Search

CN-122025005-A - Orthopedics rehabilitation effect monitoring method and system based on artificial intelligence

CN122025005ACN 122025005 ACN122025005 ACN 122025005ACN-122025005-A

Abstract

The invention relates to the technical field of patient monitoring, in particular to an orthopedics rehabilitation effect monitoring method and system based on artificial intelligence, comprising the following steps: the invention relates to a method for acquiring action time and angle sequence, combining action time and angle sequence into record set, constructing time chain, correlating angle change, comparing time difference with angle amplitude mark rhythm, judging rhythm continuity to form trend segment, distinguishing efficiency improvement and slowing and generating structured input set, in the invention, traceable action record is constructed by start-stop time and angle sequence acquired by sensor, time chain is formed by finishing time of multiple training in sequence, the training rhythm is clearly presented in a continuous structure according to the speed change of the training rhythm, the structural input set of the recovery efficiency is combined according to the rhythm attribute, different training batch performances can be compared under the same frame, progress stagnation or efficiency decline is revealed in advance, the recovery process is presented as a directional trend chain, and the recognition of a high-yield training stage and the adjustment of a low-efficiency interval are facilitated.

Inventors

  • LIU XIAOCHEN
  • BAI XUEJUAN
  • TANG XIAONAN
  • CHENG SHI

Assignees

  • 南通大学

Dates

Publication Date
20260512
Application Date
20260129

Claims (10)

  1. 1. An orthopedics rehabilitation effect monitoring method based on artificial intelligence is characterized by comprising the following steps: S1, acquiring action start-stop time recorded by a sensor, calculating the completion time, orderly arranging joint angle sequences, and combining the completion time and the angle sequences to form a rehabilitation action original record set; S2, according to the original record set of rehabilitation actions, sequentially forming a time chain of action completion time of multiple training, calculating adjacent time differences and associating the adjacent time differences with corresponding angle change sections to generate a joint action time chain structure; S3, comparing the time difference with the angle change amplitude based on the joint action time chain structure, accelerating the section marking rhythm with short time difference and continuous angle, slowing down the section marking rhythm with long time difference and slow angle, and forming an action rhythm characteristic structure; S4, judging consistency of adjacent rhythm marks according to the action rhythm characteristic structure, dividing the same mark section into continuous sections with the same trend, setting a difference mark section as a new section starting point, and sequentially combining to generate a continuous section set with a recovery trend; s5, judging attributes of the trend segments according to the continuous segment set of the recovery trend, classifying the rhythm accelerating segment into an efficiency improving segment, classifying the rhythm slowing segment into an efficiency slowing segment, and combining according to a training sequence to obtain a recovery efficiency structured input set.
  2. 2. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 1, wherein the rehabilitation action original record set comprises action start-stop time label data, joint angle change time sequence data and training action identification information, the joint action time chain structure comprises a training sequence completion time chain, adjacent action completion time difference association information and joint angle change section mapping relation, the action rhythm feature structure comprises a rhythm acceleration mark sequence, a rhythm slowing mark sequence and a rhythm feature pattern index, the recovery trend continuous section set comprises a same trend continuous section set, a trend change starting position record and section sequence index information, and the recovery efficiency structuring input set comprises an efficiency improvement section set, an efficiency slowing section set and a training sequence mapping structure.
  3. 3. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 1, wherein the specific steps of S1 are as follows: S101, acquiring time information of an action start gesture signal and time information of an action end gesture signal, calculating time quantity between start and stop of an action according to the two time information, performing corresponding verification on an action interval boundary according to time quantity and signal time sequence, and generating action duration time quantity according to the interval boundary; S102, calling a joint angle sequence formed by the sensor record based on the action duration time, correspondingly analyzing adjacent angle values and adjacent time points in the angle sequence, and classifying intervals according to the corresponding angle change condition and the action duration time to form an angle sequence arrangement order, and generating a joint angle change amplitude measurement according to the arrangement order; and S103, calling the time information of the action start gesture signal and the time information of the action end gesture signal according to the measurement of the joint angle change amplitude, sequentially matching the joint angle sequence with the two pieces of time information, and sequentially integrating the matched angle values with time points to form an action whole-process structure, and generating a rehabilitation action original record set according to the structure.
  4. 4. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 3, wherein the specific steps of S2 are as follows: S201, arranging the completion time of the rehabilitation appointed action generated by multiple training based on the original record set of the rehabilitation action according to the training sequence, comparing the arranged completion time according to the sequence relation, and carrying out chain connection on the completion time in the sequence before and after the time obtained by the comparison, and generating an action time chain sequence quantity according to the connected time sequence; s202, invoking the action time chain sequence quantity to calculate adjacent completion time in the sequence, correspondingly analyzing the calculated time difference value and adjacent time points, classifying the time difference value into continuous sections according to an adjacent relation, forming a time difference sequence structure according to the change condition of the difference value in the sections, and generating the action time difference sequence quantity according to the structure; S203, calling the joint angle change sections in the original record set of rehabilitation actions according to the action time difference sequence quantity, enabling the time difference sequence and the angle change sections to correspond item by item, sequentially integrating the corresponding sequence values and the angle change values, and only reserving the paired sequences in the integrated content to generate a joint action time chain structure.
  5. 5. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 4, wherein the specific steps of S3 are as follows: S301, comparing the time difference sequence with the joint angle change amplitude according to the joint action time chain structure, correspondingly judging the time difference value and the angle change quantity in the time difference sequence according to the sequence position, and obtaining the rhythm acceleration section quantity by taking the section with short time difference and continuous angle change as the rhythm acceleration section according to the judgment result; S302, judging the corresponding angle change quantity of an uncalibrated section call in a time difference sequence based on the rhythm accelerating section quantity, identifying a section with prolonged time difference and slowed angle change as a rhythm slowing section according to a judging result, and sorting the slowing section according to a sequence position to generate a rhythm slowing section quantity; s303, calling the rhythm acceleration section quantity according to the rhythm slowing section quantity, combining the two types of sections according to the original sequence, and carrying out a corresponding rhythm characteristic arrangement structure with the position of the combined section identification sequence and the joint action time chain structure to generate an action rhythm characteristic structure.
  6. 6. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 5, wherein the specific steps of S4 are as follows: S401, based on the action rhythm feature structure, consistency judgment is carried out on adjacent rhythm marks, the adjacent rhythm marks are compared item by item according to sequence positions, continuous paragraphs with the same marks are identified according to comparison results, the identified paragraphs are summarized and sorted according to a continuous relation, and rhythm continuous segment quantity is generated according to the sorted continuous segment structure; S402, calling the continuous segment quantity of the rhythm to judge rhythm marks inconsistent with the continuous segment in the action rhythm characteristic structure, taking the position of the inconsistent mark as a starting point of a new segment, and carrying out sequence connection on the independent segment structure with the starting point position and adjacent rhythm marks, and generating new segment quantity of the rhythm according to the connected segment structure; S403, calling continuous rhythm segment quantity according to the new rhythm segment quantity, combining the two types of segments according to the time chain sequence, and correspondingly forming trend arrangement content according to the positions of the combined segment sequences in the original action rhythm characteristic structure, and generating a continuous recovery trend segment set according to the arrangement content.
  7. 7. The orthopaedics rehabilitation effect monitoring method based on artificial intelligence according to claim 6, wherein the specific steps of S5 are as follows: S501, performing attribute judgment on the trend segments according to the continuous segment set of the recovery trend, comparing the rhythm acceleration mark and the rhythm release mark in the trend segments with the current mark of the trend segments according to the sequence positions, classifying the trend segments where the rhythm acceleration mark is positioned into an efficiency improvement class according to the comparison result, classifying the trend segments where the rhythm release mark is positioned into an efficiency trend release class, and generating trend attribute classification quantity according to the classification result; S502, calling a recovery trend continuous section set based on the trend attribute classification quantity, arranging the classified trend sections according to a training sequence, correspondingly comparing the time sequence of the trend sections after arrangement with the trend attribute classification quantity, arranging the trend sections of the same category into a combined section structure according to the continuous sequence, and generating a trend combined section quantity according to the combined section structure; s503, calling trend attribute classification quantity according to the trend combination segment quantity, integrating the efficiency lifting segment and the efficiency slowing segment in the combination segment according to the original training sequence, and performing position correspondence with the continuous segment set of the recovery trend by using the integrated segment sequence to form efficiency characteristic arrangement content, and generating a recovery efficiency structured input set according to the arrangement content.
  8. 8. The orthopedic rehabilitation effect monitoring method based on artificial intelligence according to claim 1, wherein the sensor is a sensing device fixed near a joint to be rehabilitated of a patient, can output posture information reflecting the change of the angle of the joint, and the data is derived from an inertial measurement unit, an angle encoder or a gyroscope; The angle sequence is a sequence formed by a plurality of angle values recorded by the wearable joint angle sensor according to time in the process of executing rehabilitation appointed actions by a patient; the rehabilitation action original record set is a data set obtained by combining the rehabilitation appointed action completion time and the corresponding joint angle sequence; the time chain is a time sequence formed by arranging a plurality of rehabilitation appointed action completion times according to a training occurrence sequence; the time difference is a sequence which is obtained by arranging the difference between the completion time of two adjacent rehabilitation appointed actions in the action time chain; The angle change section is an angle record set in a time range corresponding to a single rehabilitation appointed action or a single time difference in a joint angle sequence; The joint action time chain structure is a composite data structure formed by establishing association of an action time chain, a time difference sequence and a corresponding joint angle change section.
  9. 9. The artificial intelligence based orthopedic rehabilitation effect monitoring method according to claim 1, wherein the action rhythm feature structure is a structured sequence obtained by arranging rhythm marks in a rhythm feature mode in time sequence; The rhythm acceleration is a mark set for a joint angle change section with short time difference and continuous angle change; the rhythm slowing is a mark set for a joint angle change section with longer time difference and gradual angle change and is used for reflecting the rhythm slowing condition of section action execution; The trend continuous segment is a time segment formed by continuously arranging a plurality of adjacent rhythm marks with the same mark type in the action rhythm characteristic structure; the recovery trend continuous section set is a data set obtained by combining a plurality of trend continuous sections in time sequence; the efficiency improving section is a trend continuous section corresponding to the rhythm accelerating mark in the recovery trend continuous section set; the efficiency slowing section is a trend continuous section of the recovery trend, wherein the trend continuous section is concentrated and corresponds to the rhythm slowing mark; The recovery efficiency structured input set is a structured data set formed by combining an efficiency improving section and an efficiency slowing section according to a training sequence.
  10. 10. An artificial intelligence based orthopedic rehabilitation effect monitoring system for implementing the artificial intelligence based orthopedic rehabilitation effect monitoring method of any of claims 1-9, the system comprising: the action record acquisition module is used for executing S1, namely acquiring time information of an action start gesture signal and an action end gesture signal acquired by the wearable joint angle sensor, calculating rehabilitation appointed action completion time, arranging joint angle sequences formed by the sensor records according to the appearance sequence, and combining the completion time information with the angle sequences to generate a rehabilitation action original record set; The rhythm association construction module is used for executing S2, based on the original record set of rehabilitation actions, arranging the rehabilitation appointed action completion time generated by multiple training according to the training sequence to form an action time chain, executing calculation on the adjacent action completion time in the action time chain to obtain a time difference sequence, associating the sequence with the corresponding joint angle change section to generate a joint action time chain structure; The rhythm pattern extraction module is used for executing S3, according to the joint action time chain structure, comparing the time difference sequence with the joint angle change amplitude, marking the section with shorter time difference and continuous angle change, marking the section with longer time difference and gradual angle change, marking the section with slow rhythm, combining the marks in sequence and forming a rhythm characteristic pattern, and obtaining an action rhythm characteristic structure; the trend section generating module is used for executing S4, namely executing consistency judgment on adjacent rhythm marks based on the action rhythm characteristic structure, recording rhythm sections with the same mark as the same trend continuous section, taking inconsistent parts as starting points of new sections, and sequentially combining all the trend continuous sections to generate a recovery trend continuous section set; And S5, executing attribute judgment on each trend segment according to the continuous segment set of the recovery trend, classifying the segment with the accelerated marking rhythm as an efficiency improving segment, classifying the segment with the slowed marking rhythm as an efficiency slowing segment, and combining the segments into an integral structure according to a training sequence to obtain the recovery efficiency structured input set.

Description

Orthopedics rehabilitation effect monitoring method and system based on artificial intelligence Technical Field The invention relates to the technical field of patient monitoring, in particular to an orthopedics rehabilitation effect monitoring method and system based on artificial intelligence. Background The technical field of patient monitoring mainly relates to continuous acquisition and quantitative evaluation of physiological states and functional states of patients, and the core matters of the technical field comprise measurement of athletic performances of specific parts, extraction of key characteristics of rehabilitation processes and mode processing of monitoring data. The traditional orthopaedics rehabilitation effect monitoring method is a mode of recording the joint movement range and muscle strength performance of a patient around the orthopaedics function recovery process, and according to subclass standards such as a joint movement degree measuring scale and a gait observing scale, rehabilitation state interpretation is completed by manually judging the difference of movement tracks and time nodes, and the traditional mode is mostly implemented by means of manually visually measuring and recording movement amplitude, manually comparing scale grades and the like. In the prior art, in the rehabilitation monitoring process of a patient, the observer is relied on to record approximate activity performance and muscle strength grades around the joint function recovery process, the action amplitude is estimated by naked eyes, grade judgment is given according to the table entry, the state of focusing on a small amount of key time is recorded, and continuous tracking is not available for each repeated action in one training. In this operation mode, the time information often exists in a fuzzy start-stop description, so that it is difficult to form a time sequence which can be connected in series between multiple training, the completion efficiency of different schedules or different stages cannot be directly compared on the same time chain, and the evaluation result is easy to stay in a rough improvement or is not described in a clear change. Because the activity track is stored by means of site recall and simple notes of an observer, the details such as rhythm change, midway pause, acceleration or deceleration and the like in a single action process are very easy to ignore, for example, the first few rhythms are stable when a patient finishes ten flexion and extension actions, the later few rhythms are obviously slowed down due to fatigue, the traditional record always only leaves the impression of integral time or angle, and the high-efficiency training interval and the efficiency gliding interval are difficult to accurately point out. The monitoring process has higher experience dependence, different evaluators have judging differences in visual angle, feeling muscle strength and interpretation gait, and the same patient can obtain inconsistent conclusion due to subjective preference of observers when being evaluated at different time, so that the recovery curve in long-term follow-up visit lacks a quantifiable and reproducible trend structure. In the face of the rehabilitation plan adjustment requirement, the traditional record only provides scattered scale grades and text remarks, so that the problems that whether the training rhythm is too fast or too slow, whether the action control is more coherent and the like are difficult to support and make fine judgment are solved, the training intensity adjustment is easy to lag, the high-load stage is not recognized in time, or the low-efficiency rhythm is maintained for a long time, and the overall rehabilitation period and the functional reconstruction quality are influenced. Disclosure of Invention In order to solve the technical problems in the prior art, the embodiment of the invention provides an orthopedics rehabilitation effect monitoring method based on artificial intelligence, which comprises the following steps: S1, acquiring action start-stop time recorded by a sensor, calculating the completion time, orderly arranging joint angle sequences, and combining the completion time and the angle sequences to form a rehabilitation action original record set; S2, according to the original record set of rehabilitation actions, sequentially forming a time chain of action completion time of multiple training, calculating adjacent time differences and associating the adjacent time differences with corresponding angle change sections to generate a joint action time chain structure; S3, comparing the time difference with the angle change amplitude based on the joint action time chain structure, accelerating the section marking rhythm with short time difference and continuous angle, slowing down the section marking rhythm with long time difference and slow angle, and forming an action rhythm characteristic structure; S4, judging consistency of adjacent r