CN-122024404-A - Epileptic seizure early warning and alarming system and method

CN122024404ACN 122024404 ACN122024404 ACN 122024404ACN-122024404-A

Abstract

The invention discloses an epileptic seizure early warning system and method, and particularly relates to the field of epileptic seizure monitoring and early warning, which comprises the steps of executing synchronous verification on a dangerous segment chain corresponding to seizure risk and a qualification interval corresponding to a current perception state and a wireless networking state, and calling a different-position evidence supplementing node to conduct segmented evidence supplementing and result review when an evidence deficiency interval exists.

Inventors

JIANG XIAOLING
Hua Heliu
WU WENBAO
DU SHAOYI
TIAN ZHIQIANG
ZENG WEI
GUO XINGLONG
WANG QINGHUI
LAN JINKAI
LIU FENGLIN
CHEN YANGUI
PAN LIYING
ZHANG BAIXIANG

Assignees

龙岩学院

Dates

Publication Date: 20260512
Application Date: 20260407

Claims (10)

1. The epileptic seizure early warning and alarming method is characterized by comprising the following steps of: s1, acquiring physiological data, behavior data, node data and link data transmitted by a patient side wearing node, a bedside node and a central node through wireless networking in a current monitoring period, executing alignment according to a unified period identifier and an arrival time sequence, and outputting a period data sequence; s2, based on physiological data and behavior data in the periodic data sequence, extracting abnormal segments with consistent abnormal directions between the current monitoring period and the previous monitoring period, performing serial connection according to the time phase relation, and outputting a dangerous segment chain; s3, extracting effective segments which simultaneously meet the continuity of contact, the online continuity of nodes and the continuity of link transmission in the current monitoring period based on node data and link data in the period data sequence, and executing combination according to a time coverage relation to output qualification intervals; s4, performing time overlapping verification on the dangerous segment chain and the qualification interval, generating a formal alarm candidate result when the dangerous segment chain is completely covered by the qualification interval, generating a limited early warning candidate result and a evidence missing interval when the dangerous segment chain is not completely covered by the qualification interval, and outputting a layering result; S5, for the limited early warning candidate result, selecting a supplementary evidence node with a different acquisition position from the current invalid node from the wireless network according to the evidence lack section, the dangerous segment chain and the current invalid node identification, acquiring supplementary evidence data of each dangerous segment corresponding to the supplementary evidence node, executing sectional supplementary evidence, supplementary evidence repeated check and recalculation of the qualification section in the evidence lack section, and outputting an updating qualification section and an updating result.
2. The seizure early warning method of claim 1, further comprising: And S6, executing qualification sending confirmation on the formal alarm candidate result or the updated result, sending the qualified result to the care terminal and the central node, writing in an alarm record, and outputting an epileptic seizure early warning alarm result.
3. The seizure early warning method of claim 2, characterized by: The S1 comprises the following steps: S1-1, calculating time difference offset of each node relative to a central node according to node identifications, packet sending time and receiving time of a patient side wearing node, a bedside node and the central node in a wireless networking, and outputting a node time difference table; S1-2, converting the arrival time of various data into correction time under a unified period identifier based on physiological data, behavior data, node data, link data and a node time difference table, classifying the correction time into a corresponding period according to the unified period identifier, executing time sequence rearrangement according to the correction time, and outputting a correction data sequence; S1-3, identifying a cross-node connection interval and performing interval splicing according to the time interval and source node change of adjacent data in the correction data sequence, and outputting a periodic data sequence.
4. The seizure early warning method as in claim 3, wherein: The step S2 comprises the following steps: s2-1, respectively calculating the numerical value difference and the change direction of the current monitoring period relative to the previous monitoring period according to physiological data and behavior data in the period data sequence, determining data segments with the same change direction and the same numerical value difference as candidate segments, and outputting a candidate segment set; S2-2, calculating segment interval duration based on the end time and the start time of the front and rear adjacent candidate segments in the candidate segment set, performing head-tail concatenation on the candidate segments with the segment interval duration smaller than the duration of a single monitoring period and consistent change direction, and outputting a homodromous segment chain; S2-3, extracting a co-occurrence part of the simultaneous segment and a continuous extension part of the adjacent time segment according to a time overlapping interval and a front-back connection interval of the physiological data segment and the behavior data segment in the homodromous segment chain, and outputting a dangerous segment chain.
5. The seizure early warning method of claim 4, characterized by: The step S3 comprises the following steps: S3-1, respectively identifying a continuous period in which the contact state of each node is kept unchanged, a continuous period in which the online state is kept unchanged and a continuous period in which the transmission result is kept successful according to node data and link data in a periodic data sequence in a time sequence, and outputting a contact segment, an online segment and a transmission segment; S3-2, calculating a time intersection covered by the contact segment, the online segment and the transmission segment of the same node, determining the time intersection as an effective segment, and outputting an effective segment set; S3-3, sequentially executing fragment deletion, overlapping fragment merging and splicing fragments according to the time inclusion relation, the time overlapping relation and the head-to-tail connection relation of the front and back effective fragments in the effective fragment set, and outputting qualification intervals.
6. The seizure early warning method of claim 5, characterized by: the step S4 comprises the following steps: s4-1, calculating the time overlapping part of each dangerous segment and the qualification interval according to the starting and ending time of each dangerous segment in the dangerous segment chain and the starting and ending time of the qualification interval, and outputting an overlapping segment set; S4-2, calculating the residual part which does not fall into the qualification interval based on the starting and ending time of each dangerous segment and the starting and ending time of the corresponding overlapping segment, determining the residual part as the evidence-missing interval, and outputting a coverage result and the evidence-missing interval; S4-3, generating a formal alarm candidate result when all dangerous segments do not have the evidence missing section according to the coverage result and the evidence missing section, generating a limited early warning candidate result when any dangerous segment has the evidence missing section, and outputting a layering result.
7. The seizure early warning method of claim 6, characterized by: The step S5 comprises the following steps: S5-1, according to the evidence missing section, the dangerous segment chain, the current failure node identification and the wireless networking node position table in the limited early warning candidate result, respectively corresponding each evidence supplementing node with a different acquisition position from the current failure node to each dangerous segment in the evidence missing section, calculating the coverage time, the starting time difference and the ending time difference of each evidence supplementing node to the evidence missing section corresponding to each dangerous segment, and outputting the evidence supplementing node sequence corresponding to each dangerous segment; S5-2, based on the sequence of the evidence supplementing nodes corresponding to the dangerous segments, the starting and ending time of the evidence missing section and the starting and ending time of the dangerous segments, sending an evidence supplementing request to a front evidence supplementing node in the sequence of the evidence supplementing nodes, receiving returned evidence supplementing data, calculating a time overlapping section, a direction consistent section and a direction opposite section of the evidence supplementing data and the corresponding dangerous segments, determining the corresponding evidence supplementing data as the evidence supplementing segments when the time overlapping section exists and the direction consistent section is greater than the direction opposite section, otherwise, writing the corresponding dangerous segments into the segments to be rechecked, and outputting a evidence supplementing segment set and a segments to be rechecked.
8. The seizure early warning method of claim 7, characterized by: The step S5 further includes: S5-3, according to the later-order evidence supplementing node, the evidence supplementing segment set and the qualification interval in the order of the evidence supplementing node to be rechecked, sending evidence supplementing requests to the later-order evidence supplementing node and receiving returned evidence supplementing data, respectively calculating a time overlapping interval, a direction consistent interval and an end-to-end connection interval of the evidence supplementing data returned by the later-order evidence supplementing node and the corresponding dangerous segment, adding the corresponding evidence supplementing data into the evidence supplementing segment set when the time overlapping interval exists and the direction consistent interval is greater than the opposite direction interval, adding the corresponding evidence supplementing data into the evidence supplementing segment set when the time overlapping interval does not exist but is connected with the front end and the rear end of the qualification interval, otherwise, keeping the corresponding dangerous segment in the to be rechecked segment set, and outputting an updated evidence supplementing segment set and an updated to-be rechecked segment set; s5-4, based on the updated supplementary evidence segment set, the updated fragment set to be rechecked, the absent evidence interval and the qualification interval, performing segmented backfilling on the absent evidence interval covered by the updated supplementary evidence segment set and performing interval combination with the qualification interval to form an updated qualification interval, reserving an uncovered part corresponding to the updated fragment set to be rechecked as a residual absent evidence interval, and outputting an updated qualification interval and the residual absent evidence interval; s5-5, recalculating a coverage result of the dangerous segment chain according to the updated qualification interval, the remaining evidence missing interval and the dangerous segment chain, and when the dangerous segment chain is covered by the updated qualification interval and the remaining evidence missing interval is empty, rewriting the limited early warning candidate result into a formal warning candidate result, otherwise, keeping the limited early warning candidate result, writing the remaining evidence missing interval into the updated result, and outputting the updated result.
9. The seizure early warning method of claim 8, characterized by: The step S6 comprises the following steps: S6-1, extracting a result type, a starting time, an ending time and a corresponding dangerous segment chain according to a formal alarm candidate result or an updating result, calculating a time overlapping interval and a segment overlapping part with the result type, the starting time, the ending time and the corresponding dangerous segment chain of the latest sent result in an alarm record, determining a current result as a new sending result when the time overlapping interval is empty, determining the current result as a concurrent result when the time overlapping interval is not empty and the segment overlapping part exists, otherwise, determining the current result as a resending result, and outputting a sending qualification result; S6-2, based on the qualification sending result, generating complete alarm data comprising a result type, a starting time, an ending time and a corresponding dangerous segment chain for the new sending result, sending the complete alarm data to the care terminal and the central node, generating supplementary alarm data comprising a newly added time part and a newly added dangerous segment for the concurrent result, sending the supplementary alarm data to the care terminal and the central node, stopping sending the new sending result, reserving the current result, and outputting a sending result; s6-3, writing the result type, the starting time, the ending time, the corresponding dangerous segment chain and the sending state of the current result into an alarm record according to the sending qualification result and the sending result, and outputting the epileptic seizure early warning alarm result.
10. An epileptic seizure early warning system, comprising: the alignment management module is used for acquiring physiological data, behavior data, node data and link data transmitted by the patient side wearing node, the bedside node and the central node through wireless networking in the current monitoring period, executing alignment management according to the unified period identification and the arrival time sequence, and outputting a period data sequence; the segment serial connection module extracts abnormal segments with consistent abnormal directions between the current monitoring period and the previous monitoring period based on physiological data and behavior data in the period data sequence, and performs serial connection according to the time-phase connection relationship to output a dangerous segment chain; The qualification extraction module is used for extracting effective fragments which simultaneously meet the continuity of contact, the online continuity of nodes and the continuity of link transmission in the current monitoring period based on node data and link data in the period data sequence, and executing combination according to a time coverage relationship to output qualification intervals; The layering judgment module is used for performing time overlapping verification on the dangerous segment chain and the qualification interval, generating a formal alarm candidate result when the dangerous segment chain is completely covered by the qualification interval, generating a limited early warning candidate result and a failure interval when the dangerous segment chain is not completely covered by the qualification interval, and outputting a layering result; The evidence-supplementing recalculation module is used for selecting evidence-supplementing nodes with different acquisition positions from the wireless networking according to the evidence-lacking interval, the dangerous segment chains and the current invalid node identifiers for the limited early warning candidate results, acquiring evidence-supplementing data of the evidence-supplementing nodes corresponding to dangerous segments, executing segmented evidence supplementing, evidence-supplementing recalculation and qualification interval recalculation on the evidence-lacking interval, and outputting an updating qualification interval and an updating result; And the result sending module is used for executing qualification sending confirmation on the formal alarm candidate result or the updated result, sending the qualified result to the care terminal and the central node, writing in an alarm record and outputting an epileptic seizure early warning alarm result.

Description

Epileptic seizure early warning and alarming system and method Technical Field The invention relates to the technical field of epileptic seizure monitoring and early warning, in particular to an epileptic seizure early warning system and method. Background In the epileptic seizure early warning technology, the existing system mainly focuses on whether the seizure risk of a patient can be recognized as early as possible and timely triggers warning, generally, a wearable monitoring terminal, a bedside monitoring device and a mobile receiving terminal are arranged on the patient side, information such as brain electricity, heart rate, skin electricity, body movement, blood oxygen and the like is continuously collected, and is uploaded to a carer terminal or a central receiving terminal through a wireless networking link after being locally judged; Taking home night continuous monitoring as an example, a patient is in an operation condition of frequent sleep turn-over, random limb compression, wearing contact state change, fluctuation of battery allowance and dynamic change of wireless networking link quality among bedside nodes, a handheld terminal and a central receiving end, and the system is required to continuously complete risk judgment and also is required to respond to a current real and reliable monitoring basis for each alarm sent outwards; however, under this condition, in the prior art, once the front end acquisition state, the node access state or the link transmission state deviates, the alarm judgment still always uses the existing threshold value, the existing baseline or the existing uploading path to continue outputting the result, the phenomena that the same patient can directly observe and recheck during the sensing contact variation, the abnormal increase of the alarm times during the node switching or link jitter, the rising of the alarm withdrawal proportion, the inconsistent front and rear alarm conclusion under the similar scene and the like easily occur in actual use, but the on-line detection or the fault self-detection of the equipment usually only can indicate that the device is still in the working state, and cannot indicate whether the current alarm conclusion still has the sending basis under the existing sensing condition and the wireless networking state; therefore, the application aims to solve the technical problem of synchronously judging the sending qualification of the alarm conclusion by combining the current perception state and the wireless networking state in the epileptic seizure early warning process so as to avoid that an unreliable alarm is output outwards when basic distortion is monitored. Disclosure of Invention In order to overcome the above-mentioned drawbacks of the prior art, embodiments of the present invention provide an epileptic seizure early warning system and method, which solve the problems set forth in the background art above by performing synchronous verification on a dangerous segment chain corresponding to seizure risk and a qualification interval corresponding to a current perception state and a wireless networking state, and calling a different-position evidence-supplementing node to perform segment evidence-supplementing and result review when an evidence-missing interval exists. In order to achieve the purpose, the invention provides the following technical scheme that the epileptic seizure early-warning and alarming method comprises the following steps: s1, acquiring physiological data, behavior data, node data and link data transmitted by a patient side wearing node, a bedside node and a central node through wireless networking in a current monitoring period, executing alignment according to a unified period identifier and an arrival time sequence, and outputting a period data sequence; s2, based on physiological data and behavior data in the periodic data sequence, extracting abnormal segments with consistent abnormal directions between the current monitoring period and the previous monitoring period, performing serial connection according to the time phase relation, and outputting a dangerous segment chain; s3, extracting effective segments which simultaneously meet the continuity of contact, the online continuity of nodes and the continuity of link transmission in the current monitoring period based on node data and link data in the period data sequence, and executing combination according to a time coverage relation to output qualification intervals; s4, performing time overlapping verification on the dangerous segment chain and the qualification interval, generating a formal alarm candidate result when the dangerous segment chain is completely covered by the qualification interval, generating a limited early warning candidate result and a evidence missing interval when the dangerous segment chain is not completely covered by the qualification interval, and outputting a layering result; S5, for the limited early warning candidate result