CN-121971040-A - Predictive intervention system for mandibular transient syndrome

CN121971040ACN 121971040 ACN121971040 ACN 121971040ACN-121971040-A

Abstract

The technical scheme of the invention discloses a predictive intervention system for mandibular transient syndrome, which is characterized by working in a threshold calibration stage and a monitoring and intervention stage, and comprising a signal acquisition module, a processing and decision module and an intervention feedback module, wherein the intervention feedback module is used for executing a preset intervention triggering step after receiving a triggering intervention signal sent by the processing and decision module in the monitoring and intervention stage. The technical scheme of the invention has diversity in concrete implementation (such as multiple signals of video, myoelectricity and the like can be adopted), is not limited by a single technical path, has good expansibility and adaptability, and provides multiple solutions for different clinical manifestations and patient demands.

Inventors

DING XIA
XING YUE
LI JIN
ZHOU TIANYI
CAI YUCHEN
LIN MING
LIU JIANNAN
SONG XUEFEI

Assignees

上海交通大学医学院附属第九人民医院

Dates

Publication Date: 20260505
Application Date: 20260114

Claims (10)

1. A predictive intervention system for mandibular transient syndrome, wherein the predictive intervention system operates in a threshold calibration phase, in which a user is directed to perform a specified oral action, and a monitoring and intervention phase, in which the user is alerted and performs an autonomous intervention training, the predictive intervention system comprising a signal acquisition module for acquiring oral physiological signals and ocular physiological signals in real time, capable of reflecting the movements of the user's oral and ocular and the associated neuromuscular activity states, wherein in the threshold calibration phase the signal acquisition module acquires the oral physiological signals and the ocular physiological signals associated therewith, and in the monitoring and intervention phase the signal acquisition module acquires at least oral physiological signals; the processing and decision module further comprises a feature extraction unit, a threshold calibration unit which is triggered only in a threshold calibration stage and a monitoring and intervention unit which is triggered only in a monitoring and intervention stage, wherein the feature extraction unit is used for extracting the eye signal features and/or the oral signal features of the eye physiological signals and/or the oral physiological signals acquired by the signal acquisition module, the threshold calibration unit is used for selecting the eye signal features of which the eye actions of the left eye or the right eye accord with the mandibular transient syndrome features from all the eye signal features acquired in a certain time period, further acquiring all the oral signal features associated with the selected eye signal features, the processing and decision module establishes an individuation early warning model of a current user based on the obtained oral signal characteristics, the monitoring and intervention unit calls the individuation early warning model, the individuation early warning model judges whether oral actions are overlarge or not based on the relation between the real-time oral signal characteristics obtained by the signal acquisition module and the characteristic extraction unit in the monitoring and intervention stage and the oral data calibration value, if yes, a trigger intervention signal is generated, and the intervention feedback module is used for executing a preset intervention triggering step after receiving the trigger intervention signal sent by the processing and decision module in the monitoring and intervention stage.
2. A predictive intervention system for a transient mandibular syndrome according to claim 1, wherein the signal acquisition module covers simultaneously the oral area and the ocular area of the same user, thereby enabling simultaneous acquisition of the oral physiological signal and the ocular physiological signal, or the signal acquisition module covers alternately the oral area and the ocular area of the same user, thereby enabling alternate acquisition of the oral physiological signal and the ocular physiological signal.
3. A predictive intervention system for a transient mandibular syndrome according to claim 1, wherein the oral physiological signal or the ocular physiological signal is an optical image signal or the oral physiological signal or the ocular physiological signal is an electrophysiological signal.
4. A predictive intervention system for a transient mandibular syndrome in accordance with claim 3 wherein if the oral physiological signal and the ocular physiological signal acquired by the signal acquisition module are of different types during the threshold calibration phase, a time stamp based sliding window mechanism is employed to correlate the optical image signal and the electrophysiological signal such that the two types of signals are aligned in time.
5. The predictive intervention system for mandibular transient syndrome of claim 3, wherein when the oral physiological signal or the ocular physiological signal is the electrophysiological signal, the eye signal feature or the oral signal feature extracted by the feature extraction unit at least includes a time domain signal feature, the threshold calibration unit selects an eye signal feature of which eye motion of a left eye or a right eye conforms to a mandibular transient syndrome feature from all the eye signal features obtained in a certain period of time by judging a relationship between the time domain signal feature and a resting baseline value among the eye signal features, wherein the resting baseline value is obtained by calculating a time domain feature of an eye physiological signal of the same user in a resting state, and the personalized early warning model judges whether the oral motion is excessive based on a relationship between the time domain signal feature of the oral signal feature and the oral data calibration value, thereby generating the trigger intervention signal.
6. The predictive intervention system for mandibular transient syndrome of claim 5, wherein when the oral physiological signal or the ocular physiological signal is the electrophysiological signal, the ocular signal feature or the oral signal feature extracted by the feature extraction unit further comprises a frequency domain feature, wherein the threshold calibration unit judges whether a time domain signal feature and a resting baseline value in the ocular signal feature satisfy a preset relation while judging whether the time domain signal feature and the resting baseline value in the ocular signal feature satisfy the preset relation or not, and excludes the ocular signal feature outside a specific frequency spectrum range by using the frequency domain feature in the ocular signal feature to exclude electrical noise, and wherein the individuation early warning model judges whether a muscle enters a fatigue state based on the frequency domain feature in the oral signal feature in a process of judging whether the oral motion is excessive or not based on the relation between the time domain signal feature in the oral signal feature and the oral data calibration value, and updates the oral data calibration value if so.
7. The predictive intervention system for mandibular transient syndrome of claim 6, wherein the personalized early warning model does not directly generate the trigger intervention signal after judging that the oral motion amplitude is excessive at the current moment, but generates the trigger intervention signal after judging that the oral motion amplitude is excessive by the personalized early warning model for a continuous period of time.
8. A predictive intervention system for a mandibular transient syndrome as claimed in claim 3, wherein when the oral physiological signal or the ocular physiological signal is the optical image signal, the ocular signal features extracted by the feature extraction unit include at least an ocular amplitude geometrical feature for characterizing a left eye or a right eye eyelid movement amplitude obtained by an eye contour key set, the oral signal features extracted by the feature extraction unit are weighted values of a single oral amplitude geometrical feature or a plurality of oral amplitude geometrical features for characterizing an oral movement amplitude obtained by an oral contour key set, and the threshold calibration unit selects an ocular signal feature in which an ocular motion of a left eye or a right eye conforms to a mandibular transient syndrome feature among all the ocular signal features obtained over a certain period of time based on a relation between the ocular amplitude geometrical feature and a closed eye geometrical feature threshold among the ocular signal features.
9. The predictive intervention system for a transient mandibular syndrome of claim 8, wherein the eye signal features extracted by the feature extraction unit further comprise eye frequency geometric features obtained by an eye contour key set for characterizing a frequency of eye lid movement of the left or right eye when the eye physiological signal is the optical image signal, wherein the threshold calibration unit is based on a relationship between the eye frequency geometric features in the eye signal features and a preset resting threshold in addition to a relationship between the eye amplitude geometric features in the eye signal features and a closing eye geometric feature threshold when the eye signal features of the left or right eye are selected to be in accordance with the eye signal features of the transient mandibular syndrome.
10. The predictive intervention system for mandibular transient syndrome of claim 8, wherein when the oral physiological signal is the optical image signal, the personalized early warning model calculates a normalized value based on the real-time oral signal characteristic and the oral data calibration value, and judges whether the oral action is too large based on a relation between the normalized value and a preset threshold value, if the oral action of the current frame is too large, the personalized early warning model does not directly generate the trigger intervention signal, but in N continuous frames, the personalized early warning model obtains a conclusion of the oral action being too large, and the real-time oral data characteristic is incremental, and confirms the generation of the trigger intervention signal.

Description

Predictive intervention system for mandibular transient syndrome Technical Field The invention relates to a clinical intervention system for treating mandibular transient syndrome. Background The mandibular transient syndrome is a disease with a clinically manifested specific neuromuscular dysfunction, and is characterized by a central pathological feature in that specific oral movements (e.g. mouth opening, chewing, tooth indication, etc.) actively performed by the patient can trigger a series of involuntary, pathological ocular coordinated reactions. Such ocular linked responses are often manifested as upward eyelid lifting, transient eye, enlarged eyelid fissures, and the like, of the affected side eyelid. The above symptoms have various significant effects on the daily life of the patient. Firstly, in social and psychological aspects, the incapacitated and strange facial expression is very easy to cause the patient to generate strong embarrassment, anxiety and spelt, and seriously hinders the normal social interaction and interpersonal interaction. Secondly, at the functional level, frequent eyelid jump or position changes can introduce visual disturbances, creating subjective visual shaking, resulting in distraction, and thus affecting daily activities such as reading, driving, delicate operations, etc. that require visual attention. In addition, sustained abnormal facial activity itself can also cause discomfort to the patient. Therefore, the mandibular transient syndrome is not only an appearance problem, but also a functional disorder which seriously affects the life quality and physical and mental health of patients, and the requirement for effective intervention schemes is clear and urgent in clinic. At present, clinical intervention schemes aiming at mandibular transient syndrome are mainly divided into three categories, but all have significant limitations: the first category is surgical intervention, i.e. by frontal valve/levator blephar surgery to improve the eyelid position in the patient's quiescent state. However, this solution has the fundamental drawback of: Firstly, the symptoms and root causes are not cured, namely, the operation can only passively correct the static appearance, but can not block the core pathological path of the eye symptom triggered by the mouth movement, so that the involuntary linkage eye symptom still occurs when a patient opens the mouth, chews and the like; Secondly, surgery is used as an invasive treatment, with inherent anesthesia risk, infection risk, and possible complications (such as eyelid insufficiency); third, the surgical effect is often irreversible and is difficult to retrieve once it has had adverse consequences. The second type of solution is functional intervention, such as botulinum toxin injection, which can alleviate symptoms to some extent, but the focus is on weakening muscle reaction or blurring appearance, and the effect is limited on the basic goal of reducing the occurrence frequency of abnormal linkage, and side effects such as facial expression stiffness may be accompanied. In addition, current behavioral intervention and autonomic control approaches rely on the subjective consciousness of the patient to counteract oral movement or inhibit ocular response. The drawbacks of this solution are also highlighted: Firstly, a feedback mechanism which is objective and quantitative is lacking, closed-loop control cannot be formed, the training process is completely dependent on subjective feeling of a patient, and a real-time and accurate index is not used for evaluating a control effect, so that the training is difficult to systemize and standardize, and positive feedback cannot be provided for the patient to maintain training power; Secondly, the effect is very unstable, the attention and the mind of the patient are seriously relied on, the continuous mental burden is caused to the patient, the long-term adherence is difficult, and the effectiveness in various scenes of daily life cannot be ensured. There is always a lack of a fundamental method in clinic that can accurately "decouple" the pathological links of oral movements and abnormal eye muscle contractions, thereby functionally eliminating symptoms. Thus, the prior art, in the face of the clinical need for transient mandibular syndrome, has clear technical gap, lacking an effective solution that enables noninvasive, quantifiable, and real-time prediction and active intervention immediately prior to onset of symptoms. Specifically, the existing means cannot realize the treatment mode transition from 'passive response' to 'active prevention', and also fail to provide a closed-loop intervention system integrating objective monitoring, intelligent early warning and instant feedback. This gap severely limits the therapeutic effect of the disease and the quality of life of the patient, and also highlights the urgency and innovation value of the technical problem to be solved by the invention. D