CN-121747968-B - Dynamic early warning system for physiological characteristics of patient with endocrine treatment of prostate cancer

CN121747968BCN 121747968 BCN121747968 BCN 121747968BCN-121747968-B

Abstract

The invention relates to the technical field of software and discloses a dynamic early warning system for physiological characteristics of a patient with prostate cancer endocrine treatment, which comprises a core processing module, a risk score and a system, wherein the core processing module is configured to dynamically allocate analysis weights for different types of monitoring data based on the current endocrine treatment duration time of the patient, wherein the analysis weights for high-frequency physiological symptom data are increased in a treatment induction period, the analysis weights for function and behavior data are increased in a long-term maintenance period, an individuation baseline of the monitoring data is set, the analysis weights and the individuation baseline are adaptively adjusted according to the periodical clinical calibration data, and the accumulated integral of deviation of the monitoring data relative to the individuation baseline in a sliding time window is calculated according to a preset function or behavior index to obtain the risk score.

Inventors

YANG YAN
QIN YANWEN
YIN HONGFAN
XI HUIQIN
ZHANG TING
Yuan Xiuqun
PAN CHEN
WANG BEIBEI

Assignees

上海交通大学医学院附属仁济医院

Dates

Publication Date: 20260508
Application Date: 20260302

Claims (9)

1. A dynamic pre-warning system for physiological characteristics of a patient treated by endocrine treatment of prostate cancer, the system comprising: The data acquisition and access module is used for acquiring monitoring data of a patient and comprises high-frequency physiological symptom data, function and behavior data, subjective state data and periodic clinical calibration data in a hospital scene; A core processing module configured to: Dynamically allocating analysis weights of different types of monitoring data based on the current endocrine treatment duration of a patient, wherein the analysis weights of high-frequency physiological symptom data are improved in a treatment induction period, and the analysis weights of function and behavior data are improved in a long-term maintenance period; Setting an individuation baseline of the monitoring data; adaptively adjusting the analysis weights and the personalized baselines according to the periodic clinical calibration data; calculating a cumulative integral of the deviation of the monitored data from the personalized baseline over a sliding time window for a predetermined function or behavioral indicator, calculating a risk score, and The early warning output module is used for generating and outputting risk early warning information based on the dynamically allocated analysis weight, the adaptive adjustment result and the risk score; The risk score formula is: Wherein, the A monitoring index value corresponding to the moment t is represented; is an individualized baseline; is a time weighting factor; T is the integration window length.
2. The system of claim 1, wherein the data acquisition and access module further comprises a physiological signal acquisition unit for acquiring high frequency physiological symptom class data, comprising: Heart rate variability rate, namely continuously acquiring heart rate signals through a photoplethysmogram (PPG) sensor, and calculating SDNN and RMSSD; The sleep structure parameters are that based on the accelerometer and the heart rate signal, the sleep process is analyzed to obtain the deep sleep time length, the rapid eye movement sleep (REM) duty ratio and the night awakening frequency (WASO); Hot flashes occur at a frequency that, in combination with a Galvanic Skin Response (GSR) signal and a sudden heart rate rise characteristic, automatically identifies and marks the time, duration and frequency of occurrence of vasomotor symptoms.
3. The system of claim 1, wherein the data acquisition and access module further comprises a function and behavior signal parsing unit for acquiring function and behavior class data, comprising: The muscle strength related index is that a triaxial accelerometer is utilized to capture the micro-vibration amplitude change in the upper limb movement process, and a grip dynamometer is optionally matched to obtain grip strength data; Gait stability characteristics, namely calculating the variability of the step frequency and the time duty ratio of the double support phases through gait cycle analysis; The standing/walking time is that the action mode of the transition of the patient from the sedentary state to the standing state is identified, and the time required for the patient to stand from the static sitting posture to the stable standing posture is calculated; daily activity, counting the effective number of steps per day and the duration of medium-high strength physical activity (MVPA).
4. The system of claim 1, wherein the data acquisition and access module further comprises a subjective state interaction unit, the subjective state interaction unit acquires subjective state data actively reported by the patient through the mobile terminal, and the subjective state data comprises: Normalizing the fatigue score; psychological stress scoring; occurrence of a near fall event; Hot flashes were self-rated.
5. The system of claim 1, wherein the data acquisition and access module further comprises a hospital low frequency periodic calibration module that obtains clinical test data with a time resolution of a month or a season, and periodically calibrates baseline parameters of a home monitoring algorithm to obtain periodic clinical calibration data, the periodic clinical calibration data comprising: prostate Specific Antigen (PSA) levels; Serum testosterone levels; bone metabolism related index, including bone density T value and/or bone transition marker.
6. The system of claim 1, wherein setting an individualized baseline for the monitored data comprises: Taking the latest hospital period calibration data of the patient as a reference, and on the premise of confirming that the relevant low-frequency clinical indexes are in a stable or normal range, selecting a home continuous monitoring data average value in a first preset period after the calibration node as an individuation baseline of the monitoring data.
7. The system of claim 1, further comprising a risk classification and response output unit comprising: The weighted scores of the monitoring indexes and the corresponding accumulated risk scores are all in a preset safety interval, no abnormal trend or event signal is detected, And storing and archiving the relevant monitoring data in a silent mode, displaying a long-term change trend to a patient or a medical care end in a low-frequency mode, and not triggering instant intervention prompt.
8. The system of claim 1, further comprising a risk classification and response output unit comprising: The single or short term monitoring value does not exceed the traditional alarm threshold, but the accumulated risk score exceeds the corresponding trend early warning threshold, prompting the existence of long-term and progressive functional or state decay risks, Triggering trend early warning prompt and outputting targeted self-management or rehabilitation suggestion instructions to the patient according to the risk characteristic category.
9. The system of claim 1, further comprising a risk classification and response output unit comprising: any of the following is detected: occurrence of a near fall event or other high risk behavioural event; The abrupt abnormal change of the high-frequency physiological signal and the subjective discomfort state show nonlinear cooperative change in the time dimension, and the potential acute risk amplification is indicated; the high priority risk response is triggered immediately, including but not limited to outputting an immediate alert prompt to the patient terminal and sending a high priority risk notification request to the background healthcare management end, prompting that further evaluation or manual intervention is required.

Description

Dynamic early warning system for physiological characteristics of patient with endocrine treatment of prostate cancer Technical Field The invention relates to the technical field of software, in particular to a dynamic early warning system for physiological characteristics of a patient with prostate cancer endocrine treatment. Background Endocrine treatment of prostate cancer is often a chronic intervention process of long duration, and the treatment cycle often lasts for years, even requiring life-long management. During this process, the physiological state and daily activities of the patient are constantly changing, and the large amount of monitoring data generated thereby also presents significant stepwise differences. Most of the current home health management systems or general follow-up software still commonly adopt a static fixed threshold judgment mode in the design of a risk identification algorithm. In addition, in the data processing mode, the existing home management system mostly adopts decision logic based on discrete single-point sampling, namely, only the current numerical value acquired at a certain time point is compared with a preset standard to make a decision, and the continuity analysis of historical data is not carried out. It can be seen that existing data processing techniques are difficult to adapt to the actual need for data features to evolve over time throughout the cycle of prostate cancer ADT treatment and generally lack the ability to identify long-term, weak, but sustained cumulative risk. The traditional system based on static rules and discrete judgment is difficult to effectively reflect the real change of the risk state of a patient when facing a long-period home management scene. Disclosure of Invention The invention aims to solve the technical defects caused by the common adoption of a static fixed threshold value and a single-point discrete sampling risk judgment algorithm in the existing prostate cancer ADT home management system. Specifically, the existing system is easy to generate a large number of false positive alarms due to high-frequency low-risk fluctuation of physiological indexes in the treatment induction period, and is difficult to identify progressive risks with slow numerical variation and continuously accumulated in the long-term maintenance period, so that missed judgment on partial hidden risks is formed. A dynamic pre-warning system for physiological characteristics of a patient treated by endocrine treatment of prostate cancer, comprising: The data acquisition and access module is used for acquiring monitoring data of a patient and comprises high-frequency physiological symptom data, function and behavior data, subjective state data and periodic clinical calibration data in a hospital scene; A core processing module configured to: Dynamically allocating analysis weights of different types of monitoring data based on the current endocrine treatment duration of a patient, wherein the analysis weights of high-frequency physiological symptom data are improved in a treatment induction period, and the analysis weights of function and behavior data are improved in a long-term maintenance period; Setting an individuation baseline of the monitoring data; adaptively adjusting the analysis weights and the personalized baselines according to the periodic clinical calibration data; calculating a cumulative integral of the deviation of the monitored data from the personalized baseline over a sliding time window for a predetermined function or behavioral indicator, calculating a risk score, and And the early warning output module is used for generating and outputting risk early warning information based on the dynamically allocated analysis weight, the adaptive adjustment result and the risk score. The data acquisition and access module also comprises a physiological signal acquisition unit, wherein the physiological signal acquisition unit is used for acquiring high-frequency physiological symptom data and comprises the following components: Heart rate variability rate, namely continuously acquiring heart rate signals through a photoplethysmogram (PPG) sensor, and calculating SDNN and RMSSD; The sleep structure parameters are that based on the accelerometer and the heart rate signal, the sleep process is analyzed to obtain the deep sleep time length, the rapid eye movement sleep (REM) duty ratio and the night awakening frequency (WASO); Hot flashes occur at a frequency that, in combination with a Galvanic Skin Response (GSR) signal and a sudden heart rate rise characteristic, automatically identifies and marks the time, duration and frequency of occurrence of vasomotor symptoms. The data acquisition and access module also comprises a function and behavior signal analysis unit, wherein the function and behavior signal analysis unit is used for acquiring function and behavior data and comprises the following components: The muscle strength related index is that a