CN111420177A

CN111420177A - Venous transfusion extravasation detection alarm system and detection method

Info

Publication number: CN111420177A
Application number: CN202010338838.7A
Authority: CN
Inventors: 张广渊; 王康; 高晓楠; 赵峰; 王朋; 谢振华
Original assignee: Shandong Jiaotong University
Current assignee: Shandong Jiaotong University
Priority date: 2020-04-26
Filing date: 2020-04-26
Publication date: 2020-07-17

Abstract

The invention discloses a venous transfusion extravasation detection alarm system which comprises a shell, wherein an industrial camera, a line laser generator and a transfusion blocking device are arranged at the front end of the shell, a processor and a laser generator moving device driving the line laser generator to move are arranged in the shell, the industrial camera and the line laser generator are connected with the processor through wires, and a computer vision method is adopted for monitoring, so that the result has higher accuracy and sensitivity.

Description

Venous transfusion extravasation detection alarm system and detection method

Technical Field

The invention relates to the technical field of medical instruments, in particular to a venous transfusion extravasation detection alarm system and a detection method.

Background

Infusion extravasation refers to the leakage of an infused medical fluid into soft tissue outside a vein for a variety of reasons during an infusion process. It is usually manifested as swelling, distending pain, moderate or severe pain, burning, stabbing pain, local red swelling, no blood return, dark purple skin and hard skin. Once extravasation of intravenous fluids occurs, serious consequences can occur if aggressive and correct preventive and care measures are not taken.

Infusion extravasation is a problem often encountered in clinical work, resulting in a variety of causes, including: the patient touches the needle head accidentally or releases the hand to take the object, so that the needle head slips off to cause extravasation; the infant cry and cry, is not matched, and has small and small blood vessels which are not exposed, so that the difficulty is increased for venipuncture, the infant moves and the transfusion part is difficult to fix in the transfusion process, the needle head is easy to slip off and the veins are easy to damage, and meanwhile, the exosmosis is easier to aggravate because the infant has weak expression ability; the old people can cause the displacement of a needle or the damage of veins due to the weakened behavior control capability, loose skin and fragile veins, and the carelessness can cause the slow reaction and the reduced pain threshold of the old people and also is an important reason for the aggravation of the extravasation; patients with disturbance of consciousness are prone to extravasation due to coma, mania, and sensory disturbance; the infusion concentration is too high, the infusion speed is too fast, the medicine irritation is too large, such as mannitol, calcium agent and the like, the blood vessel wall can be damaged, the permeability is increased, and the extravasation is caused; cancer is a risk factor for extravasation because cancer patients receive chemotherapy repeatedly, veins are fragile and difficult to puncture; the diabetic is easy to cause extravasation due to sugar and fat metabolism disorder and angiosclerosis; patients with elevated venous pressure, such as right heart failure patients, systemic venous stasis, and blood backflow are hindered, and extravasation is easy to occur.

The existing venous transfusion extravasation alarm judges whether extravasation exists or not by sensing the local skin uplift degree, but is not suitable for the conditions of too small extravasation amount, loose skin and the like; some vascular extravasation monitoring technologies are used for sensing the change of the biological impedance when the cannula needle is extravasated, but complex and expensive equipment needs to be connected for processing and analysis when the cannula needle is used, so that the method has no popularization value; there is also a portable alarm device which monitors the skin surface bio-impedance through a skin surface bio-impedance electrode to thereby sense extravasation of various degrees, but when the alarm device is fastened too tightly or pressed by an external force, it is easy to cause the extravasation of intravenous infusion, and it may hinder the observation of the puncture point. The existing equipment is contact equipment, needs a patient to wear the equipment, inevitably does not cause discomfort of the patient, and increases the contradiction between doctors and patients.

Disclosure of Invention

The invention aims to solve the problems and provides a venous transfusion extravasation detection alarm system and a detection method, which can effectively simplify the working content of nursing staff and reduce the tissue damage of a patient caused by transfusion extravasation.

The technical scheme adopted by the invention for solving the technical problems is as follows:

the utility model provides a venous transfusion extravasation detection alarm system, includes the shell, the shell front end is equipped with industry camera, line laser generator and infusion blocking device, is equipped with the laser generator mobile device that treater and drive line laser generator removed in the shell, industry camera, line laser generator pass through the wire and are connected with the treater.

Furthermore, a display screen is arranged at the upper end of the shell;

and a switch, a power supply connector and a buzzer are arranged on the side surface of the shell.

Further, the industrial camera and the line laser generator are placed in parallel.

Further, line laser generator includes casing, point laser generator and line laser conversion head, and point laser generator, line laser conversion head are located the casing, point laser generator passes through the wire and is connected with the treater, point laser generator is located line laser conversion head front portion.

Furthermore, the laser generator moving device comprises an upright post and a cross beam, the upright post is fixed on the shell, a sliding plate is arranged between the cross beam and the upright post, the sliding plate is connected with the upright post in a sliding manner, and the cross beam is connected with the sliding plate in a sliding manner;

the utility model discloses a line laser beam machine, including stand, first motor, synchronous belt drive slide, second motor, gear and rack, the stand upper end is equipped with first motor, and first motor reciprocates through synchronous belt drive slide, be equipped with the second motor on the slide, be equipped with the rack on the crossbeam, and second motor output is equipped with the gear, and the gear and rack cooperation reciprocates through first motor drive slide, moves about through second motor drive crossbeam, and line laser generator installs on the crossbeam.

Further, the infusion blocking device comprises a push rod, a third motor and a fixer for fixing the infusion tube flow regulator, the third motor is installed on the shell and is connected with the push rod through a roller lead screw, the fixer is fixed on the shell, a pulley is arranged at the upper end of the infusion tube flow regulator, and the push rod is matched with the pulley.

A venous transfusion extravasation detection method utilizes the venous transfusion extravasation detection alarm system and comprises the following steps;

(1) starting the device, and waiting for the system initialization;

(2) when medical staff or machines puncture, the industrial camera continuously collects images and transmits the images to the ARM processor;

(3) a neural network module in the ARM processor identifies the injection needle in real time and calculates the position of the injection needle;

the neural network module utilizes the plurality of convolution layers to efficiently extract image features, learns the feature representation of the injection needle from a large number of actual injection needle samples, identifies the feature representation and returns the identified needle position;

(4) when the needle head slowly advances to disappear, the puncture is considered to be finished, and the position of the injection needle head before the disappearance is recorded;

(5) the medical staff fixes the infusion tube flow regulator into the fixer of the infusion blocking system;

(6) the ARM processor gives a corresponding movement instruction to a motor system of a linear laser generator movement system according to the obtained needle position and direction, and the motor rotates according to the instruction to drive a transmission mechanism to operate, so that the linear laser generator is aligned to the puncture point;

(7) starting a linear laser generator to irradiate the puncture point;

(8) the industrial camera collects images with wired laser and transmits the images to the ARM processor;

(9) for the obtained image frame, extracting the skeleton line type of the line laser by an image processing technology of an infusion extravasation monitoring module of an ARM processor;

(10) after obtaining the skeleton line of the line laser, monitoring the change of the skeleton line type in real time, and when detecting that the skeleton line type is changed violently, determining that infusion extravasation occurs, and giving an infusion extravasation signal by an ARM processor;

after obtaining the skeleton line of the line laser, storing the skeleton line, marking the skeleton line as line, and processing the obtained image frame by using the same algorithm, wherein the difference is that after obtaining the skeleton line, further clustering the skeleton line by using a K _ Means algorithm to obtain K cluster centers, respectively calculating the minimum distance between each center point and the line, summing the minimum distance and marking the minimum distance as DIST, accumulating the success times count when the DIST is smaller than THRESHO L D _ DIST preset in a program, then continuously processing the next frame image by using the same method, marking the line as an initial line type start _ line when the count is accumulated to a preset value, storing the frame gray image start _ gray and extracting characteristic start _ feature, calculating the curvature of each point of the start _ line and calculating the average value thereof, and marking as start _ curve, otherwise, resetting the line, and executing the step (9) again;

after obtaining an initial linear start _ line, extracting a framework line from a subsequent frame, and then judging infusion extravasation by combining two aspects, namely, on one hand, clustering K cluster centers by applying a K _ Means algorithm to the framework line, respectively calculating the distance between each center point and the start _ line, summing the distances and recording as loss1, on the other hand, calculating the curvature of each point on the framework line to obtain an average value, calculating the Euclidean distance between a current and the start _ current, recording as loss2, weighting loss1 and loss2 to obtain loss (w 1 loss1+ w2 loss2), comparing the loss with THRESHO L D _ L OSS preset in a program, if the loss exceeds the THRESHO 35L D _ L OSS preset in the program, considering that the infusion extravasation occurs, an ARM processor sends an infusion extravasation signal, and otherwise, repeating the step (10) on the next frame image;

the image curvature calculation formula is as follows:

wherein: u shape_x，U_y，U_xx，U_yy，U_xyRespectively, a partial derivative of the image U along the x direction, a partial derivative along the y direction, a second-order partial derivative along the x direction, a second-order partial derivative along the y direction and a second-order mixed partial derivative;

(11) after detecting the infusion exosmosis, the ARM processor transmits a corresponding instruction to the infusion blocking system, the control push rod pushes the infusion tube flow regulator to block the infusion, the voice prompt module is controlled to give out voice broadcast, and the buzzer gives out a buzzer alarm to wait for further processing of medical staff.

Further, the image processing method used in step (9) includes the steps of:

1) applying HSV color tracking to roughly separate the area where the line laser exists;

2) using morphological operations such as corrosion, expansion and the like to corrode and remove noise points and expand breakpoints possibly existing in connection;

3) extracting the outer contour line of the line laser by contour detection, filling all areas in the outer contour line, and eliminating holes;

4) thinning the filled line laser line area to preliminarily extract the line shape of the line structure light skeleton;

further, in the step 4), due to noise and the like, branch burrs inevitably exist on the thinned framework line, and in order to avoid the influence of the branch burrs on subsequent calculation, the branch burrs need to be trimmed, and the trimming step is as follows;

the structure tree building part:

and a, searching branch points in the size of a specified neighborhood from the leftmost point of the skeleton line as a center, and when the branch points are not searched in the neighborhood, expanding the neighborhood at the speed of 2 times to continue searching until the branch points are searched, and marking the branch points as root nodes. When the search neighborhood is as large as the whole graph and no branch point exists, the framework line is considered to have no branch, and the algorithm is exited;

b, searching in the branch direction by taking the root node as a starting point, establishing a new tree node when a branch point is encountered, taking the new tree node as a sub-node of the last branch point, and recording corresponding path information; establishing leaf nodes each time an end point is encountered, and stopping searching on the path;

c, after traversing all available pixel points, establishing a structure tree of the skeleton line;

and (3) branch cutting part:

and d, traversing the structure tree, starting from the leaf node, tracing the father node, traversing each child node at the father node, cutting off child nodes with the child path length smaller than a specified threshold value, selecting the child path with the longest length as a main direction, reserving the child path with the direction closest to the child path, and cutting off the rest child paths. Recalculating the type of the trimmed father node, and traversing the structure tree again;

and e, repeating d until all leaf nodes are traversed, and obtaining a smooth skeleton line.

Further, during the infusion, the movement of the piercing part may be caused by the patient's intentional or unintentional actions, at which point the previously obtained information will no longer be available, and the solving steps are as follows:

a, modeling an image by using a Gaussian mixture model, separating the foreground and the background of the image, setting the foreground as white and the background as black;

b, calculating a proportion of the foreground, foreground _ RATIO, and when the foreground _ RATIO is greater than a specified threshold L D _ RATIO, considering that the puncture part of the patient moves and does not perform skeleton extraction any more;

c, prompting the patient to recover the position through voice broadcasting;

d after the forcedly _ ratio is stable, considering that the patient recovers the position, converting the image frame into a gray image and extracting the feature of the gray image, performing feature matching on the feature and the start _ feature, obtaining new puncture area information by taking the high matching area as the moved puncture area, sending a corresponding instruction to the line laser generator control moving system by the ARM processor, moving the line laser generator to a new puncture point, and repeating the step (9).

The invention has the beneficial effects that:

1. the invention adopts a computer vision method for monitoring, so that the result has higher accuracy and sensitivity.

2. The invention adopts an ARM embedded system, is small and flexible, is convenient to move and install, and overcomes the defect that the traditional PC equipment is huge and is inconvenient to move.

3. The invention does not need to be in direct contact with the patient in the whole monitoring process, and does not increase any discomfort of the patient.

4. The invention displays the processing result on the display, which is beneficial for medical staff and patients to observe the injection condition.

5. The invention is not only suitable for the condition of venous transfusion extravasation, but also can be used in any situation requiring bulge detection by slight modification, has high expandability and good prospect, and has market popularization value.

Drawings

In order to more clearly illustrate the embodiments or technical solutions in the prior art of the present invention, the drawings used in the description of the embodiments or prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without creative efforts.

FIG. 1 is a front view of the present invention;

FIG. 2 is a left side view of the present invention;

FIG. 3 is a top view of the present invention;

FIG. 4 is a schematic view of a line laser generator according to the present invention;

FIG. 5 is a schematic view of an infusion blocking device in accordance with the present invention;

FIG. 6 is a schematic view of a moving device of the laser generator of the present invention

FIG. 7 is a flow chart of the detection method of the present invention.

In the figure: the device comprises a shell 1, an industrial camera 2, a line laser generator 3, a processor 4, a laser generator moving device 5, a conducting wire 6, an infusion blocking device 7, a display screen 8, a switch 9, a power connector 10, a buzzer 11, a shell 12, a point laser generator 13, a line laser conversion head 14, a point laser beam 15, a line laser beam 16 upright post 17, a cross beam 18, a sliding plate 19, a first motor 20, a synchronous belt 21, a second motor 22, a rack 23, a fixer 24, a push rod 25, a third motor 26, an infusion tube flow regulator 27 and a pulley 28.

Detailed Description

In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention, and it is obvious that the described embodiment is only a part of the embodiment of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1-3, a venous transfusion extravasation detection alarm system, including

shell

1, 1 front end of shell is equipped with industry camera 2, line laser generator 3 and infusion blocking device 7, is equipped with laser generator mobile device 5 that ARM treater 4 and drive line laser generator 3 removed in the shell 1, industry camera 2, line laser generator 3 pass through wire 6 and are connected with ARM treater 4, and industry camera 2 is used for gathering near puncture point regional image, and ARM treater 4 handles the analysis and sends various control signal to the image that industry camera 2 gathered. The ARM processor 4 comprises a neural network module, an infusion extravasation monitoring module and a voice prompt module. The neural network module identifies and tracks the needle head to obtain the position information of the puncture point; the infusion extravasation monitoring module processes images collected by the industrial camera to obtain the line type of the structured light skeleton, monitors the change of the line type of the structured light skeleton in real time and judges whether infusion extravasation occurs or not; the voice prompt module plays different prompt tones to the patient according to the conditions judged by the processor.

As shown in fig. 1 to 3, a display screen 8 is arranged at the upper end of the housing 1, the display screen 8 is used for displaying images acquired by the industrial camera 2 and processing results in real time, and is connected with the ARM processor 4 through a wire 6; the display screen 8 is placed at the top of the device and keeps a certain inclination angle with the device, so that medical staff and patients can observe conveniently; the side of the shell 1 is provided with a switch 9, a power connector 10 and a buzzer 11, the buzzer 11 is used for buzzing and alarming, the power connector 10 is used for power supply of the whole device and is designed to be an external socket plug, and the switch 9 is used for starting and stopping the whole device.

As shown in fig. 1, the industrial camera 2 and the line laser generator 3 are placed in parallel, and similar to the arrangement of human eyes, after the position of the injection needle under the camera coordinate system is obtained, the industrial camera 2 coordinate system can be conveniently mapped to the line laser generator 3 coordinate system, so that the line laser generator 3 can be conveniently moved in the next step.

As shown in fig. 4, the line laser generator 3 includes a housing 12, a point laser generator 13 and a line laser conversion head 14, the point laser generator 13 and the line laser conversion head 14 are located in the housing 12, the line laser generator 3 is used for irradiating the structural light near the puncture point, the point laser generator 13 is connected with the ARM processor 4 through a wire, the point laser generator 13 is started and stopped according to an instruction of the ARM processor 4, and the point laser generator 13 is located in front of the line laser conversion head 14 and is used for converting a point laser beam 15 into a line laser beam 16.

As shown in fig. 6, the laser generator moving device 5 includes an upright post 17 and a cross beam 18, the upright post 17 is fixed on the housing 1, a sliding plate 19 is arranged between the cross beam 18 and the upright post 17, the sliding plate 19 is slidably connected with the upright post 17 through a guide rail and slider pair, the sliding plate 19 moves in the vertical direction, the cross beam 18 is slidably connected with the sliding plate 19 through a guide rail and slider pair, and the cross beam 18 moves in the horizontal direction; the stand 17 upper end is equipped with first motor 20, and first motor 20 reciprocates through hold-in range 21 drive slide 19, be equipped with second motor 22 on the slide 19, be equipped with rack 23 on the crossbeam 18, second motor 22 output is equipped with the gear, wheel and rack 23 cooperation, reciprocate through first motor 20 drive slide 19, remove about second motor 22 drive crossbeam 18, line laser generator 3 installs on crossbeam 18, first motor 20 and second motor 22 pass through the wire and link to each other with ARM treater 4, accept the instruction that ARM treater 4 was linked to each other and was assigned, start driving motor and rotate, and then drive line laser generator 3 and carry out horizontal flat vertical motion.

As shown in fig. 5, the infusion blocking device 7 is used for blocking infusion after receiving an infusion extravasation signal, the infusion blocking device 7 includes a push rod 25, a third motor 26 and a fixer 24 for fixing an infusion tube flow regulator 27, the third motor 26 is installed on the housing 1, the third motor 26 is connected with the push rod 25 through a roller screw, the roller screw converts the number of turns of the motor into the displacement of the push rod 25 moving transversely, the fixer 24 is fixed on the housing 1, the fixer 24 is a clamp, a pulley 28 is arranged at the upper end of the infusion tube flow regulator 27, the push rod 25 is matched with the pulley 28, the third motor 26 is connected with the ARM processor 4 through a wire, the push rod 25 pushes the pulley 28 on the infusion tube flow regulator 27 to regulate the infusion speed under the driving of the third motor 26, and the infusion speed is completely blocked when the push rod is pushed to the bottom.

The infusion extravasation detection method adopts the following principle: when structured light irradiates the surface of an object, if the object is not moved, the whole structured light cannot be changed; otherwise, when the object moves, the structured light pattern on the object can change greatly. According to the principle, structured light (in the embodiment, line laser) is irradiated to the vicinity of the puncture, when infusion extravasation occurs, the line type of the structured light at the extravasation part is changed, and the occurrence of infusion extravasation is judged by detecting the change of the line type.

As shown in fig. 7, a method for detecting extravasation of intravenous infusion, which utilizes the system for detecting extravasation of intravenous infusion, comprises the following steps;

(1) starting the device, and waiting for the system initialization;

(7) starting a linear laser generator to irradiate the puncture point;

(9) for the obtained image frame, the skeleton line type of the line laser is extracted by an image processing technology of an infusion extravasation monitoring module of an ARM processor, and the image processing comprises the following steps:

in the step 4), because of noise and the like, branch burrs inevitably exist on the thinned framework line, and the framework line needs to be trimmed to avoid the influence of the branch burrs on subsequent calculation, wherein the trimming step is as follows, and a tree structure is adopted to record the framework line structure;

the structure tree building part:

and (3) branch cutting part:

the image curvature calculation formula is as follows:

During the infusion, the movement of the piercing part can be caused by the patient's intentional or unintentional actions, and the information obtained before is no longer available, and the solving steps are as follows:

c, prompting the patient to recover the position through voice broadcasting;

In the description of the present invention, it should be noted that the terms "left", "right", "upper", "lower", and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience in describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention.

In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; they may be mechanically or electrically connected, directly or indirectly through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.

Claims

1. The utility model provides a venous transfusion extravasation detection alarm system, includes shell (1), its characterized in that, shell (1) front end is equipped with industry camera (2), line laser generator (3) and infusion blocking device (7), is equipped with laser generator mobile device (5) that treater (4) and drive line laser generator (3) removed in shell (1), industry camera (2), line laser generator (3) are connected with treater (4) through wire (6).

2. The venous transfusion extravasation detection alarm system of claim 1, wherein a display screen (8) is arranged at the upper end of the shell (1);

the side surface of the shell (1) is provided with a switch (9), a power connector (10) and a buzzer (11), and the switch (9), the power connector (10) and the buzzer (11) are all connected with the processor (4) through leads.

3. The intravenous infusion extravasation detection alarm system of claim 1, wherein the industrial camera (2) and the line laser generator (3) are placed in parallel.

4. The venous transfusion extravasation detection alarm system of claim 1, wherein the line laser generator (3) comprises a shell (12), a point laser generator (13) and a line laser conversion head (14), the point laser generator (13) and the line laser conversion head (14) are positioned in the shell (12), the point laser generator (13) is connected with the processor (4) through a lead, and the point laser generator (13) is positioned in front of the line laser conversion head (14).

5. The venous transfusion extravasation detection alarm system of claim 1, wherein the laser generator moving device (5) comprises a vertical column (17) and a cross beam (18), the vertical column (17) is fixed on the shell (1), a sliding plate (19) is arranged between the cross beam (18) and the vertical column (17), the sliding plate (19) is connected with the vertical column (17) in a sliding way, and the cross beam (18) is connected with the sliding plate (19) in a sliding way;

the utility model discloses a line laser generator, including stand (17), first motor (20) drive slide (19) through hold-in range (21) and reciprocate, be equipped with second motor (22) on slide (19), be equipped with rack (23) on crossbeam (18), second motor (22) output is equipped with the gear, and wheel and rack (23) cooperation reciprocates through first motor (20) drive slide (19), moves about through second motor (22) drive crossbeam (18), and line laser generator (3) are installed on crossbeam (18).

6. The venous transfusion extravasation detection alarm system of claim 1, wherein the transfusion blocking device (7) comprises a push rod (25), a third motor (26) and a fixer (24) for fixing a transfusion tube flow regulator (27), the third motor (26) is installed on the shell (1), the third motor (26) is connected with the push rod (25) through a roller lead screw, the fixer (24) is fixed on the shell (1), a pulley (28) is arranged at the upper end of the transfusion tube flow regulator (27), and the push rod (25) is matched with the pulley (28).

7. The intravenous fluid extravasation detection method of claim 1, utilizing the intravenous fluid extravasation detection alarm system of any one of claims 1 to 6, comprising the steps of;

(1) starting the device, and waiting for the system initialization;

(7) starting a linear laser generator to irradiate the puncture point;

the image curvature calculation formula is as follows:

(10) after detecting the infusion exosmosis, the ARM processor transmits a corresponding instruction to the infusion blocking system, the control push rod pushes the infusion tube flow regulator to block the infusion, the voice prompt module is controlled to give out voice broadcast, and the buzzer gives out a buzzer alarm to wait for further processing of medical staff.

8. The intravenous infusion extravasation detection method of claim 7, wherein the image processing method used in step (9) comprises the steps of:

4) and thinning the filled line laser line area to preliminarily extract the line structure light skeleton line type.

9. The method for detecting extravasation of intravenous infusion according to claim 8, wherein in step 4), branch burrs inevitably exist on the thinned skeleton line due to noise and the like, and the branch burrs need to be trimmed to avoid the influence of the branch burrs on subsequent calculation, and the trimming step is as follows;

the structure tree building part:

and (3) branch cutting part:

10. The method of claim 7, wherein the step of detecting extravasation in an intravenous infusion, wherein the movement of the piercing portion during the infusion is caused by an intentional or unintentional action by the patient, wherein the previously obtained information is no longer available, comprises the steps of:

c, prompting the patient to recover the position through voice broadcasting;