TWI798002B - Oral eating ability assessment system - Google Patents

Abstract

An evaluation system for oral feeding ability includes a sucking piece, a sensing device and an analyzing device. The sensing device includes a respiration sensor for sensing respiration and outputting a respiration sensing signal, a heart rate sensor for sensing heart rate and outputting a heart rate sensing signal, and a respiration sensor for sensing oral cavity A motion sensor that operates and outputs a motion sensing signal. The analyzing device is communicatively connected to the sensing device and is used to receive the breathing sensing signal, the heart rate sensing signal and the motion sensing signal, and output an analysis result according to a smart oral eating ability evaluation module. The sensing device senses respiration, heart rate and oral movement, and judges oral feeding ability based on the analysis result of the analyzing device. The sensing and analyzing process is objective and the evaluation accuracy can be improved.

Description

Oral eating ability assessment system

The invention relates to an evaluation system, in particular to an evaluation system for oral feeding ability.

Patients with central nervous system damage or infants with immature nerve development due to premature birth may lack the ability to coordinate neuromuscular movements between eating, swallowing, and breathing in the oral cavity. Pneumonia and suffocation problems.

In order to judge and evaluate the restoration or development of the patient's oral feeding ability, the existing evaluation method is mainly to observe and score by medical personnel with the oral movement development scale. However, manual observation will be affected by subjective judgment and clinical experience, the evaluation results are not objective, and the evaluation accuracy is lacking, which needs to be improved.

Therefore, the object of the present invention is to provide an oral feeding ability evaluation system that can improve the evaluation accuracy.

Therefore, the oral feeding ability evaluation system of the present invention includes a sucking member, a sensing device and an analyzing device. The suction piece is adapted to be placed in the mouth. The sensing device includes a respiration sensor arranged on the sucking part for sensing respiration and outputting a respiration sensing signal, a respiration sensor suitable for being arranged on the chest for sensing heart rate and outputting a heart rate sensing signal A heart rate sensor, and a motion sensor arranged on the sucking part and used to sense oral motion and output a motion sensing signal. The analyzing device is communicatively connected to the sensing device and has a built-in intelligent mouth-feeding ability evaluation module. The analysis device is used to receive the breathing sensing signal, the heart rate sensing signal and the motion sensing signal, and output an analysis result according to the intelligent oral eating ability evaluation module.

The effect of the present invention is: the sensing device senses respiration, heart rate and oral movement, and judges oral feeding ability based on the analysis result of the analysis device. The sensing and analysis process is objective and the evaluation accuracy can be improved.

Referring to FIG. 1 and FIG. 2 , an embodiment of the oral feeding ability evaluation system of the present invention is suitable for evaluating the oral feeding ability of a subject 9 . This embodiment includes a suction piece 1 , a sensing device 2 and an analysis device 3 .

The sucking part 1 includes a sucking part 11 adapted to be put into the oral cavity of the subject 9, and a stop part 12 connected to the sucking part 11 and exposed to the oral cavity. The stop portion 12 is used to stop outside the oral cavity, preventing the subject 9 from accidentally swallowing the sucking part 1 . In this embodiment, the sucking part 1 is made of silicone, but in other variations, the sucking part 1 can also be made of silicone or other flexible and biocompatible materials.

The sensing device 2 is disposed on the sucking part 1 and includes a breathing sensor 21 , a heart rate sensor 22 , a motion sensor 23 and a transmission part 24 .

The breath sensor 21 is disposed on the stop portion 12 of the suction member 1 and is used for sensing the breath of the subject 9 and outputting a breath sensing signal. In this embodiment, the breath sensor 21 is a thermometer for sensing breath temperature.

The heart rate sensor 22 is arranged on the chest of the subject 9 for sensing heart rate and outputting a heart rate sensing signal.

The motion sensor 23 is disposed on the suction portion 11 of the suction member 1 and is used for sensing oral motion and outputting a motion sensing signal. In this embodiment, the motion sensor 23 is a three-axis accelerometer.

The transmission part 24 is electrically connected to the breathing sensor 21, the heart rate sensor 22 and the motion sensor 23 and is communicatively connected to the analysis device 3, and is used for the sniff sensing signal, the heart rate sensing signal and the The motion sensing signal is sent to the analyzing device 3 .

In this embodiment, the transmission element 24 is a miniature single-board computer (Single-Board Computers, SBCs), which is connected to the analysis device 3 by wired communication.

The analysis device 3 is communicatively connected to the transmission part 24 of the sensing device 2 and has a built-in intelligent oral feeding ability evaluation module. The analysis device 3 is used to receive the breathing sensing signal, the heart rate sensing signal and the motion sensing signal, and calculate and output an analysis result by the intelligent oral eating ability evaluation module.

In this embodiment, the intelligent oral feeding ability evaluation module built in the analyzing device 3 is trained to complete an artificial neural network (Artificial Neural Network) by means of deep learning.

Among them, deep learning is one of machine learning (Machine Learning), and machine learning is also divided into other types such as Supervised Learning, Unsupervised Learning and Reinforcement Learning. Deep learning is an algorithm based on representation learning (Representation Learning) of data. Its observation value (for example, an image) can be represented by a variety of feature methods, such as a vector of each pixel intensity value, or more abstractly. into a series of edges, regions of a specific shape, etc. The advantage of deep learning is that it replaces manual feature acquisition with unsupervised or semi-supervised feature learning and efficient algorithms for hierarchical feature extraction.

The intelligent oral feeding ability evaluation module is trained by inputting a plurality of training data, and each training data includes an input data and an output data. The input data includes values of the respiration sensing signal, the heart rate sensing signal and the motion sensing signal obtained by the respiration sensor 21 , the heart rate sensor 22 and the motion sensor 23 .

The present invention builds a decision-making assistance system through deep learning of artificial intelligence to evaluate the sensory and action responses in the oral cavity and builds:

1. Development of a diagnostic model: by collecting the breathing sensations obtained by the breathing sensor 21, the heart rate sensor 22 and the motion sensor 23 of babies of different gestational weeks (28-40 weeks of gestation) The values of the heart rate sensing signal, the heart rate sensing signal and the motion sensing signal are constructed and used for diagnosing developmental delay conditions.

2. Model of sucking characteristics: According to the development of oral movements of premature infants, distinguish the following three characteristics of sucking patterns: normal sucking pattern, disorganized sucking pattern and dysfunctional sucking pattern.

The normal sucking pattern is that newborns can suck continuously and regularly for 10 to 30 times, and present a regular rhythm of sucking-swallowing-breathing rhythm. The respiration, heart rate and oral movement are sensed by the respiration sensor 21, the heart rate sensor 22 and the motion sensor 23, and then the intelligent oral eating ability evaluation module analyzes the respiration sensing signal, The continuity, rhythm and regularity of the heart rate sensing signal and the motion sensing signal.

Disorganized sucking pattern is the newborn's inability to sustain suck-swallow-breath or movement without regular rhythm.

Dysfunctional sucking patterns are neonatal motor responses with abnormal or absent jaw and tongue movements.

The correct judgment answer of the output data is supplemented by medical experts collecting the breathing sensing signal, the heart rate sensing signal and the motion sensing signal and matching the corrected age of the baby (corrected age: the age calculated from the day of the expected date of delivery) and clinical oral feeding Assessment scale (Neonatal Oral Motor Assessment Scale, NOMAS) assessment and judgment.

The training method of the intelligent oral eating ability evaluation module is to calculate the error value between the output result obtained after inputting the input data into the artificial neural network and the correct judgment answer of the output data, using the gradient descent method (Gradient Descent) adjusts the connection weight (weight) and the bias value (bias) in the neuron between the middle layer and the layer neuron (neuron) of the artificial neural network, so as to reduce the difference between the output result and the correct decision answer. .

The evaluation process used to evaluate the oral ability of the subject 9 in this embodiment is described as follows:

First, put the sucking part 11 of the sucking piece 1 into the oral cavity of the subject 9, and let the subject 9 breathe, suck the sucking part 11 and swallow by himself. Simultaneously, the respiration sensor 21 senses the respiration temperature of the subject 9 and outputs the respiration sensing signal, the heart rate sensor 22 senses the heart rate of the subject 9 and outputs the respiration sensing signal, and The motion sensor 23 senses the oral motion of the subject 9 and outputs the motion sensing signal. The transmission member 24 receives the breathing sensing signal, the breathing sensing signal and the motion sensing signal and transmits them to the analyzing device 3 . Finally, the analysis device 3 inputs the breathing sensing signal, the breathing sensing signal and the motion sensing signal into the intelligent oral feeding ability evaluation module and outputs the analysis result as: normal, disordered or dysfunctional sucking patterns .

In the present invention, the sensing device 2 senses respiration, heart rate and oral movements, and judges oral feeding ability based on the analysis results of the analysis device 3 . The sensing and analysis process is objective and can improve evaluation accuracy. Therefore, the object of the present invention can indeed be achieved.

But what is described above is only an embodiment of the present invention, and should not limit the scope of the present invention. All simple equivalent changes and modifications made according to the patent scope of the present invention and the content of the patent specification are still within the scope of the present invention. Within the scope covered by the patent of the present invention.

1: sucking piece 11: sucking part 12: stop part 2: Sensing device 21: Breathing sensor 22: Heart rate sensor 23: Motion sensor 24: Transmission parts 3: Analysis device 9: Subject

Other features and effects of the present invention will be clearly presented in the implementation manner with reference to the drawings, wherein: Fig. 1 is a schematic perspective view illustrating an embodiment of the oral feeding ability evaluation system of the present invention used to evaluate a subject's oral feeding ability; and Fig. 2 is a circuit diagram of this embodiment.

1: sucking piece

11: sucking part

12: stop part

2: Sensing device

21: Breathing sensor

22: Heart rate sensor

23: Motion sensor

24: Transmission parts

3: Analysis device

9: Subject

Claims (6)

An oral feeding ability evaluation system, comprising: a suction piece, suitable for insertion into the oral cavity; A sensing device, including a respiration sensor arranged on the sucking part for sensing respiration and outputting a respiration sensing signal, a respiration sensor suitable for being arranged on the chest for sensing heart rate and outputting a heart rate sensing signal a heart rate sensor, and a motion sensor disposed on the sucking part for sensing oral motion and outputting a motion sensing signal; and An analysis device, which is connected to the sensing device in communication and has a built-in intelligent oral feeding ability evaluation module, the analysis device is used to receive the breathing sensing signal, the heart rate sensing signal and the motion sensing signal, and according to The intelligent oral eating ability evaluation module outputs an analysis result. The oral feeding ability evaluation system according to claim 1, wherein the sucking part includes a sucking part suitable for putting into the oral cavity, and a stop part connected to the sucking part and exposed to the oral cavity, and the breathing sensor set on the stopper. The oral feeding ability assessment system according to claim 1, wherein the breathing sensor is a thermometer for sensing breathing temperature. The oral feeding ability evaluation system as claimed in Claim 1, wherein the motion sensor is a three-axis accelerometer. The oral feeding ability assessment system according to claim 1, wherein the sensing device further includes a device that is electrically connected to the breathing sensor, the heart rate sensor and the motion sensor and is used to communicate with the analysis device The transmission part is a miniature single-board computer. The oral feeding ability evaluation system according to claim 1, wherein the intelligent oral feeding ability evaluation module built in the analysis device is trained by machine learning.

Images