A near-infrared phototherapy regulation method suitable for cognitive intervention of neurodegenerative diseases and affective disorders

Abstract

The application discloses a transcranial near-infrared phototherapy regulation method suitable for cognitive intervention of neurodegenerative diseases and emotional disorders, and belongs to the technical field of neural regulation and rehabilitation medical treatment. The method is based on a transcranial near-infrared phototherapy regulation system, penetrates the scalp and skull with 800nm-1100nm near-infrared light to intervene in the brain function network through a non-thermal photochemical effect; cognitive function evaluation data and electroencephalogram, heart rate variability and blood oxygen saturation physiological parameters of a subject are collected, an individualized phototherapy parameter is output through a regulation strategy model constructed by machine learning, an array light source is driven to irradiate a target point of a target brain area, ATP of brain cells is improved, and metabolic waste in the brain is promoted to be removed. Physiological feedback is collected in real time during the irradiation process, parameters are dynamically adjusted or the irradiation is terminated according to a safety threshold in a closed loop, and a course management and remote updating function is matched. The application realizes individualized and closed-loop non-invasive phototherapy intervention, is high in safety, and is used for improving cognitive impairment caused by diseases such as Alzheimer's disease, Parkinson's disease and depression.

Description

Technical Field

[0001] This invention relates to the field of neuromodulation and rehabilitation medicine technology, specifically to a transcranial near-infrared phototherapy modulation method suitable for cognitive intervention in neurodegenerative diseases and emotional disorders. Background Technology

[0002] Neurodegenerative diseases and mood disorders are two major types of brain diseases that seriously threaten human health. Among them, neurodegenerative diseases are primarily Alzheimer's disease. Alzheimer's Disease AD), Parkinson's disease ( Parkinson's Disease PD (Prostatic Disorder) is a typical example, and its core pathological features include progressive neuronal loss, abnormal protein accumulation in the brain (such as β-amyloid plaques, τ-protein neurofibrillary tangles, and α-synuclein Lewy bodies), and brain energy metabolism disorders; mood disorders are mainly characterized by depression (…). Major Depressive Disorder MDD (prefrontal cortex disorder) is a typical example, and its core pathological features include abnormal prefrontal-limbic circuitry, decreased neuroplasticity, reduced expression of brain-derived neurotrophic factor (BDNF), and disordered brain network connectivity. All of these diseases are accompanied by varying degrees of cognitive impairment, including decreased attention, reduced working memory, and executive dysfunction, severely affecting patients' quality of life and social function.

[0003] Currently, clinical interventions for these diseases mainly fall into two categories: drug therapy and physical neuromodulation. In terms of drug therapy, cholinesterase inhibitors (such as donepezil) and NMDA receptor antagonists (such as memantine) are commonly used for Alzheimer's disease, levodopa preparations and dopamine receptor agonists are commonly used for Parkinson's disease, and selective serotonin reuptake inhibitors (SSRIs) are commonly used for depression. However, drug therapy generally has limitations such as slow onset of action, large individual variability, significant side effects with long-term use, and insufficient response rates in some patients. For example, approximately 30% of patients with depression do not respond to first-line antidepressants (i.e., treatment-resistant depression); existing drugs for Alzheimer's disease can only delay symptom progression, not reverse the pathological process; and long-term use of levodopa in Parkinson's disease patients is prone to motor complications (such as end-of-dose phenomenon and dyskinesia).

[0004] In the field of physical neuromodulation, transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) have been applied clinically to some extent, but each has its limitations: TMS devices are bulky and expensive, and their stimulation depth to deep brain regions is limited; tDCS has low current intensity, insufficient spatial resolution, and significant individual differences in efficacy. Therefore, there is an urgent clinical need for a novel neuromodulation technique that is non-invasive, highly safe, can penetrate deep brain regions, and can simultaneously intervene in energy metabolism and protein clearance pathways.

[0005] Transcranial near-infrared phototherapy ( transcranial Photobiomodulation Near-infrared phototherapy (tPBM) is a non-invasive neuromodulation technique that has emerged in recent years. Its basic principle is to utilize near-infrared light of a specific wavelength (mainly within the "optical therapy window" of 800nm–1100nm) to penetrate the scalp and skull, reaching the cerebral cortex and even deep brain regions. The photons are absorbed by cytochrome C oxidase (CcO) in the mitochondria of neurons, triggering a series of photochemical reactions: including increasing mitochondrial membrane potential, promoting ATP synthesis, increasing cerebral blood flow and oxygen supply, regulating reactive oxygen species (ROS) signaling pathways, activating transcription factors (such as NF-κB and Nrf2), promoting BDNF expression, inhibiting neuroinflammatory responses, and enhancing the efficiency of the glymphatic system in clearing metabolic waste products (such as β-amyloid protein). Unlike traditional phototherapy, tPBM is a non-thermal effect-driven neuromodulation technique; its therapeutic effect does not depend on tissue temperature rise but rather on the precise intervention of photochemical signals on brain functional networks, thus exhibiting higher safety and tolerability.

[0006] However, existing transcranial near-infrared phototherapy technology still has the following shortcomings: First, most devices use fixed parameter outputs, which cannot be individually adjusted according to the patient's cognitive function level and physiological state, resulting in inconsistent efficacy; Second, the lack of real-time physiological feedback and safety closed-loop control makes it difficult to avoid potential safety risks while ensuring efficacy; Third, the spatial distribution of the light source array and the corresponding accuracy of the brain region target points are insufficient, affecting the effective deposition of light energy in the target brain region; Fourth, existing technologies mostly focus on single indications and lack differentiated control strategies for different pathological characteristics of neurodegenerative diseases and emotional disorders; Fifth, most devices lack treatment course management and remote follow-up functions, which is not conducive to long-term rehabilitation management.

[0007] While some existing technologies propose using machine learning to optimize neuromodulation parameters, most focus on electrical stimulation devices. In the field of transcranial near-infrared phototherapy, integrated solutions that fully combine cognitive assessment, multi-source physiological input, multi-dimensional phototherapy parameter output, real-time multi-indicator safety closed loop, and treatment iteration are still relatively rare. Most solutions only focus on the phototherapy hardware itself and lack a complete technical chain from baseline assessment, individualized parameter generation, safety protection, to multi-treatment follow-up iteration.

[0008] Therefore, developing a personalized, closed-loop, safe and controllable transcranial near-infrared phototherapy modulation method and system based on the patient's cognitive function and physiological state is of great clinical significance and application value for improving the cognitive intervention effect of neurodegenerative diseases and emotional disorders. Summary of the Invention

[0009] The purpose of this invention is to overcome the shortcomings of the prior art and provide a transcranial near-infrared phototherapy modulation method and system suitable for cognitive intervention in neurodegenerative diseases and emotional disorders, so as to achieve individualized, closed-loop, safe and controllable transcranial near-infrared phototherapy modulation, improve the effect of cognitive intervention and promote the clearance of metabolic waste in the brain.

[0010] Most existing transcranial near-infrared phototherapy protocols use preset, fixed phototherapy parameters, relying solely on scalp temperature for simple overheat protection. They fail to incorporate pre-planned, individualized parameters based on the subject's cognitive impairment, EEG, and autonomic nervous system physiological state, nor do they differentiate target points and phototherapy paradigms according to the pathological characteristics of different diseases. Furthermore, they ignore the attenuation differences of near-infrared light between the scalp and skull tissues, limiting only the irradiance of the scalp surface, making it difficult to ensure effective light doses to deep brain regions. Additionally, they lack a complete rehabilitation pathway that combines single-intervention safety protection with multi-treatment iterative optimization. This invention addresses these shortcomings by combining cognitive assessment, multi-source physiological signal acquisition, machine learning parameter planning, multi-dimensional safety closed-loop protection, target light dose compensation, and dynamic iterative treatment, while retaining a channel for manual intervention by clinical personnel, balancing automated equipment output with the needs of complex clinical cases.

[0011] To achieve the above objectives, the present invention adopts the following technical solution: A transcranial near-infrared phototherapy modulation method applicable to cognitive intervention for neurodegenerative diseases and affective disorders. The method is based on a transcranial near-infrared phototherapy modulation system, utilizing near-infrared light within an optical treatment window of 800nm–1100nm to penetrate the subject's scalp and skull to reach the cerebral cortex and deep brain regions, intervening in brain functional networks with photochemical signals dominated by non-thermal effects. The method includes the following steps: S1. Obtain cognitive function assessment data and physiological state parameters of the subject. The cognitive function assessment data includes at least one score of attention, memory, and executive function. The physiological state parameters include at least one of electroencephalogram (EEG) signal, heart rate variability, and blood oxygen saturation. S2. Based on the cognitive function assessment data and physiological state parameters, transcranial near-infrared phototherapy parameters are generated through a preset regulation strategy model. The transcranial near-infrared phototherapy parameters include light wavelength, irradiance, irradiation duration, irradiation mode, and irradiation target area. S3. Drive the near-infrared light source module in the transcranial near-infrared phototherapy control system according to the parameters of the transcranial near-infrared phototherapy, and irradiate the preset brain region target point on the subject's head with transcranial near-infrared light, so that the photons are absorbed by the mitochondria of neurons to enhance the ATP energy of brain cells and promote the removal of metabolic waste. S4. During the irradiation process, the physiological feedback signal of the subject is collected in real time and compared with the preset safety threshold. When the physiological feedback signal exceeds the preset safety threshold, the transcranial near-infrared phototherapy parameters are dynamically adjusted or the irradiation is terminated.

[0012] Furthermore, the transcranial near-infrared phototherapy control system includes: a main control module, a near-infrared light source module, a physiological signal acquisition module, a human-computer interaction module, and a power supply module; the main control module is electrically connected to the near-infrared light source module, the physiological signal acquisition module, and the human-computer interaction module, respectively, and is used to execute the control strategy model and output control commands; the near-infrared light source module includes multiple near-infrared emitting units, which are arranged in an array and correspond to different brain regions of the subject's head target points, and the emission wavelength of the near-infrared emitting units is 808nm~1064nm; the physiological signal acquisition module is used to acquire the subject's electroencephalogram (EEG) signals, heart rate variability, and blood oxygen saturation; the human-computer interaction module is used to input cognitive function assessment data and display phototherapy parameters and physiological feedback signals; the power supply module is used to supply power to each module of the system.

[0013] Furthermore, the irradiance range of the near-infrared emitting unit is 1 mW / cm². 2 ~100mW / cm 2 The output power of each near-infrared emitting unit is independently adjustable; the near-infrared light acts on brain tissue in a non-thermal manner, and the temperature of the subject's scalp surface does not rise by more than 2°C during irradiation.

[0014] Furthermore, the near-infrared light source module also includes a heat dissipation unit and a light intensity detection unit. The light intensity detection unit collects the actual output light intensity of the near-infrared light-emitting unit in real time and feeds it back to the main control module. The main control module performs closed-loop correction based on the deviation between the actual output light intensity and the target light intensity. Main control module optical intensity closed-loop correction: ; ; E target Target irradiance, E real The actual irradiance collected by the light intensity detection unit; K p ,K i ,K d These are PID control parameters; u k This is the output control quantity.

[0015] Furthermore, the method for constructing the preset regulation strategy model in step S2 includes: collecting cognitive function assessment data, physiological state parameters, and corresponding transcranial near-infrared phototherapy response data of patients with neurodegenerative diseases and emotional disorders to construct a training dataset; using the degree of improvement in cognitive function and the efficiency of clearing metabolic waste in the brain as labels, training the training dataset with a machine learning algorithm to obtain a mapping relationship model; using the mapping relationship model as the regulation strategy model to output the optimal transcranial near-infrared phototherapy parameters based on the input cognitive function assessment data and physiological state parameters; The model input feature vector X : ; In the formula: X age For age; X gender Gender coding; X cog Assess cognitive function; X EEG Characteristics of the brainwave frequency band; X HRV This is an indicator of heart rate variability. X SpO2 Blood oxygen saturation; X disease Code the disease type.

[0016] Model outputs phototherapy parameter vector F : ; In the formula: For the completed mapping model, These are the model weight parameters; Near-infrared wavelength; E I represents irradiance; dur This refers to the duration of irradiation. f pulse The pulse frequency; D cycle Duty cycle; R roi is the target region encoding.

[0017] Model training optimizes the objective loss function: ; Predicting loss based on the degree of improvement in cognitive function; Predicting losses in the efficiency of clearing metabolic waste from the brain; , These are the weighting coefficients for the two objectives. This is an L2 regularization term, used to suppress model overfitting.

[0018] Furthermore, the irradiation mode in step S3 includes a continuous wave mode, a pulse modulation mode, and a combination of the two; the pulse frequency range of the pulse modulation mode is 1Hz to 100Hz, and the duty cycle range is 10% to 90%.

[0019] Furthermore, the irradiation target area in step S3 includes at least one of the prefrontal cortex, the scalp projection area corresponding to the hippocampus, and the temporoparietal junction; the near-infrared light source module is fixed to the subject's head by a wearable headband or helmet-like structure, so that the positional deviation between each near-infrared emitting unit and the corresponding irradiation target area does not exceed 5mm.

[0020] Furthermore, the preset safety thresholds in step S4 include: scalp surface temperature not exceeding 40°C, blood oxygen saturation fluctuation not exceeding ±10% of the baseline value, and abnormal spike wave occurrence rate in EEG signals not exceeding a preset frequency; when any one of them exceeds the corresponding threshold, the main control module reduces the irradiance or shortens the irradiation time; when two or more of them exceed the threshold simultaneously, the irradiation is immediately terminated and an alarm is issued. The security discrimination logic satisfies: ; S represents the count of out-of-limit indicators; This is an indicator function; it takes the value 1 if the condition is true and 0 if it is false. When S=1, the parameters are dynamically adjusted. When S≥2, an alarm to terminate irradiation is triggered.

[0021] Furthermore, the method also includes a treatment management step: setting a treatment cycle according to the subject's disease type and degree of cognitive impairment, with each treatment cycle containing multiple phototherapy interventions, and each intervention interval being no less than 24 hours; after each treatment cycle, re-acquiring cognitive function assessment data, and updating the transcranial near-infrared phototherapy parameters for the next treatment cycle based on the assessment results.

[0022] Furthermore, the transcranial near-infrared phototherapy control system also includes a wireless communication module. The main control module is connected to a remote terminal through the wireless communication module to upload cognitive function assessment data, physiological state parameters and phototherapy records to the remote terminal, and to receive the control strategy model update parameters sent by the remote terminal.

[0023] It should be noted that the phototherapy parameters output by the control strategy model of this invention are reference intervention plans provided by the system, not fixed instructions that cannot be changed. Medical staff can manually modify the light wavelength, irradiance, irradiation duration, and target area within the limits allowed by the system hardware parameters, based on the individual clinical condition of the subject. The modified parameters will also be recorded and saved by the system for subsequent treatment evaluation and model iteration optimization. For subjects with severe brain atrophy and individual differences in scalp thickness, the system can differentiate the output irradiance of each luminescent unit based on the depth of the target brain region. The irradiance of the target points corresponding to deep brain regions is appropriately increased, while the output of the target points in the superficial cortex is moderately reduced to offset the energy attenuation of near-infrared light during its propagation through the scalp and skull, ensuring that the target brain region receives an effective therapeutic light dose while avoiding local light energy overload on the scalp. As an auxiliary rehabilitation method, the intervention method of this invention should be assessed by a clinician before implementation to determine the suitability of the subject and exclude situations that are unsuitable for near-infrared phototherapy, such as photosensitive diseases, scalp damage, and intracranial metal implants. The intervention process respects the subject's subjective feelings, and even if the physiological indicators do not reach the safety threshold, manual pause or termination of phototherapy is supported.

[0024] It should be noted that the regulatory strategy model of this invention operates in the pre-intervention parameter generation stage, completing the planning and output of a complete set of phototherapy parameters based on the subject's baseline cognitive and physiological data. During the actual phototherapy irradiation process, machine learning inference calculations are no longer performed; instead, regularized physiological feedback closed-loop control is executed based on preset safety thresholds. This achieves a two-tiered division of labor between pre-intervention intelligent parameter planning and in-session safety protection. It utilizes data-driven approaches to obtain individualized intervention plans while reducing the real-time computational overhead of the main control module, ensuring the operational stability of the intervention process. This invention uses both the degree of improvement in cognitive function and the efficiency of brain metabolic waste clearance as dual indicators as model training labels, taking into account both subjective cognitive scale outcomes and objective pathological metabolic improvement goals, avoiding parameter bias caused by a single label.

[0025] The beneficial effects of this invention are as follows: (1) The present invention utilizes near-infrared light within the 800nm-1100nm optical treatment window, which can effectively penetrate the scalp and skull with low energy loss, reach the cerebral cortex and deep brain regions, and achieve non-invasive intervention on deep brain functional networks. (2) The present invention uses photochemical signals dominated by non-thermal effects as the intervention mechanism. Photons are absorbed by cytochrome C oxidase in the mitochondria of neurons, which increases the ATP energy of brain cells and promotes the clearance of metabolic waste (such as β amyloid protein) in the brain, while avoiding the risk of thermal damage. (3) The present invention generates phototherapy parameters individually based on the cognitive function assessment data and physiological state parameters of the subjects through a preset control strategy model, thereby achieving precise and personalized intervention and overcoming the defects of inconsistent efficacy of fixed parameter equipment. (4) The present invention adopts real-time physiological feedback and safety threshold closed-loop control, which dynamically adjusts parameters or terminates irradiation while ensuring efficacy, significantly improving the safety and tolerability of treatment. (5) The near-infrared light source module of the present invention adopts an array distribution and each light-emitting unit is independently adjustable. Combined with the wearable headband / helmet structure, it can achieve precise target positioning and improve the effective deposition of light energy in the target brain region. (6) The present invention provides treatment management and remote communication functions, supports long-term rehabilitation management and remote follow-up, which is beneficial to the patient's continuous rehabilitation and consolidation of therapeutic effects.

[0026] (7) This invention takes into account both large-scale clinical department use and home rehabilitation. At the hardware level, it supports wearable structure adaptation and has the ability to manually correct parameters, retain intervention records, screen contraindications, and adjust parameters in a human-machine collaborative manner. This improves the adaptability of the solution to complex individual differences and expands the practical feasibility of transcranial near-infrared phototherapy in the context of neurodegenerative diseases and cognitive rehabilitation of emotional disorders.

[0027] (8) This invention distinguishes between a two-level mechanism: pre-machine learning parameter planning and threshold-based safety protection during the irradiation stage. The model only completes parameter output before intervention. During the irradiation stage, a lightweight threshold discrimination logic is used to reduce the computing power burden of embedded hardware and adapt to the deployment conditions of wearable embedded main control modules. At the same time, light dose compensation is performed for different brain regions to offset the light energy attenuation caused by the scalp and skull, ensuring that deep target brain tissues receive effective photochemical stimulation and avoiding the problem of individual deviation in intracranial actual dose caused by using scalp surface irradiance as the only indicator. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the overall process of the transcranial near-infrared phototherapy control method of the present invention; Figure 2 This is a structural block diagram of the transcranial near-infrared phototherapy control system of the present invention; Figure 3 This is a schematic diagram of the array distribution of the near-infrared light source module of the present invention; Figure 4 This is a schematic diagram illustrating the construction process of the control strategy model of the present invention; Figure 5 This is a schematic diagram of the safety closed-loop control process of the present invention; Detailed Implementation

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0030] I. Overall System Structure

[0031] like Figure 2 As shown, the transcranial near-infrared phototherapy control system of the present invention includes a main control module 1, a near-infrared light source module 2, a physiological signal acquisition module 3, a human-computer interaction module 4, a power supply module 5, and an optional wireless communication module 6.

[0032] The main control module 1 uses an embedded microcontroller (such as an ARM Cortex-M4 or a higher performance processor), with built-in regulation strategy model algorithms and safety control logic, and is the core control unit of the entire system. The main control module 1 is electrically connected to the near-infrared light source module 2, the physiological signal acquisition module 3, and the human-computer interaction module 4, respectively. It is used to receive cognitive function assessment data and physiological state parameters, execute the regulation strategy model to generate phototherapy parameters, output PWM control commands to drive the near-infrared light source module 2, and perform safety closed-loop control based on real-time physiological feedback signals.

[0033] The near-infrared light source module 2 includes multiple near-infrared light-emitting units 21, a heat dissipation unit 22, and a light intensity detection unit 23. For example... Figure 3 As shown, multiple near-infrared emitting units 21 are arrayed on the inner side of the wearable headband or helmet-like structure, their spatial positions corresponding to different brain regions of the subject's head (such as the prefrontal cortex, the scalp projection area corresponding to the hippocampus, the temporoparietal junction, etc.). The near-infrared emitting units 21 employ laser diodes or high-power LEDs with wavelengths ranging from 808nm to 1064nm, preferably one or a combination of 810nm, 850nm, 940nm, and 1064nm. The output power of each near-infrared emitting unit 21 is independently adjustable, with an irradiance range of 1mW / cm² to 100mW / cm², which can be flexibly configured according to the target area and treatment needs. The heat dissipation unit 22 uses a metal heat sink combined with air cooling or phase-change cooling to ensure stable temperature of the light source module during prolonged irradiation. The light intensity detection unit 23 uses a photodiode to collect the actual output light intensity of each near-infrared emitting unit 21 in real time and feed it back to the main control module 1. The main control module 1 performs PID closed-loop correction based on the deviation between the actual output light intensity and the target light intensity to ensure the stability and accuracy of the light intensity output.

[0034] The physiological signal acquisition module 3 includes an EEG acquisition submodule 31, a heart rate variability acquisition submodule 32, a blood oxygen saturation acquisition submodule 33, and a scalp temperature acquisition submodule 34. The EEG acquisition submodule 31 uses multi-channel dry or wet electrodes with a sampling rate of no less than 250Hz to acquire EEG signals from brain regions such as the frontal and parietal lobes of the subject, extracting power characteristics in frequency bands such as δ, θ, α, β, and γ. The heart rate variability acquisition submodule 32 uses photoplethysmography (PPG) or electrocardiogram electrodes to acquire heart rate and heart rate variability indices (such as SDNN, RMSSD, and LF / HF ratio). The blood oxygen saturation acquisition submodule 33 uses a clip-on or ear-worn pulse oximeter to monitor peripheral blood oxygen saturation (SpO2) in real time. The scalp temperature acquisition submodule 34 uses an infrared temperature sensor or an NTC thermistor, positioned near each irradiation target point, to monitor the scalp surface temperature in real time.

[0035] The human-computer interaction module 4 includes a touch screen display 41 and a voice prompt unit 42. The touch screen display 41 is used to display cognitive function assessment scales, input basic information of subjects, display real-time phototherapy parameters and physiological feedback signals, and provide an operation interface and alarm prompts; the voice prompt unit 42 is used to issue voice prompts when treatment begins, parameters are adjusted, treatment ends, and alarms occur, thereby improving the user experience.

[0036] Power module 5 uses a rechargeable lithium battery pack (such as a combination of 18650 cells with a capacity of not less than 5000mAh), and works with a power management chip to achieve charge and discharge management and overvoltage, overcurrent and overheat protection, providing a stable DC power supply for each module of the system.

[0037] The wireless communication module 6 uses a Wi-Fi (IEEE 802.11b / g / n) or Bluetooth (BLE 5.0) module. The main control module 1 connects to a remote terminal (such as a doctor's workstation, cloud server, or mobile APP) through the wireless communication module 6 to upload cognitive function assessment data, physiological state parameters, and phototherapy records, and to receive the updated parameters of the control strategy model and treatment plan sent by the remote terminal.

[0038] The hardware of this system fully considers the differentiated needs of clinical departments and home rehabilitation scenarios at the engineering implementation level. For hospital clinical scenarios, the wearable headband / helmet can be fixed to the headrest of the treatment chair, further reducing the load on the subject's head and minimizing target point displacement caused by shaking. For home rehabilitation scenarios, the device simplifies external wiring and relies on a built-in lithium battery to complete a full course of phototherapy intervention. The device is designed with hardware fault self-checking logic. During startup, it automatically performs self-checks on the light source module, physiological signal acquisition pathway, and communication link. If a light source unit failure or sensor signal abnormality is detected, the human-computer interaction module directly outputs a fault prompt and prohibits phototherapy irradiation, avoiding phototherapy intervention without feedback in the event of sensor failure. The wearable headband or helmet has a flexible cushioning pad inside, ensuring a close fit between the light-emitting unit and the scalp while reducing discomfort from prolonged wear. The pad is removable and washable, meeting the hygiene and disinfection requirements for multiple subjects using the device alternately.

[0039] This system supports synchronous time-aligned processing of multi-source signals for physiological signal acquisition. The sampling timestamps for EEG, heart rate variability, blood oxygen saturation, and scalp temperature are kept consistent, preventing misalignment of acquisition times from different sensors from affecting the effectiveness of the input data for the regulatory strategy model. For subjects with dense scalp hair, the wearable headband can be equipped with an adjustable pressure block structure, allowing the light-emitting unit to penetrate the hair and adhere closely to the scalp surface, reducing the scattering loss of near-infrared light energy and ensuring the actual light dose incident on the brain tissue. The system's human-computer interaction module also includes built-in operation guidance prompts, providing both text and voice prompts for device wearing procedures, baseline acquisition precautions, and intervention process precautions, lowering the operational threshold for clinical operators and home users.

[0040] II. Overall Process of Control Methods

[0041] like Figure 1 As shown, the transcranial near-infrared phototherapy modulation method of the present invention includes the following steps: Step S1: Data Acquisition.

[0042] First, the subject's basic information (age, gender, disease type, disease course, medication, etc.) is input through the human-computer interaction module 4, and they complete cognitive function assessment scales (such as the Mini-Mental State Examination (MMSE), the Montreal Cognitive Assessment (MoCA), the cognitive section of the ADAS-Cog Alzheimer's Disease Assessment Scale, the motor section of the UPDRS-III Unified Rating Scale for Parkinson's Disease, the Hamilton Depression Rating Scale (HAMD), the Beck Self-Rating Depression Scale (BDI), the Stroop Color-Word Test, the Digit Span Test, and the N-back Working Memory Task, etc.) to obtain cognitive function assessment data. Simultaneously, the subject's baseline physiological parameters, including resting-state EEG signals, heart rate variability, blood oxygen saturation, and scalp temperature, are collected through the physiological signal acquisition module 3.

[0043] It should be noted that during the collection of baseline physiological parameters, subjects need to remain seated quietly to minimize interference from limb movements and emotional fluctuations on physiological signals such as EEG and heart rate variability. If the physiological signal noise is too high during the collection process, the system will prompt for re-collection of baseline data to ensure the data quality input into the regulatory strategy model. For cognitive function assessment, medical staff can complete the scale scoring, or the human-computer interaction module can guide subjects to complete self-assessment through the built-in scale interface. For subjects with severe cognitive impairment who cannot complete the answers independently, caregivers can assist in the information collection.

[0044] During the baseline data collection phase, the system performs a simple validity check on the input data. When there are obvious logical contradictions in the cognitive scale scores or the physiological signal-to-noise ratio is lower than the preset threshold, the system will provide a prompt, guiding the operator to review the entered information or re-collect physiological signals. This prevents abnormal data from being directly input into the control strategy model and avoids outputting unreasonable phototherapy intervention parameters. After baseline collection is completed, the system automatically saves the complete baseline file, using the baseline physiological indicators as a reference benchmark for subsequent S4 safety closed-loop judgments. Subsequent treatment evaluations can also directly retrieve this baseline file for before-and-after comparison.

[0045] Step S2: Parameter generation.

[0046] The main control module 1 inputs the cognitive function assessment data and physiological state parameters obtained in step S1 into a preset regulation strategy model. The model outputs individualized transcranial near-infrared phototherapy parameters, including light wavelength (one or a combination of 808 nm / 850 nm / 940 nm / 1064 nm) and irradiance at each target point (1 mW / cm²~100 mW / cm²). 2 Irradiation duration (5 min to 30 min), irradiation mode (continuous wave / pulse modulation / combined mode, pulse frequency 1 Hz to 100 Hz, duty cycle 10% to 90%), and irradiation target area (prefrontal cortex, hippocampal corresponding scalp projection area, temporoparietal junction, etc.).

[0047] Step S3: Phototherapy.

[0048] After the subject wears a wearable headband or helmet-like structure and confirms that the positional deviation between each near-infrared emitting unit 21 and the corresponding irradiation target area does not exceed 5 mm, the main control module 1 outputs PWM control commands according to the generated phototherapy parameters, driving the near-infrared light source module 2 to irradiate the preset brain region target points on the subject's head with transcranial near-infrared light. Near-infrared light in the 800 nm–1100 nm range penetrates the scalp and skull to reach the cerebral cortex and deep brain regions. Photons are absorbed by cytochrome C oxidase in the mitochondria of neurons, triggering a photochemical reaction that increases ATP energy in brain cells, increases cerebral blood flow and oxygen supply, promotes the clearance of metabolic waste products in the brain (such as β-amyloid protein, τ protein, α-synuclein, etc.), and regulates brain functional network connections. The entire irradiation process is dominated by non-thermal effects, with the scalp surface temperature rising by no more than 2°C.

[0049] During phototherapy, near-infrared light experiences energy attenuation as it travels from the body surface to the deep brain. The array light source automatically distributes output energy according to the depth of the target area, allocating higher irradiance to deeper scalp projection targets to compensate for light attenuation, while appropriately reducing output to superficial cortical targets. This ensures that target brain regions at different depths receive an effective light dose while avoiding localized scalp overload. Throughout the intervention, the subject remains seated with eyes closed and relaxed. Low-volume white noise can be used to help maintain a quiet state, which is beneficial for improving the actual effectiveness of the phototherapy intervention.

[0050] Step S4: Safety closed-loop control.

[0051] like Figure 5 As shown, during the irradiation process, the physiological signal acquisition module 3 collects the subject's physiological feedback signals (EEG, heart rate variability, blood oxygen saturation, and scalp temperature) in real time. The main control module 1 compares each physiological feedback signal with preset safety thresholds: scalp surface temperature does not exceed 40℃, blood oxygen saturation fluctuation does not exceed ±10% of the baseline value, and the incidence of abnormal spikes in the EEG signal does not exceed a preset frequency (e.g., no more than 3 times per minute). When any item exceeds the corresponding threshold, the main control module 1 reduces the irradiance (e.g., reduces it to 50% of the original irradiance) or shortens the irradiation time; when two or more items exceed the threshold simultaneously, the main control module 1 immediately terminates the irradiation and issues an audible and visual alarm through the human-computer interaction module 4, while recording the abnormal event for subsequent analysis. Define the set of Boolean conditions for security judgment: ; T skin Real-time scalp temperature; real-time SpO2 blood oxygenation; SpO2 and blood oxygen levels; N spike Counting of abnormal spike waves on electroencephalogram (EEG); N th Spike frequency threshold.

[0052] like If a single indicator exceeds its limit, the execution parameter should be lowered. like If two or more exceed the limit, the irradiation will be terminated and an alarm will be triggered. If all conditions Establishment: Maintain current phototherapy parameters and continue irradiation.

[0053] Furthermore, this invention also includes a treatment management step: The treatment cycle is set according to the subject's disease type and degree of cognitive impairment (e.g., 4 weeks, 8 weeks, or 12 weeks per cycle), with each cycle containing multiple phototherapy interventions (e.g., 3 to 5 times per week), with an interval of no less than 24 hours between each intervention. After each cycle, cognitive function assessment data and physiological state parameters are reacquired, and the main control module 1 updates the transcranial near-infrared phototherapy parameters for the next cycle based on the assessment results (e.g., adjusting irradiance, irradiation duration, or target combination), achieving dynamically optimized long-term rehabilitation management.

[0054] III. Construction of Regulation Strategy Model

[0055] like Figure 4 As shown, the preset control strategy model is constructed using the following method: (1) Data collection: Patients with neurodegenerative diseases (Alzheimer's disease, Parkinson's disease, etc.) and affective disorders (depression, etc.) were recruited as research subjects. Their baseline cognitive function assessment data (MMSE, MoCA, ADAS-Cog, UPDRS-III, HAMD scores, etc.) and physiological state parameters (power of each frequency band of EEG, heart rate variability index, blood oxygen saturation, etc.) were collected. After receiving transcranial near-infrared phototherapy intervention with different parameter combinations, the degree of improvement in cognitive function (the difference in scores before and after intervention) and the efficiency of brain metabolic waste clearance (which can be indirectly assessed by cerebrospinal fluid Aβ42 / Aβ40 ratio, PET amyloid imaging, serum neurofilament light chain NfL, etc.) were collected as response data.

[0056] (2) Dataset construction: The above data are organized into a training dataset, where the input features include disease type, age, gender, baseline cognitive score, EEG features, heart rate variability features, blood oxygen saturation, etc.; the output labels include optimal light wavelength, irradiance, irradiation duration, irradiation mode parameters, irradiation target combination, and the corresponding degree of improvement in cognitive function and metabolic waste clearance efficiency.

[0057] (3) Model training: The training dataset is trained using machine learning algorithms (such as random forest, gradient boosting tree XGBoost, support vector regression SVR or artificial neural network). The weighted sum of the degree of improvement in cognitive function and the efficiency of clearing metabolic waste in the brain is used as the optimization objective. The mapping relationship model between input features and optimal phototherapy parameters is obtained through cross-validation and hyperparameter tuning.

[0058] Optimize the mathematical expression of the objective: ; Constraints: ; In the formula: Y cog True cognitive improvement label The model predicts the improvement in cognition. Y waste True metabolic waste removal efficiency label. The model predicts the clearance efficiency; the constraint set is used to limit the physical boundaries of the phototherapy parameters to ensure that the output parameters fall within the hardware-achievable range.

[0059] (4) Model Deployment and Update: The trained mapping model is deployed to the main control module 1 as a control strategy model, which is used to output the optimal transcranial near-infrared phototherapy parameters based on the real-time input cognitive function assessment data and physiological state parameters. The system can receive new training data and model update parameters sent by the remote terminal through the wireless communication module 6 to realize the continuous optimization and iteration of the model.

[0060] This invention selects the degree of improvement in cognitive function and the efficiency of clearing metabolic waste in the brain as dual training labels. On the one hand, these labels correspond to the clinically observable cognitive scale behavioral outcomes, and on the other hand, they correspond to the objective pathological improvement goals of tPBM phototherapy on brain tissue. The model trained under dual label constraints outputs parameters that pursue both cognitive and behavioral improvement and the clearing of metabolic waste in the brain, avoiding parameter bias caused by relying solely on a single cognitive outcome label and improving the comprehensive adaptability of the generated phototherapy plan.

[0061] Once deployed, the phototherapy parameters output by the control strategy model will be limited by the system to the range physically feasible by the hardware, and will not output parameter combinations exceeding the capabilities of the light source module hardware. The model output parameters are the optimal reference scheme. Medical staff can manually fine-tune the output parameters through the system's human-computer interface based on the actual clinical condition of the subject. For example, for subjects with sensitive skin, the local target irradiance can be manually reduced; for patients with severe brain atrophy, the irradiation duration and target area can be appropriately adjusted. The parameter schemes output by the model and the records of manual modifications will be completely saved in the phototherapy log for easy review of subsequent treatments and iterative training of the model. It is important to note that the model is only used to provide intervention parameter suggestions and cannot replace the diagnosis of clinicians. Before any intervention is carried out, a clinician must evaluate and confirm that the subject is suitable for this transcranial near-infrared phototherapy intervention.

[0062] It should be noted that all phototherapy parameters output by the control strategy model fall within the inherent parameter range of the system hardware, and will not output intervention plans that exceed the hardware limits. The model focuses on providing individualized reference plans based on population statistical patterns, but due to limitations in sample size and individual pathological conditions, it cannot completely cover all special cases. Therefore, a channel for manual intervention by medical staff is retained. Each manually modified parameter plan and the reason for the modification will be stored together with the original model output plan in the phototherapy file. This historical file can be used by physicians for review and analysis, and, while ensuring privacy protection, can also serve as a reference data source for subsequent model iterations.

[0063] IV. Example 1: Cognitive Intervention for Alzheimer's Disease

[0064] Basic information of the subjects: The patient is a 72-year-old retired male teacher who presented with a chief complaint of "progressive memory decline for over two years, worsening with personality changes for the past six months." He has a 10-year history of hypertension, which he regularly manages with amlodipine. He has no history of diabetes, heart disease, head trauma, or a family history of mental illness.

[0065] Baseline assessment: Cognitive function: MMSE score 21 (moderate cognitive impairment), MoCA score 16, ADAS-Cog score 28.5; the main manifestations are: impaired episodic memory (digit span: forward 5, backward 3), decreased executive function (Stroop color-word test reaction time increased by 35%), and orientation impairment (time orientation 2 / 5, place orientation 3 / 5).

[0066] Imaging findings: Cranial MRI showed bilateral hippocampal atrophy (MTA score grade 2 on the left and grade 3 on the right) and mild enlargement of the ventricular system; PET-CT showed positive amyloid deposition in the bilateral parietal and temporal lobes (SUVR value 1.85).

[0067] Cerebrospinal fluid: Aβ42 380 pg / mL (decreased), Aβ42 / Aβ40 ratio 0.062 (decreased), τ protein phosphorylation p-tau181 65 pg / mL (increased).

[0068] Physiological state: Resting-state EEG alpha band power decreased (occipital region alpha index 0.35), theta band power increased (frontal region theta / alpha ratio 1.8), heart rate variability SDNN 28ms (decreased), SpO2 96%, scalp temperature 36.5℃.

[0069] Intervention plan: Individualized phototherapy parameters are generated based on the regulation strategy model: Light wavelength: 810nm (primary) + 1064nm (secondary) combination, 810nm is used for cortical intervention, and 1064nm enhances deep penetration; Irradiation target areas: bilateral prefrontal cortex (Fp1, Fp2), bilateral hippocampal corresponding scalp projection areas (near P3, P4, based on the 10-20 system localization), and left temporoparietal junction area (CP5). Irradiance: 40 mW / cm² for the prefrontal cortex target area, and 60 mW / cm² for the hippocampal projection area target area. 2 Target area of ​​temporoparietal junction: 35mW / cm²; Irradiation mode: Pulse modulation mode, pulse frequency 10Hz (α-wave rhythm coupling), duty cycle 50%; Irradiation duration: 8 minutes per target, total irradiation duration 24 minutes; Treatment course: 3 times a week (Monday, Wednesday, and Friday), for 8 consecutive weeks as one course of treatment, for a total of 24 interventions.

[0070] Real-time security control: Scalp temperature, SpO2, EEG, and heart rate variability were monitored in real time during each intervention. The highest scalp temperature during the intervention was 37.8℃ (temperature rise of 1.3℃, below the 2℃ non-thermal effect threshold), SpO2 fluctuated between 95% and 97% (fluctuation range ±1%, within the ±10% safety range), and no abnormal spikes were observed in the EEG, nor were the safety thresholds triggered.

[0071] Treatment course assessment (after 8 weeks): Cognitive function: MMSE score improved to 25 (+4), MoCA score improved to 21 (+5), ADAS-Cog score decreased to 22.0 (-6.5); digit span improved to 7 in forward and 5 in backward; Stroop color word test reaction time decreased by 22%; time orientation recovered to 4 / 5 and place orientation to 5 / 5.

[0072] Imaging: Follow-up PET-CT showed that the amyloid SUVR value of bilateral parietal and temporal lobes decreased to 1.62 (a decrease of 12.4%); cranial MRI showed no further shrinkage of hippocampal volume.

[0073] Cerebrospinal fluid: Aβ42 increased to 445 pg / mL, Aβ42 / Aβ40 ratio increased to 0.078, and p-tau181 decreased to 52 pg / mL.

[0074] Physiological status: The α index in the occipital region increased to 0.52, the θ / α ratio in the frontal region decreased to 1.2, and the SDNN time increased to 42ms.

[0075] Safety: No adverse events such as headache, dizziness, scalp burn, or seizures occurred during the entire treatment course, and patients tolerated it well.

[0076] As shown in Table 1, patients showed significant improvements in the three core cognitive indicators of MMSE, MoCA, and ADAS-Cog, and the amyloid protein load in the brain was reduced, suggesting that the method of the present invention has a dual effect of improving cognitive function and promoting the clearance of metabolic waste in Alzheimer's disease.

[0077] Table 1 Comparison of cognitive scores before and after intervention in Alzheimer's disease patients in Example 1

[0078] V. Example 2: Cognitive and Motor Intervention for Parkinson's Disease

[0079] Basic information of the subjects: The patient is a 65-year-old female who presented with a 3-year history of right-sided limb tremor and bradykinesia, and recent onset of memory loss and depression. She has no prior medical history and is otherwise healthy. She has been taking Levodopa (dopacarzine) 0.25g three times a day and Pramipexole 0.25mg three times a day, which has provided adequate symptom control, but she has experienced end-of-dose symptoms for the past six months.

[0080] Baseline assessment: Motor function: UPDRS-III score (off period) 38 points, Hoehn-Yahr stage 2.5; right limb resting tremor (++), rigidity (++), bradykinesia (++++), postural instability (+).

[0081] Cognitive function: MMSE score of 24, MoCA score of 19 (mild cognitive impairment, MCI), mainly manifested as decreased executive function (completion time of 125 seconds for Part B of the connecting test), reduced working memory (accuracy of 62% in the 2-back task), and decreased visuospatial ability (5 / 7 in the clock drawing test).

[0082] Mood: HAMD score of 18 (moderate depression), PDQ-39 quality of life questionnaire score of 48.

[0083] Physiological status: Resting-state EEG β band power decreased (central β index 0.28), θ band power increased, heart rate variability LF / HF ratio 1.8 (sympathetic hyperactivity), SpO2 97%, scalp temperature 36.3℃.

[0084] Intervention plan: Individualized phototherapy parameters are generated based on the regulation strategy model: Optical wavelength: 850nm (primary) + 940nm (secondary) combination; Irradiation target areas: bilateral primary motor cortex (C3, C4), bilateral dorsolateral prefrontal cortex (F3, F4), and the scalp projection area corresponding to the right basal ganglia (based on individualized MRI localization, corresponding to 3.5 cm lateral to Cz). Irradiance: 50 mW / cm at the target site of the motor cortex 2 Prefrontal cortex target 40mW / cm 2 The target point in the basal ganglia projection area is 70mW / cm² (using 940nm to enhance deep penetration). Irradiation mode: Combined mode - 20Hz pulse modulation (β-wave rhythm coupling, duty cycle 40%) is used for the motor cortex target, continuous wave mode is used for the prefrontal cortex target, and 10Hz pulse modulation (duty cycle 60%) is used for the basal ganglia projection area. Irradiation duration: 6 minutes per target, total 18 minutes; Treatment course: 4 times a week (Tuesday to Friday), for 6 consecutive weeks as one course of treatment, for a total of 24 interventions; each intervention is performed during the drug "off period" (before taking the medication in the morning).

[0085] Real-time security control: During the intervention, the highest scalp temperature was 37.6℃ (temperature rise of 1.3℃), SpO2 remained stable at 96%–98%, no abnormal discharges were observed in EEG monitoring, no drastic fluctuations were observed in heart rate variability, and the safety threshold was not triggered.

[0086] Treatment course assessment (after 6 weeks): Motor function: UPDRS-III score (off period) decreased to 26 (-12), Hoehn-Yahr stage decreased to stage 2; right limb tremor decreased to (+), muscle rigidity decreased to (+), and bradykinesia significantly improved; the time to end-of-dose phenomenon increased from 3 hours to 4.5 hours after administration, and the daily equivalent dose of levodopa decreased by 12.5%.

[0087] Cognitive function: MMSE score improved to 27 (+3), MoCA score improved to 24 (+5); completion time of Part B of the connecting test was shortened to 85 seconds (-32%), accuracy of the 2-back task improved to 81% (+19%), and score of 7 / 7 in the clock drawing test improved.

[0088] Mood: HAMD score dropped to 9 (mild depression), PDQ-39 score dropped to 32.

[0089] Physiological status: The central β index rose to 0.45, and the LF / HF ratio decreased to 1.2 (autonomic balance improved).

[0090] Safety: No adverse events such as worsening of movement disorders, headache, nausea, or insomnia were reported; patients reported feeling "more relaxed and clearer-headed".

[0091] As shown in Table 2, patients showed improvement in UPDRS-III motor scores, MMSE / MoCA cognitive scores, and HAMD mood scores, indicating that the method of this invention has a multidimensional intervention effect on Parkinson's disease involving motor, cognitive, and emotional aspects, and can be used as an effective adjunct to drug therapy.

[0092] Table 2 Comparison of motor and cognitive scores before and after intervention in Parkinson's disease patients in Example 2 of the present invention.

[0093] VI. Example 3: Cognitive and Emotional Intervention for Depression

[0094] Basic information of the subjects: The patient is a 28-year-old male graduate student who presented with a chief complaint of "depressed mood, loss of interest, and difficulty concentrating for 6 months, worsening in the past 2 weeks." He has no prior history of mental illness, but his mother has a history of depression. His symptoms recently worsened due to academic pressure. He previously took sertraline 50mg once daily for 4 weeks, but discontinued the medication on his own due to gastrointestinal side effects and is currently not taking any medication.

[0095] Baseline assessment: Mood: HAMD-17 score of 24 (major depression), BDI-II score of 32, and suicidal ideation scale score of 4 (passive suicidal ideation, no plan); core symptoms include persistent low mood, loss of interest, decreased energy, sleep disturbances (difficulty falling asleep + early awakening), and decreased appetite.

[0096] Cognitive function: MMSE score of 28 (normal range), but MoCA score of 23 (mild cognitive impairment), mainly manifested as decreased attention (6 numbers forward and 3 numbers backward), impaired executive function (40% increased interference effect of Stroop color word test), reduced working memory (55% accuracy in 2-back task), and slowed information processing speed (42 out of 90 numbers completed in digit symbol conversion test).

[0097] EEG: Resting-state prefrontal cortex alpha band power asymmetry (left F3 alpha power significantly higher than right F4, Alpha lateralization index -0.35, suggesting insufficient activity in the left prefrontal cortex), and increased theta band power (frontal cortex / alpha ratio 2.1).

[0098] Physiological status: Heart rate variability SDNN 22ms (significantly reduced), LF / HF ratio 2.3 (sympathetic hyperactivity), SpO2 98%, scalp temperature 36.6℃.

[0099] Intervention plan: Individualized phototherapy parameters are generated based on the regulation strategy model: Light wavelength: 810nm (single wavelength); Irradiation target areas: dorsolateral prefrontal cortex of the left frontal lobe (F3, for insufficient activity of the left frontal lobe), dorsolateral prefrontal cortex of the right frontal lobe (F4, low-dose balanced stimulation), and scalp projection areas corresponding to the bilateral anterior cingulate cortex (near Fz). Irradiance: 50 mW / cm at the left F3 target point 2 (High-dose activation), right-side F4 target 20mW / cm 2 (Low-dose equilibrium), anterior cingulate cortex Fz target 40mW / cm 2 ; Irradiation mode: Pulse modulation mode, pulse frequency 40Hz (γ wave rhythm, promoting neural plasticity and BDNF expression), duty cycle 50%; Irradiation duration: 10 minutes for F3 on the left side, 5 minutes for F4 on the right side, 8 minutes for Fz on the right side, for a total irradiation duration of 23 minutes; Treatment course: 5 times a week (Monday to Friday), for 4 consecutive weeks as one course of treatment, for a total of 20 interventions; each intervention is conducted from 9:00 to 11:00 am (in conjunction with circadian rhythm regulation).

[0100] Real-time security control: During the intervention, the highest scalp temperature was 37.9℃ (temperature rise of 1.3℃), SpO2 remained stable at 97%–99%, no abnormalities were found in EEG monitoring, and heart rate variability showed a benign trend of gradually increasing SDNN during the intervention, without triggering the safety threshold.

[0101] Treatment course assessment (after 4 weeks): Mood: HAMD-17 score decreased to 8 (relief, reduction rate 66.7%), BDI-II score decreased to 10, and suicidal ideation scale score was 0; the patient reported significant improvement in mood, restored interest, improved sleep (sleep onset time shortened from 60 minutes to 20 minutes, early awakening disappeared), and normal appetite.

[0102] Cognitive function: MoCA score improved to 27 (+4); number span improved to 8 in forward and 6 in backward; interference effect of Stroop color word test decreased by 35%; accuracy of 2-back task improved to 82% (+27%); number of completed digit symbol conversion test improved to 68 / 90 (+62%).

[0103] EEG: The prefrontal alpha lateralization index rose to -0.08 (close to normal symmetry), and the frontal θ / α ratio dropped to 1.3.

[0104] Physiological status: SDNN increased to 45ms (+104.5%), and the LF / HF ratio decreased to 1.1 (autonomic balance was restored).

[0105] Follow-up: One month after the end of the treatment, the HAMD-17 score remained at 9 points, with no signs of relapse, and cognitive function remained stable.

[0106] Safety: No adverse events such as headache, manic phase, or worsening insomnia were observed, and patient tolerance and compliance were good.

[0107] As shown in Table 3, patients showed significant improvements in HAMD depression scores, MoCA cognitive scores, and 2-back working memory accuracy. Furthermore, the asymmetry of the prefrontal cortex and autonomic nerve function returned to normal, suggesting that the method of this invention has a comprehensive regulatory effect on the emotion-cognition-brain function network of depression, and is especially suitable for non-pharmacological intervention in patients with drug intolerance or treatment-resistant depression.

[0108] Table 3 Comparison of depression scale and cognitive score before and after intervention in patients with depression in Example 3

[0109] VII. Other Implementation Methods and Extensions

[0110] In addition to the three typical embodiments mentioned above, this invention can also be applied to cognitive intervention for other neurodegenerative diseases and emotional disorders, including but not limited to: vascular dementia (VaD), Lewy body dementia (DLB), frontotemporal dementia (FTD), Huntington's disease, multiple sclerosis-related cognitive impairment, post-traumatic brain injury (TBI) cognitive impairment, post-stroke cognitive impairment, anxiety disorder, bipolar disorder, cognitive impairment in schizophrenia, attention deficit hyperactivity disorder (ADHD), and autism spectrum disorder (ASD). Targeting the pathological characteristics of different diseases, the regulatory strategy model can output differentiated combinations of phototherapy parameters: for example, for vascular dementia, it can increase the weight of targets and parameters related to cerebral blood flow; for anxiety disorder, it can use a low-irradiance, long-duration relaxation mode; and for ADHD, it can use high-frequency pulses (such as 20Hz–40Hz) to enhance prefrontal cortex arousal levels.

[0111] In terms of hardware implementation, the near-infrared emitting unit 21 can adopt a fiber-coupled laser diode, a vertical cavity surface-emitting laser (VCSEL) array, or a high-power near-infrared LED, with wavelengths selectable or combined from 808nm, 830nm, 850nm, 905nm, 940nm, 980nm, and 1064nm; the wearable structure can be designed as a headband, helmet, cap, or half-helmet to adapt to different usage scenarios (clinical, home, portable); the physiological signal acquisition module 3 can further integrate a functional near-infrared spectroscopy (fNIRS) channel to realize real-time monitoring of cerebral blood oxygenation and cerebral blood flow, thereby providing more accurate central feedback signals.

[0112] In terms of algorithms, the regulation strategy model can further incorporate deep learning architectures (such as convolutional neural networks (CNN) to process EEG temporal signals and graph neural networks (GNN) to process brain functional connectivity networks), and combine federated learning to achieve multi-center data collaborative training without leaking patient privacy; the safety closed-loop control can introduce predictive maintenance algorithms to predict potential risks and adjust parameters in advance based on historical physiological data.

[0113] In terms of system integration, this invention can be linked with virtual reality (VR) / augmented reality (AR) cognitive training systems to perform cognitive task training while undergoing near-infrared phototherapy, and enhance the cognitive training effect by utilizing the neural plasticity time window induced by phototherapy; it can also be used in conjunction with other neuromodulation technologies such as transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) to form a multimodal neuromodulation scheme.

[0114] In clinical application, this invention's phototherapy intervention, as an adjunct rehabilitation method, does not preclude the subject's existing standard drug treatment, nor does it force the subject to discontinue their original clinical medications. Before intervention, it is necessary to exclude individuals with contraindications to phototherapy, such as those with scalp lesions, intracranial metal implants, a history of photosensitizing drug use, or photosensitivity diseases. During the intervention, close attention is paid to the subject's subjective experience. If the subject experiences subjective adverse sensations such as dizziness, scalp tingling, or eye discomfort, even if physiological indicators have not yet reached the safety threshold, manual suspension or termination of phototherapy is supported to further improve the safety of the subject's use.

[0115] The treatment duration, intervention frequency, and combinations of targets and irradiance given in the various embodiments of this invention are merely illustrative and not intended to limit the scope of protection of this invention. Those skilled in the art can flexibly adjust the intervention frequency, single irradiation duration, target combinations, and irradiance of each target within the parameter range described in this invention, based on the severity of the subject's condition, age, individual differences in scalp and skull anatomy, and tolerance level. The intervention process can be entirely automated by the device or conducted under the full supervision of medical personnel; when used in home rehabilitation scenarios, it is recommended that family members or caregivers accompany the subject to facilitate timely capture of subjective discomfort feedback.

[0116] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A transcranial near-infrared phototherapy modulation method suitable for cognitive intervention in neurodegenerative diseases and emotional disorders, characterized in that, The method is based on a transcranial near-infrared phototherapy control system, utilizing near-infrared light within the 800nm–1100nm optical treatment window to penetrate the subject's scalp and skull to reach the cerebral cortex and deep brain regions, intervening in brain functional networks with photochemical signals dominated by non-thermal effects; the method includes the following steps: S1. Obtain cognitive function assessment data and physiological state parameters of the subject. The cognitive function assessment data includes at least one score of attention, memory, and executive function. The physiological state parameters include at least one of electroencephalogram (EEG) signal, heart rate variability, and blood oxygen saturation. S2. Based on the cognitive function assessment data and physiological state parameters, transcranial near-infrared phototherapy parameters are generated through a preset regulation strategy model. The transcranial near-infrared phototherapy parameters include light wavelength, irradiance, irradiation duration, irradiation mode, and irradiation target area. S3. Drive the near-infrared light source module in the transcranial near-infrared phototherapy control system according to the parameters of the transcranial near-infrared phototherapy, and irradiate the preset brain region target point on the subject's head with transcranial near-infrared light, so that the photons are absorbed by the mitochondria of neurons to enhance the ATP energy of brain cells and promote the removal of metabolic waste. S4. During the irradiation process, the physiological feedback signal of the subject is collected in real time and compared with the preset safety threshold. When the physiological feedback signal exceeds the preset safety threshold, the transcranial near-infrared phototherapy parameters are dynamically adjusted or the irradiation is terminated.

2. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The transcranial near-infrared phototherapy control system includes: a main control module, a near-infrared light source module, a physiological signal acquisition module, a human-computer interaction module, and a power supply module; The main control module is electrically connected to the near-infrared light source module, the physiological signal acquisition module, and the human-computer interaction module, respectively, and is used to execute the regulation strategy model and output control commands. The near-infrared light source module includes multiple near-infrared light-emitting units, which are arranged in an array and correspond to different brain regions of the subject's head. The emission wavelength of the near-infrared light-emitting units is 808nm to 1064nm. The physiological signal acquisition module is used to acquire the subject's electroencephalogram (EEG) signals, heart rate variability, and blood oxygen saturation. The human-computer interaction module is used to input cognitive function assessment data and display phototherapy parameters and physiological feedback signals; The power module is used to supply power to the various modules of the system.

3. The transcranial near-infrared phototherapy modulation method according to claim 2, characterized in that, The irradiance range of the near-infrared emitting unit is 1 mW / cm². 2 ~100mW / cm 2 The output power of each near-infrared emitting unit is independently adjustable; the near-infrared light acts on brain tissue in a non-thermal manner, and the temperature of the subject's scalp surface does not rise by more than 2°C during irradiation.

4. The transcranial near-infrared phototherapy modulation method according to claim 2, characterized in that, The near-infrared light source module also includes a heat dissipation unit and a light intensity detection unit. The light intensity detection unit collects the actual output light intensity of the near-infrared light-emitting unit in real time and feeds it back to the main control module. The main control module performs closed-loop correction based on the deviation between the actual output light intensity and the target light intensity.

5. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The method for constructing the preset regulation strategy model in step S2 includes: We collected cognitive function assessment data, physiological state parameters, and corresponding transcranial near-infrared phototherapy response data from patients with neurodegenerative diseases and emotional disorders to construct a training dataset. Using the degree of improvement in cognitive function and the efficiency of clearing metabolic waste in the brain as labels, a machine learning algorithm was used to train the training dataset to obtain a mapping relationship model; The mapping relationship model is used as the regulation strategy model to output the optimal transcranial near-infrared phototherapy parameters based on the input cognitive function assessment data and physiological state parameters.

6. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The irradiation mode in step S3 includes continuous wave mode, pulse modulation mode, and a combination of the two; the pulse frequency range of the pulse modulation mode is 1Hz to 100Hz, and the duty cycle range is 10% to 90%.

7. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The target area for irradiation in step S3 includes at least one of the prefrontal cortex, the scalp projection area corresponding to the hippocampus, and the temporoparietal junction; the near-infrared light source module is fixed to the subject's head by a wearable headband or helmet-like structure, so that the positional deviation between each near-infrared emitting unit and the corresponding target area does not exceed 5mm.

8. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The preset safety thresholds in step S4 include: scalp surface temperature not exceeding 40°C, blood oxygen saturation fluctuation not exceeding ±10% of the baseline value, and abnormal spike wave occurrence rate in EEG signals not exceeding a preset frequency. When any one of these exceeds the corresponding threshold, the main control module reduces the irradiance or shortens the irradiation time. When two or more exceed the threshold simultaneously, the irradiation is immediately terminated and an alarm is issued.

9. The transcranial near-infrared phototherapy modulation method according to claim 1, characterized in that, The method also includes a treatment management step: setting a treatment cycle according to the subject's disease type and degree of cognitive impairment, with each treatment cycle containing multiple phototherapy interventions, and each intervention interval being no less than 24 hours; after each treatment cycle, re-acquiring cognitive function assessment data, and updating the transcranial near-infrared phototherapy parameters for the next treatment cycle based on the assessment results.

10. The transcranial near-infrared phototherapy modulation method according to claim 2, characterized in that, The transcranial near-infrared phototherapy control system also includes a wireless communication module. The main control module is connected to a remote terminal through the wireless communication module to upload cognitive function assessment data, physiological state parameters and phototherapy records to the remote terminal, and to receive the control strategy model update parameters sent by the remote terminal.

Images