CN115252962A - Intelligent infusion alarm method and system based on machine vision - Google Patents

Intelligent infusion alarm method and system based on machine vision Download PDF

Info

Publication number
CN115252962A
CN115252962A CN202210805180.5A CN202210805180A CN115252962A CN 115252962 A CN115252962 A CN 115252962A CN 202210805180 A CN202210805180 A CN 202210805180A CN 115252962 A CN115252962 A CN 115252962A
Authority
CN
China
Prior art keywords
infusion
data
time
transfusion
speed data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210805180.5A
Other languages
Chinese (zh)
Inventor
商临萍
郭威
赵文婷
孙沛
杨雅茜
刘美荣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
First Hospital of Shanxi Medical University
Original Assignee
First Hospital of Shanxi Medical University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by First Hospital of Shanxi Medical University filed Critical First Hospital of Shanxi Medical University
Priority to CN202210805180.5A priority Critical patent/CN115252962A/en
Publication of CN115252962A publication Critical patent/CN115252962A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61MDEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M5/00Devices for bringing media into the body in a subcutaneous, intra-vascular or intramuscular way; Accessories therefor, e.g. filling or cleaning devices, arm-rests
    • A61M5/14Infusion devices, e.g. infusing by gravity; Blood infusion; Accessories therefor
    • A61M5/168Means for controlling media flow to the body or for metering media to the body, e.g. drip meters, counters ; Monitoring media flow to the body
    • A61M5/16831Monitoring, detecting, signalling or eliminating infusion flow anomalies
    • A61M5/1684Monitoring, detecting, signalling or eliminating infusion flow anomalies by detecting the amount of infusate remaining, e.g. signalling end of infusion

Landscapes

  • Health & Medical Sciences (AREA)
  • Vascular Medicine (AREA)
  • Engineering & Computer Science (AREA)
  • Anesthesiology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Hematology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Infusion, Injection, And Reservoir Apparatuses (AREA)

Abstract

The invention discloses an intelligent transfusion alarm method based on machine vision, which belongs to the technical field of medical treatment and machine vision, and is characterized in that a monitoring camera in the existing transfusion scene is used for collecting images of a transfusion container and images of a drip cup based on the prior art, and a pre-trained machine learning model is used for identifying the images, so that volume data of different transfusion containers and diameter data of drip tubes of different types can be effectively obtained, and then dripping speed data is obtained, so that timely alarm on transfusion conditions can be realized; in addition, the infusion monitoring device does not need to set a fixed reference line or a fixed reference value, and can perform relative change setting according to different infusion container types and infusion container capacities, so that intelligent infusion monitoring can be performed on different infusion container types, infusion container capacities and infusion speeds, the universality is high, and different use scenes can be met.

Description

Intelligent infusion alarm method and system based on machine vision
Technical Field
The invention relates to the technical field of medical treatment and machine vision, in particular to an intelligent infusion alarm method and system based on machine vision.
Background
The intravenous infusion is a method for infusing a large amount of sterile liquid, electrolyte and medicine into the body from veins by utilizing the principles of atmospheric pressure and hydrostatic pressure, and is a conventional administration technology of clinical medical treatment; in the process of intravenous infusion of a patient, an attendant or a nurse is often required to effectively monitor the infusion process so as to prevent the blood return phenomenon caused by the completion of the infusion, and if the patient is not reminded in time, great accidents are likely to be caused; at present, besides tracking and monitoring the infusion condition by an attendant or a nurse, methods for monitoring the infusion condition by a weight sensor, an infrared sensor or a pressure sensor and the like and giving an alarm appear, for example: chinese patent No. CN103463698B discloses a pressure-sensitive sensor infusion alarm system, which needs to pre-install a sensor on an infusion bottle or an infusion bag to realize infusion monitoring alarm, which undoubtedly increases the production cost of the infusion bottle or the infusion bag, and further increases the treatment cost of patients, even if the sensors are externally arranged and reused in a reusable equipment manner, the cost is very high, and in a scene with many patients, each infusion bottle or infusion bag needs to be additionally installed with one-to-one reusable equipment, the subsequent maintenance cost is higher, and compatibility needs to be considered, and one-to-many infusion monitoring cannot be realized;
compared with a sensor-type monitoring alarm system, the above problems can be solved well by using image recognition, and at present, some methods for performing infusion monitoring alarm through image recognition appear, for example: chinese patent No. CN111298238B discloses an infusion early warning method based on image recognition, but the method needs to preset an early warning line on an infusion container, and the practical applicability is low; in addition, the Chinese patent No. CN105498042B discloses a non-light-shading type transfusion automatic alarm method and a non-light-shading type transfusion automatic alarm device based on videos, although the invention can realize real-time alarm according to the liquid level in a drip cup and the liquid speed in the drip cup, the alarm is too urgent and does not reserve the buffer time for changing medicines or removing needles for nurses, and if the nurses face a plurality of scenes, the hands and feet of the nurses cannot take measures; in addition, because the types of the infusion containers (infusion bottles or infusion bags are common), the capacity of the infusion containers and the infusion speed adjusted by nurses are different, the method of setting fixed preset values is difficult to be applied practically, the preset values of the speed and the preset values of the early warning lines of the infusion containers are difficult to be unified due to the influence of multiple factors of audiences, and if the preset values are input one by nurses, the risk of misoperation is undoubtedly caused;
therefore, an intelligent infusion alarm method is needed to solve the above problems.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides an intelligent infusion alarm method and system based on machine vision.
In order to achieve the purpose, the invention adopts the following technical scheme:
an intelligent infusion alarm method based on machine vision comprises the following steps:
1) Acquiring a patient transfusion container image at a first transfusion time point based on a monitoring camera in the existing transfusion scene, and identifying the transfusion container image based on a first machine learning model to acquire volume data of a transfusion container;
2) After volume data of the infusion container are obtained, collecting images of the drip cup based on the monitoring camera, and identifying based on a second machine learning model to obtain diameter data of the drip tube;
3) After diameter data of a dropper are acquired, a dropper video at a second infusion time point is acquired by using the monitoring camera, and the dropper video is subjected to video processing based on a preset time interval to calculate first dropping speed data in the dropper, wherein the preset time interval is 1 minute;
4) Correcting the first dripping speed data based on the diameter data, calculating a time difference value between the second transfusion time point and the first transfusion time point, and simultaneously calculating the time difference value and the corrected first dripping speed data to obtain error data;
5) Subtracting error data from the volume data of the infusion container to obtain current infusion volume data, and calculating the remaining infusion time according to the corrected first dripping speed data;
6) Judging whether the residual transfusion time meets a preset early warning time range in real time, and if so, performing early warning for the first time;
7) When the residual transfusion time meets a preset early warning time range, secondarily collecting the drip cup video in real time by using the monitoring camera, calculating second dripping speed data in the drip cup, simultaneously judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio, and if so, carrying out secondary emergency warning.
Furthermore, the first machine learning model and the second machine learning model are generated based on the marked basic data in the infusion device database as a training set, and the basic data in the infusion device database at least comprises volume data of infusion containers of different types and diameter data of droppers of different types.
Further, the first machine learning model and the second machine learning model are specifically at least one of logistic regression, random forest, K nearest neighbor, support vector machine, linear discriminant analysis, naive bayes, and neural network.
Further, the video processing is performed on the water dropping kettle based on the preset time interval to calculate the first dropping speed data in the water dropping kettle, and the method comprises the following steps:
carrying out time line distinguishing on the collected drip cup videos, and extracting 1-minute videos;
and performing key frame marking on the 1-minute video, namely taking the first time that the infusion drop falls on the liquid level as a first key frame, taking the second time that the infusion drop falls on the liquid level as a second key frame, sequentially circulating until all key frames in 1 minute are obtained, and obtaining first dripping speed data in the drip cup according to the key frames.
Further, the error data refers to the transfusion volume data of the human body which is input in the time difference value between the second transfusion time point and the first transfusion time point, namely the transfusion volume which is completed in the time period from the time point when the patient starts to input the transfusion volume data to the time point after the data processing is completed.
Further, the correcting the first drop speed data based on the diameter data, calculating a time difference between the second infusion time point and the first infusion time point, and calculating the time difference and the corrected first drop speed data to obtain error data includes:
firstly, correcting the first dripping speed data according to the diameter data, namely the dripping amount required by 1 milliliter of dropper with different diameters, so as to obtain the corrected first dripping speed data;
then, subtracting the first infusion time point from the second infusion time point to obtain a time difference value;
and finally, multiplying the corrected first drop speed data by the time difference to obtain error data, wherein the formula is as follows:
SW=VX×tc
in the formula: s. theWRepresenting error data; vXRepresenting the corrected first drop velocity data; t is tcRepresenting the time difference.
Further, the remaining infusion time is calculated by the following formula:
Figure BDA0003736814350000051
in the formula: t isZIndicating the remaining infusion time; sZRepresenting infusion container volume data; sWRepresenting error data; vXIndicating the corrected first drop velocity data.
A machine vision based intelligent infusion alarm system comprising:
the first acquisition module is used for acquiring an image of a patient transfusion container at a first transfusion time point;
the first identification module is used for identifying the infusion container image based on a first machine learning model so as to acquire volume data of the infusion container;
the second acquisition module is used for acquiring a drip cup image after acquiring volume data of the infusion container;
the second identification module is used for identifying the drip cup image based on a second machine learning model so as to obtain diameter data of the drip tube;
the third acquisition module is used for acquiring a drip cup video at a second infusion time point after acquiring diameter data of the drip tube;
the first calculation processing module is used for performing video processing on the drip cup video based on a preset time interval so as to calculate first dripping speed data in the drip cup;
the second calculation processing module is used for calculating the time difference value between the second infusion time point and the first infusion time point and calculating and acquiring error data;
the third calculation processing module is used for subtracting error data from the volume data of the infusion container to obtain the current infusion volume data and calculating the residual infusion time;
the infusion early warning module is used for judging whether the residual infusion time meets a preset early warning time range in real time according to the residual infusion time, if so, performing early warning once, and displaying the residual infusion time at the same time;
and the emergency alarm module is used for secondarily acquiring the drip cup video in real time when the residual transfusion time meets a preset early warning time range, calculating second dripping speed data in the drip cup, simultaneously judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio, and if so, carrying out secondary emergency alarm.
Further, the second calculation processing module comprises a drop speed correction unit, and the drop speed correction unit is used for correcting the first drop speed data according to the diameter data, namely according to the drop volume required by 1 ml of dropper with different diameters, so as to obtain the corrected first drop speed data.
Compared with the prior art, the invention has the beneficial effects that:
(1) The application discloses an intelligent transfusion alarm method and system based on machine vision, which collects images of a transfusion container and images of a drip cup by taking a monitoring camera in the existing transfusion scene as a basis and identifies by depending on a pre-trained machine learning model, can effectively acquire volume data of different transfusion containers and diameter data of different types of drip tubes, and then can realize timely alarm on transfusion conditions by acquiring drip speed data;
(2) The application discloses an intelligent transfusion alarm method and system based on machine vision, which corrects dropping speed data according to diameter data of droppers of different types to obtain accurate residual transfusion time, and realizes timely early warning of transfusion conditions by comparing the accurate residual transfusion time with a preset early warning time range, thereby being beneficial to reserving enough buffer time for nurses to change medicines, prepare medicines and dismantle needles for a plurality of people; in addition, by judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio in real time, secondary emergency alarm can be carried out to prevent a nurse from forgetting the operation of changing medicine or removing a needle for a certain patient due to more patients;
(3) Compared with the existing method for monitoring and alarming infusion by image recognition, the intelligent infusion alarming method and system based on machine vision do not need to set a fixed reference line or a fixed reference value, and can carry out relative change setting according to different infusion container types and infusion container capacities, so that the intelligent infusion alarming method and system based on machine vision can carry out intelligent infusion monitoring on different infusion container types, infusion container capacities and infusion speeds, has strong universality, can meet different use scenes, and has certain social and economic benefits.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
FIG. 1 is a general flow chart of an intelligent infusion alarm method based on machine vision according to the present invention;
fig. 2 is a schematic diagram of the overall structure of an intelligent infusion alarm system based on machine vision according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In the description of the present invention, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention.
In one embodiment, referring to fig. 1, a machine vision-based intelligent infusion alarm method is provided, which comprises the following steps:
1) Acquiring a patient transfusion container image at a first transfusion time point based on a monitoring camera in the existing transfusion scene, and identifying the transfusion container image based on a first machine learning model to acquire volume data of a transfusion container;
2) After volume data of the infusion container are obtained, collecting images of the drip cup based on the monitoring camera, and identifying based on a second machine learning model to obtain diameter data of the drip tube;
specifically, the first machine learning model and the second machine learning model are generated based on labeled basic data in a transfusion device database as a training set, and the basic data in the transfusion device database at least comprises volume data of different types of transfusion containers and diameter data of different types of droppers.
Specifically, the first machine learning model and the second machine learning model are at least one of logistic regression, random forest, K nearest neighbor, support vector machine, linear discriminant analysis, naive bayes and neural network;
at this point it should be noted that: the first machine learning model and the second machine learning model may be the same model or different models, and any model capable of realizing the technical point may be applied, and the invention is not particularly limited thereto.
3) After diameter data of a dropper are acquired, a dropper video at a second infusion time point is acquired by using the monitoring camera, and the dropper video is subjected to video processing based on a preset time interval to calculate first dropping speed data in the dropper, wherein the preset time interval is 1 minute;
specifically, the video processing of the drip chamber based on the preset time interval to calculate the first dripping speed data in the drip chamber includes:
carrying out time line distinguishing on the collected drip cup videos, and extracting 1-minute videos;
and performing key frame marking on the 1-minute video, namely taking the first time that the infusion drop falls on the liquid level as a first key frame, taking the second time that the infusion drop falls on the liquid level as a second key frame, sequentially circulating until all key frames in 1 minute are obtained, and obtaining first dripping speed data in the drip cup according to the key frames.
4) Correcting the first dripping speed data based on the diameter data, calculating a time difference value between the second transfusion time point and the first transfusion time point, and simultaneously calculating the time difference value and the corrected first dripping speed data to obtain error data;
specifically, the error data refers to the transfusion volume data of the human body input in the time difference between the second transfusion time point and the first transfusion time point, that is, the transfusion volume completed in the period from the time point when the patient starts to input the transfusion volume data to the time point after the data processing is completed.
Specifically, the correcting the first drop velocity data based on the diameter data, calculating a time difference between the second infusion time point and the first infusion time point, and calculating the time difference and the corrected first drop velocity data to obtain error data includes:
firstly, correcting the first dripping speed data according to the diameter data, namely the dripping amount required by 1 milliliter of dropper with different diameters, so as to obtain the corrected first dripping speed data;
then, subtracting the first infusion time point from the second infusion time point to obtain a time difference value;
and finally, multiplying the corrected first drop speed data by the time difference to obtain error data, wherein the formula is as follows:
SW=VX×tc
in the formula: sWRepresenting error data; vXRepresenting the corrected first drop velocity data; t is tcRepresenting the time difference.
5) Subtracting error data from the volume data of the infusion container to obtain current infusion volume data, and calculating the remaining infusion time according to the corrected first dripping speed data;
specifically, the calculation formula of the remaining infusion time is as follows:
Figure BDA0003736814350000101
in the formula: t isZIndicating the remaining infusion time; sZRepresenting infusion container volume data; sWRepresenting error data; vXIndicating the corrected first drop velocity data.
6) Judging whether the residual transfusion time meets a preset early warning time range in real time, and if so, performing early warning for the first time;
7) And when the residual transfusion time meets a preset early warning time range, secondarily collecting the drip cup video in real time by using the monitoring camera, calculating second dripping speed data in the drip cup, simultaneously judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio, and if so, carrying out secondary emergency warning.
In one embodiment, referring to fig. 2, there is provided a machine vision based intelligent infusion alarm system comprising:
the first acquisition module is used for acquiring an image of a patient transfusion container at a first transfusion time point;
the first identification module is used for identifying the infusion container image based on a first machine learning model so as to acquire volume data of the infusion container;
the second acquisition module is used for acquiring a drip cup image after acquiring volume data of the infusion container;
the second identification module is used for identifying the drip cup image based on a second machine learning model so as to obtain diameter data of the drip tube;
the third acquisition module is used for acquiring a drip cup video at a second infusion time point after acquiring diameter data of the drip tube;
the first calculation processing module is used for performing video processing on the drip cup video based on a preset time interval so as to calculate first dripping speed data in the drip cup;
the second calculation processing module is used for calculating the time difference value between the second infusion time point and the first infusion time point and calculating and acquiring error data;
the third calculation processing module is used for subtracting error data from the volume data of the infusion container to obtain the current infusion volume data and calculating the residual infusion time;
the infusion early warning module is used for judging whether the residual infusion time meets a preset early warning time range in real time according to the residual infusion time, if so, performing early warning once, and displaying the residual infusion time at the same time;
and the emergency alarm module is used for secondarily acquiring the drip cup video in real time when the residual transfusion time meets a preset early warning time range, calculating second dripping speed data in the drip cup, simultaneously judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio, and if so, carrying out secondary emergency alarm.
The second calculation processing module comprises a dripping speed correction unit, and the dripping speed correction unit is used for correcting the first dripping speed data according to the diameter data, namely the first dripping speed data according to the amount of liquid drops required by 1 milliliter of dropper with different diameters, so that the corrected first dripping speed data can be obtained.
In one embodiment, the intelligent infusion alarm method and system based on machine vision proposed in the present application will be explained in further detail with reference to specific cases;
in this embodiment, assuming that a patient performs infusion in 18 hours 00 minutes and 15 seconds, a monitoring camera in an infusion scene collects an image of a patient infusion container at a first infusion time point (assuming that the first infusion time point is 18 hours 01 minutes and 00 seconds) and identifies the image of the infusion container based on a first machine learning model to obtain volume data of the infusion container;
in this embodiment, it is assumed that the infusion container image is identified as an infusion bottle type, and the volume data of the infusion container of the infusion bottle is 200 ml;
then, collecting images of the drip cup based on the monitoring camera, and identifying based on a second machine learning model to obtain diameter data of the drip tube;
in the present embodiment, it is assumed that the diameter data of the dropper is 0.5 cm and corresponds to 0.5 ml per drop, that is, 1 ml per two drops on average;
to be noted here is: the diameter data of the dropper and the drop volume corresponding to each drop are not necessarily the same as those of the prior infusion device, and are only used for illustration;
after diameter data of a dropper are obtained, a dropper video at a second infusion time point (assuming that the first infusion time point is 18 hours, 05 minutes and 00 seconds) is collected and is subjected to video processing based on a preset time interval so as to calculate first dropping speed data in the dropper;
in the present embodiment, it is assumed that the first dropping speed data in the drip chamber within 1 minute is 30 drops/minute;
then, the first dripping speed data is corrected according to the diameter data, namely the dripping amount required by 1 milliliter of dropper with different diameters, namely the corrected first dripping speed data is obtainedThe first drop rate data of (a) is 15 drops/min; then, subtracting the first infusion time point from the second infusion time point to obtain a time difference value; finally, by formula SW=VX×tcThe corrected first dripping speed data and the time difference value are multiplied to obtain error data of 60 milliliters;
then, subtracting error data from the volume data of the infusion container, namely 200-60=140 ml;
then, according to the formula
Figure BDA0003736814350000131
Calculating the remaining transfusion time as 140 ÷ 15 ≈ 9.3 minutes;
finally, judging whether the residual transfusion time meets a preset early warning time range in real time, and if the preset early warning time range is 1 minute, if the residual transfusion time is less than 1 minute, the system can give an early warning;
meanwhile, when the residual transfusion time meets a preset early warning time range, the monitoring camera can secondarily acquire the drip cup video in real time, calculate second dripping speed data in the drip cup, and judge whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio or not, if so, judge whether the ratio exceeds the preset ratio or not;
in this embodiment, assuming that the second droplet speed data is 25 droplets/minute, the ratio of the second droplet speed data to the first droplet speed data is 25: 30, i.e. 5: 6;
in this embodiment, assuming that the preset ratio is 1, when the second dropping speed data is 25 drops/min, the second dropping speed data is less than the preset ratio, so that the second emergency alarm is performed.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.

Claims (9)

1. An intelligent infusion alarm method based on machine vision is characterized by comprising the following steps:
1) Acquiring a patient transfusion container image at a first transfusion time point based on a monitoring camera in the existing transfusion scene, and identifying the transfusion container image based on a first machine learning model to acquire volume data of a transfusion container;
2) After volume data of the infusion container are obtained, collecting images of the drip cup based on the monitoring camera, and identifying based on a second machine learning model to obtain diameter data of the drip tube;
3) After diameter data of a dropper are acquired, a dropper video at a second infusion time point is acquired by using the monitoring camera, and the dropper video is subjected to video processing based on a preset time interval to calculate first dropping speed data in the dropper, wherein the preset time interval is 1 minute;
4) Correcting the first dripping speed data based on the diameter data, calculating a time difference value between the second transfusion time point and the first transfusion time point, and simultaneously calculating the time difference value and the corrected first dripping speed data to obtain error data;
5) Subtracting error data from the volume data of the infusion container to obtain current infusion volume data, and calculating the remaining infusion time according to the corrected first dripping speed data;
6) Judging whether the residual transfusion time meets a preset early warning time range in real time, and if so, performing early warning for the first time;
7) And when the residual transfusion time meets a preset early warning time range, secondarily collecting the drip cup video in real time by using the monitoring camera, calculating second dripping speed data in the drip cup, simultaneously judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio, and if so, carrying out secondary emergency warning.
2. The intelligent infusion alarm method based on machine vision as claimed in claim 1, wherein the first machine learning model and the second machine learning model are generated based on labeled basic data in an infusion device database as a training set, and the basic data in the infusion device database at least comprises volume data of infusion containers of different types and diameter data of droppers of different types.
3. The intelligent infusion alarm method based on machine vision is characterized in that the first machine learning model and the second machine learning model are at least one of logistic regression, random forest, K nearest neighbor, support vector machine, linear discriminant analysis, naive Bayes and neural network.
4. The intelligent infusion alarm method based on machine vision according to claim 1, characterized in that the video processing is performed on the infusion alarm method based on the preset time interval to calculate the first dropping speed data in the drip chamber, and the method comprises the following steps:
carrying out time line distinguishing on the collected drip cup videos, and extracting 1-minute videos;
and performing key frame marking on the 1-minute video, namely taking the first time that the infusion drop falls on the liquid level as a first key frame, taking the second time that the infusion drop falls on the liquid level as a second key frame, sequentially circulating until all key frames in 1 minute are obtained, and obtaining first dripping speed data in the drip cup according to the key frames.
5. The machine-vision-based intelligent infusion alarm method as claimed in claim 1, wherein the error data is infusion volume data input into the human body in the time difference between the second infusion time point and the first infusion time point, i.e. the infusion volume completed in the time period from the time point when the patient starts to input the infusion volume data to the time point after the data processing is completed.
6. The method as claimed in claim 1, wherein the step of correcting the first infusion speed data based on the diameter data and calculating the time difference between the second infusion time point and the first infusion time point and simultaneously calculating the time difference with the corrected first infusion speed data to obtain error data comprises:
firstly, correcting the first dripping speed data according to the diameter data, namely the dripping amount required by 1 milliliter of dropper with different diameters, so as to obtain the corrected first dripping speed data;
then, subtracting the first infusion time point from the second infusion time point to obtain a time difference value;
and finally, multiplying the corrected first drop speed data by the time difference to obtain error data, wherein the formula is as follows:
SW=VX×tc
in the formula: sWRepresenting error data; vXRepresenting the corrected first drop velocity data; t is tcRepresenting the time difference.
7. The intelligent infusion alarm method based on machine vision according to claim 1, characterized in that the remaining infusion time is calculated according to the following formula:
Figure FDA0003736814340000031
in the formula: t is a unit ofZIndicating the remaining infusion time; sZRepresenting infusion container volume data; s. theWRepresenting error data; vXIndicating the corrected first drop velocity data.
8. An intelligent infusion alarm system based on machine vision, comprising:
the first acquisition module is used for acquiring an image of a patient transfusion container at a first transfusion time point;
the first identification module is used for identifying the infusion container image based on a first machine learning model so as to acquire volume data of the infusion container;
the second acquisition module is used for acquiring a drip bottle image after acquiring volume data of the infusion container;
the second identification module is used for identifying the drip cup image based on a second machine learning model so as to obtain diameter data of the drip tube;
the third acquisition module is used for acquiring a drip cup video at a second infusion time point after acquiring diameter data of the drip tube;
the first calculation processing module is used for performing video processing on the drip cup video based on a preset time interval so as to calculate first dripping speed data in the drip cup;
the second calculation processing module is used for calculating the time difference value between the second infusion time point and the first infusion time point and calculating and acquiring error data;
the third calculation processing module is used for subtracting error data from the volume data of the infusion container to obtain the current infusion volume data and calculating the residual infusion time;
the infusion early warning module is used for judging whether the residual infusion time meets a preset early warning time range in real time according to the residual infusion time, if so, performing early warning once, and displaying the residual infusion time at the same time;
and the emergency alarm module is used for secondarily acquiring the drip cup video in real time when the residual transfusion time meets a preset early warning time range, calculating second dripping speed data in the drip cup, judging whether the ratio of the second dripping speed data to the first dripping speed data exceeds a preset ratio or not, and performing secondary emergency alarm if the ratio exceeds the preset ratio.
9. The intelligent infusion alarm system based on machine vision as claimed in claim 8, wherein the second calculation processing module comprises a drop speed correction unit, and the drop speed correction unit is configured to correct the first drop speed data according to the diameter data, that is, according to the amount of drops required by 1 ml of dropper with different diameters, so as to obtain the corrected first drop speed data.
CN202210805180.5A 2022-07-08 2022-07-08 Intelligent infusion alarm method and system based on machine vision Pending CN115252962A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210805180.5A CN115252962A (en) 2022-07-08 2022-07-08 Intelligent infusion alarm method and system based on machine vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210805180.5A CN115252962A (en) 2022-07-08 2022-07-08 Intelligent infusion alarm method and system based on machine vision

Publications (1)

Publication Number Publication Date
CN115252962A true CN115252962A (en) 2022-11-01

Family

ID=83765513

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210805180.5A Pending CN115252962A (en) 2022-07-08 2022-07-08 Intelligent infusion alarm method and system based on machine vision

Country Status (1)

Country Link
CN (1) CN115252962A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116421824A (en) * 2023-04-11 2023-07-14 天津医科大学总医院 Infusion monitoring method and system
CN117990938A (en) * 2024-04-03 2024-05-07 西安交通大学 Liquid dropping speed and volume measuring method and system

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116421824A (en) * 2023-04-11 2023-07-14 天津医科大学总医院 Infusion monitoring method and system
CN116421824B (en) * 2023-04-11 2024-01-19 天津医科大学总医院 Infusion monitoring method and system
CN117990938A (en) * 2024-04-03 2024-05-07 西安交通大学 Liquid dropping speed and volume measuring method and system

Similar Documents

Publication Publication Date Title
CN115252962A (en) Intelligent infusion alarm method and system based on machine vision
WO2020135226A1 (en) Drainage monitoring system and method based on image recognition
CN107315904B (en) Patient medicine taking remote monitoring system and method
CN109464722A (en) A kind of active pressure formula Intelligent transfusion apparatus, transfusion control system and control method
CN106362234A (en) Intravenous infusion management device and intravenous infusion management system
WO2020029814A1 (en) Doctor health monitoring system based on medical apparatus
CN105664297A (en) Infusion monitoring method, system and device
CN109481785A (en) A kind of dropping liquid monitoring system for residual and its control method
CN104225721A (en) Medical non-contacting drop monitoring system
CN114209925A (en) Intelligent non-gravity closed-loop closed infusion management system
CN105498042B (en) A kind of non-light-proof automatic alarm for infusion method and device thereof based on video
CN112642022B (en) Infusion monitoring system and monitoring method
CN201572380U (en) Transfusion reminding apparatus
CN215194492U (en) Intelligent infusion system
CN206660233U (en) A kind of infusion monitoring system
CN104524665A (en) Medical infusion monitoring device
CN112121258A (en) Infusion monitoring device and infusion monitoring method
CN216855425U (en) Intelligent non-gravity closed loop closed type transfusion management system
CN115732067A (en) Bedside infusion entity checking system and method based on computer vision
CN113663167B (en) Intelligent transfusion alarm device based on multiple sensors and monitoring method
CN112915316A (en) Infusion control system
CN114588402B (en) Infusion process monitoring system and method based on visual identification
CN210277862U (en) Full-automatic intravenous drip system
CN202554640U (en) Medical infusion dropping speed control system
CN117323499A (en) Infusion speed alarm and infusion speed monitoring alarm system thereof

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination