assessment ensures that the assessment remains stationary even at extreme or marginal values. e The effects on brain function are typically assessed using data obtained from electroencephalography (EEG) or functional magnetic resonance imaging (fMRI). b The weighting coefficient T represents the effect on brain function. prog Indicates treatment progress data, w T The weighting coefficients representing treatment progress data, i.e., the weighting coefficients for brain function effects, are usually set between 0.2 and 0.5, because they directly reflect the impact of treatment on the patient's brain activity. The weighting coefficients are set based on historical data and stored in the database, and are dynamically adjusted according to real-time changes.

[0046] In this implementation plan, by applying complex calculation formulas to evaluate treatment effectiveness, the system can integrate multiple clinical and neurological function data to generate an accurate treatment effectiveness assessment value. This assessment includes data on clinical feedback, disease history, neurological function effects, and treatment progress. Each data point is appropriately weighted using predefined weighting factors to ensure the comprehensiveness and balance of the assessment. Clinical feedback and disease history information provide direct insights into the patient's response to treatment and past treatment experiences, while neurological function effects and treatment progress data directly reflect the actual impact of treatment on the patient's health. This comprehensive assessment allows the treatment plan to be dynamically adjusted based on the specific effect values ​​obtained in real time, improving the accuracy and personalization of treatment and ensuring that patients receive appropriate treatment outcomes.

[0047] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0048] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation (TMS) therapy, comprising a data acquisition and preprocessing module, a feature extraction module, a treatment parameter prediction module, and a treatment effect evaluation module, characterized in that: The data acquisition and preprocessing module is used to acquire patients’ brain function data in real time, obtain target weights, stimulation frequency and stimulation intensity, acquire clinical data, and preprocess the brain function data and clinical data. The specific process of acquiring real-time brain function data from patients, obtaining target weights, stimulation frequency, and stimulation intensity, and acquiring clinical data is as follows: Brain function data includes electroencephalogram (EEG) data and brain region oxygenation concentration change data. EEG data is collected through an EEG sensor, and brain region oxygenation concentration change data is collected through an fNIRS sensor. The activity of relevant brain regions is detected through EEG and functional near-infrared spectroscopy. The left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex stimulation targets are selected. The appropriate stimulation frequency and intensity are selected based on the patient's EEG analysis results. Clinical data is automatically collected by the system based on user input. The feature extraction module is used to extract features from the preprocessed brain function data and clinical data, and to mark the extracted data as feature data. The treatment parameter prediction module is used to predict transcranial magnetic stimulation treatment parameters based on feature data, and to determine a treatment plan suitable for the patient's current condition based on the transcranial magnetic stimulation treatment parameter prediction results. The specific process of predicting transcranial magnetic stimulation (TMS) parameters based on feature data and determining a suitable treatment plan for the patient's current condition based on the predicted TMS parameters is as follows: The stimulation targets in the left, right, and medial prefrontal lobes, along with their stimulation frequency, intensity, oxygen saturation in the left and right prefrontal lobes, and temporal changes in the left and right prefrontal cortex, are obtained. Predicted treatment parameters are calculated by comprehensively considering these parameters, and a suitable treatment plan is determined based on the patient's current condition. The specific calculation method for the predicted values ​​of the treatment parameters is as follows: ； In the formula This represents the predicted value of the treatment parameter. This represents a weighted sum of the stimulation targets in the left prefrontal cortex, right prefrontal cortex, and medial prefrontal cortex. This represents the weight coefficient of the i-th target point. This represents the index of the i-th target point. Indicates the frequency of stimulation. Weighting coefficients representing stimulus frequency Indicates the intensity of the stimulus. Weighting coefficients representing stimulus intensity This indicates the blood oxygen concentration in the left prefrontal cortex. This indicates the blood oxygen concentration in the right prefrontal cortex. The weighting coefficient represents the product of the cerebral blood oxygen concentrations in the left and right prefrontal lobes. This indicates the phase changes in the left prefrontal cortex. This indicates the phase changes in the right prefrontal cortex. The weighting coefficients represent the product of the phase changes in blood oxygen concentration in the left and right prefrontal lobes of the brain; The treatment effect evaluation module is used to continuously monitor the patient's brain function data during the treatment process, obtain brain function effects, evaluate treatment effects, and make corresponding optimization and adjustment measures based on the treatment effects.

2. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 1, characterized in that: The specific process for preprocessing brain function data and clinical data is as follows: The brain function data and clinical data are denoised and standardized, and then merged to generate the patient's brain function status characteristics.

3. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 1, characterized in that: The specific process for feature extraction from preprocessed brain function data and clinical data is as follows: The alpha wave frequency of the occipital lobe was extracted from the EEG data, and the blood oxygen concentration of the left prefrontal lobe, the blood oxygen concentration of the right prefrontal lobe, the phase changes of the left prefrontal cortex, and the phase changes of the right prefrontal cortex were extracted from the brain region blood oxygen concentration change data. We obtain patients' clinical feedback and disease history information from clinical data, and record treatment progress data from clinical practice.

4. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 1, characterized in that: The specific process of determining a suitable treatment plan based on the predicted values ​​of treatment parameters for the patient's current condition is as follows: Based on the predicted values ​​of treatment parameters, a treatment plan suitable for the patient's current condition is determined. Based on the predicted stimulation target, a suitable stimulation target is selected for the patient. Based on the predicted stimulation frequency, the frequency of the transcranial magnetic stimulation device is set. Based on the predicted stimulation intensity, the intensity of the transcranial magnetic stimulation device is adjusted.

5. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 1, characterized in that: The specific process for evaluating treatment effectiveness and making corresponding optimization and adjustment measures based on the treatment effectiveness is as follows: We obtain clinical feedback, disease history information, treatment progress data, and brain function effects. By comprehensively calculating the treatment effect evaluation value based on the clinical feedback, disease history information, treatment progress data, and brain function effects, we can make corresponding optimization and adjustment measures based on the treatment effect evaluation value.

6. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 5, characterized in that: The specific process for making corresponding optimization and adjustment measures based on the treatment effect evaluation value is as follows: The treatment effect assessment value is compared with the threshold in real time. When the treatment effect assessment value is greater than the threshold, no adjustment is made. When the treatment effect assessment value is less than or equal to the threshold, adjust the specific location of the target, the stimulation frequency, and the stimulation intensity until the treatment effect assessment value is raised above the threshold.

7. The artificial intelligence-based individualized parameter optimization system for transcranial magnetic stimulation therapy according to claim 5, characterized in that: The specific calculation method for the treatment effect evaluation value is as follows: ； In the formula This indicates the evaluation value of treatment effectiveness. This indicates clinical feedback. The weighting coefficient representing clinical feedback. Indicates disease history information, The weighting coefficients represent the historical information of the disease. Indicates the effect on brain function. Weighting coefficients representing the effects on brain function. Indicates treatment progress data, Weighting coefficients representing treatment progress data.

Images