WO2023150960A1 - 超声波侦测中耳积水之数组式测量与判读装置与方法 - Google Patents

超声波侦测中耳积水之数组式测量与判读装置与方法 Download PDF

Info

Publication number
WO2023150960A1
WO2023150960A1 PCT/CN2022/075810 CN2022075810W WO2023150960A1 WO 2023150960 A1 WO2023150960 A1 WO 2023150960A1 CN 2022075810 W CN2022075810 W CN 2022075810W WO 2023150960 A1 WO2023150960 A1 WO 2023150960A1
Authority
WO
WIPO (PCT)
Prior art keywords
middle ear
array
hydrops
ultrasonic
measurement
Prior art date
Application number
PCT/CN2022/075810
Other languages
English (en)
French (fr)
Inventor
陈锦国
崔博翔
Original Assignee
长庚医疗财团法人林口长庚纪念医院
长庚大学
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 长庚医疗财团法人林口长庚纪念医院, 长庚大学 filed Critical 长庚医疗财团法人林口长庚纪念医院
Priority to PCT/CN2022/075810 priority Critical patent/WO2023150960A1/zh
Priority to CN202280090598.0A priority patent/CN118714978A/zh
Publication of WO2023150960A1 publication Critical patent/WO2023150960A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves

Definitions

  • the present disclosure relates to an array measurement and interpretation device and method for ultrasonic detection of middle ear hydrops, in particular to an ultrasonic detection method based on a single-array element ultrasonic probe combined with the probability density function parameters of multiple array measurement regions.
  • Otitis media is a common disease in children. It is currently the most commonly used antibiotic disease. It is also one of the main reasons for children to undergo surgery. However, otitis media is complicated by a high probability of middle ear fluid, which is not noticed, and middle ear fluid If it is not treated, it will lead to many complications and is a potential killer of hearing loss. Therefore, the diagnosis of hydrocephalus in the middle ear has important clinical value.
  • otoscopy is the most basic and direct way to check for middle ear effusion. It mainly checks the appearance of the eardrum or observes the vibration of the eardrum by blowing air to determine whether there is effusion in the middle ear.
  • diagnosis requires a high degree of Relying on the experience of doctors can easily fall into subjective judgment.
  • a tympanogram is another method commonly used to assess middle ear fluid. It is similar to blowing air on the eardrum and measuring the movement of the eardrum in response to pressure changes, but requires the patient to cooperate with the sealing and compression of the ear canal. For the test, children are often afraid or crying, and it is difficult to communicate, which affects the tightness of the earplugs and the ear canal, and the test time is also difficult to grasp.
  • Medical imaging systems such as computer tomography and nuclear magnetic resonance imaging are currently the standard methods for detecting middle ear hydrops.
  • computer tomography has concerns about radiation, and the above two
  • the measurement time required by the method is relatively long, and the measurement environment is relatively airtight.
  • computed tomography and magnetic resonance imaging are also limited by the performance of the device.
  • ultrasound provides a non-invasive, radiation-free, suitable for any place and any time, and can be used to evaluate the interpretation of middle ear hydrops before and after surgery.
  • one object of the present disclosure is to provide an array measurement and interpretation device for ultrasonic detection of middle ear hydrops, including: an ultrasonic probe, which is attached to the mastoid surface behind a user's ear and sends an ultrasonic wave , and the surface of the mastoid is divided into a plurality of array-type measurement areas, and an ultrasonic echo signal is generated in each array-type measurement area according to the water accumulation condition of the user's middle ear; an ultrasonic transmitter receiver is connected with The ultrasonic probe is connected and receives the ultrasonic echo signals of each array type measurement area; an analog-to-digital converter is connected with the ultrasonic transmitting receiver, and converts the ultrasonic echo signals into a digital signal; and an analysis unit, which is connected with the analog-to-digital converter, and calculates a probability density function (probability density) parameter of each array measurement area according to these digital signals, and then uses a pre-trained model to find out the corresponding An importance weight for the probability density function parameters
  • the ultrasound probe is a low frequency delay probe.
  • the analysis unit is a personal computer.
  • Another object of the present disclosure is to provide an array measurement and interpretation method for ultrasonic detection of hydrops in the middle ear, which includes the following steps: attaching an ultrasonic probe to the mastoid surface behind a user's ear and sending an ultrasonic wave, Divide the mastoid surface into a plurality of array-type measurement areas, and generate an ultrasonic echo signal in each array-type measurement area according to the water accumulation condition of the user's middle ear, and use an ultrasonic transmitter receiver to receive each The ultrasonic echo signals of the array measurement area; the ultrasonic echo signals are converted into a corresponding digital signal with an analog-to-digital converter; and according to the digital signals, an analysis unit is used to calculate A probability density function parameter of the array-type measurement area, and use the pre-training model to find an importance weight corresponding to the probability density function parameters of each array-type measurement area, and then combine the probability density function parameters of each array-type measurement area with the The corresponding importance weights are multiplied and summed up to generate a
  • the importance weights of the probability density function parameters corresponding to each array-type measurement area are found by using a linear discriminant analysis (Linear discriminant analysis, LDA) to perform a pre-training model.
  • LDA Linear discriminant analysis
  • the surface of the mastoid is divided into 12 array-type measurement areas.
  • the method of array measurement and interpretation for ultrasonic detection of middle ear water is to improve the accuracy of judging the severity of the user's middle ear water.
  • the method of array measurement and interpretation for ultrasonic detection of middle ear hydrops is to improve the accuracy of judging the nature of the user's middle ear hydrops.
  • the middle ear diseases include various otitis media, middle ear hydrops, mastoid hydrops, mastoiditis, and follow-up before and after ear tube implantation.
  • the disclosed array measurement and interpretation device and method for ultrasonic detection of middle ear effusion uses a single-array ultrasonic probe to use the mastoid surface behind the subject's ear as an audio window to send ultrasonic signals, and then receive And analyze the ultrasonic echo signal returned according to the water condition in the middle ear of the subject to perform non-invasive measurement; wherein, due to the large area at the mastoid, and there is no relevant prior art disclosure that can reflect The detection position of the best measurement result. Therefore, if the measurement position cannot be accurately grasped, the result will be inaccurate due to the uncertainty of the measurement position.
  • the invasive measurement method and its device further divide the surface of the mastoid of the subject into multiple array-type measurement areas, and obtain the ultrasonic echo signals of each array-type measurement area, After calculating the parameters of the probability density function respectively, the data analysis software is used in the analysis unit, and the linear discriminant analysis based on machine learning is used for pre-training to find out the importance of each separated measurement area in judging the water status of the middle ear.
  • the new parameters can not only maintain a good ability to judge whether the patient has middle ear hydrops, but also can be used It can be used to judge the severity of middle ear hydrops, and can also effectively improve the ability to judge the fluid nature (ie, serous or mucous) of middle ear hydrops.
  • the disclosed array measurement and interpretation device and method for ultrasonic detection of middle ear hydrops can improve the existing shortcomings of the prior art, and has the advantages of non-invasiveness, no radiation, applicable to any place and time, etc., especially It is suitable for measuring the condition of middle ear hydrops in children, and if the user's middle ear hydrops are measured repeatedly before and after surgery using the present disclosure, it can be used to assess whether the middle ear fluid has cleared before and after surgery, and most importantly, The disclosed method of data analysis can also reduce the uncertainty caused by subjective judgment, make the diagnosis of middle ear hydrops faster, the operation more convenient, and the detection results more accurate.
  • FIG. 1A is a schematic diagram of components of an array measurement and interpretation device for ultrasonic detection of middle ear hydrops according to the present disclosure.
  • FIG. 1B is a schematic diagram for defining the user's mastoid surface into a plurality of array-type measurement areas.
  • FIG. 2A is a receiver operating characteristic curve for measuring and interpreting middle ear hydrops in normal people and middle ear hydrops patients using the method and device for array measurement and interpretation of middle ear hydrops by ultrasonic waves of the present disclosure.
  • FIG. 2B is a receiver operating characteristic curve for measuring and interpreting middle ear hydrops in normal people and middle ear hydrops patients using only a single probability density function parameter.
  • Fig. 3A is a method and device for the array measurement and interpretation of middle ear hydrops by ultrasonic detection of the present disclosure, measuring and interpreting the status of middle ear hydrops in patients with mild to moderate hydrops and severe middle ear hydrops receiver operating characteristic curve.
  • FIG. 3B is a receiver operating characteristic curve for measuring and interpreting the middle ear effusion status of patients with mild to moderate effusion and severe effusion using only a single probability density function parameter.
  • Fig. 4A is a diagram for measuring and interpreting the middle ear hydrops in patients with serous hydrops and mucous hydrops in the method and device for array measurement and interpretation of middle ear hydrops using ultrasonic waves of the present disclosure receiver operating characteristic curve.
  • FIG. 4B is a receiver operating characteristic curve for measuring and interpreting the middle ear hydrops in patients with serous hydrocephalus and mucous hydrocephalus using only a single probability density function parameter.
  • the operating procedures and parameter conditions of ultrasonic detection and reception including ultrasonic probes and ultrasonic transmitter receivers, fall within the professional accomplishment and routine technical scope of those who are familiar with this technology.
  • the array measurement and interpretation device 1 for ultrasonic detection of middle ear hydrops in the present disclosure includes: an ultrasonic wave Probe 11, an ultrasonic transmitter-receiver 12, an analog-to-digital converter 13, and an analysis unit 14; wherein, the ultrasonic probe 11 is connected to the ultrasonic transmitter-receiver 12, and the ultrasonic transmitter-receiver 12 is connected to the ultrasonic transmitter-receiver 12
  • the analog-to-digital converter 13 is connected, and the analog-to-digital converter 13 is connected to the analysis unit 14 .
  • the method of using the array type measurement and interpretation device 1 for ultrasonic detection of middle ear hydrops in the present disclosure is as follows: First, attach the ultrasonic probe 11 to the surface of the mastoid behind the ear of a user to provide sound to the breast as the sound window. The protruding surface sends an ultrasonic wave, and generates a corresponding ultrasonic echo signal for non-invasive measurement according to the user's middle ear water condition.
  • the results are distorted.
  • the user's mastoid surface will first be defined as a plurality of arrays.
  • the measurement area considering the size of the ultrasonic probe 11 and the size of the user's mastoid, those skilled in the art can adjust the number of array measurement areas, and as shown in Figure 1B, it is best to define 12 arrays type measurement area, and generate an ultrasonic echo signal respectively for each array type measurement area; These ultrasonic echo signals are respectively converted into a corresponding digital signal, and then these digital signals of each array measurement area are sent to the analysis unit 14 for further analysis and calculation of these digital signals.
  • a probability density function (probability density) parameter (or called Statistical dynamic difference values of different orders)
  • the analysis unit 14 will use machine learning-based linear discriminant analysis (Linear discriminant analysis, LDA) ( Fisher Linear Discriminant (or Fisher Linear Discriminant) is pre-trained to find out the importance weights corresponding to the probability density function parameters (statistical dynamic differences of different orders) of each array measurement area, and then each array
  • LDA Linear discriminant analysis
  • Fisher Linear Discriminant Fisher Linear Discriminant
  • the probability density function parameters of the formula measurement area are multiplied by the corresponding importance weights and then summed up to obtain a weighted probability density function parameter, and the probability density function parameters of each array-type measurement area are obtained through
  • the Nakagami parameter is quantified, so when the value of the probability density function parameter is higher, it means that the user's situation is closer to the condition of middle ear hydrops to be detected (including: whether the middle ear has hydrops, the degree of middle ear
  • the analysis unit 14 is a personal computer. And according to the present disclosure, the analysis unit 14 can be used with commonly available data analysis software in the market, such as MATLAB, and provide functions of calculating probability density functions and importance weights.
  • the ultrasonic probe 11 is a low-frequency delay probe, so the measurement signal and the excitation pulse can be separated to increase the accuracy of the measurement signal and probability density function analysis. Spend.
  • control group did not separate the surface of the mastoid into multiple array-type measurement areas, and tested the same ethnic group, that is, only a single probability density function parameter was used to detect middle ear hydrops. Condition measurement and interpretation.
  • the receiver operating characteristic curve (receiver operating characteristic curve, ROC, hereinafter referred to as ROC curve) was drawn for each experimental group with statistical software (for example: sigmaplot) for analysis, and compared between each experimental group and The area under the ROC curve (Area under the Curve of ROC, AUROC) of the control group is used to evaluate the accuracy of the method and device for the array measurement and interpretation of ultrasonic detection of middle ear hydrops in the present disclosure; wherein, the ROC curve is often used for analysis
  • the accuracy of a detection method which is a binary analysis model, that is, the output result has only two types of models, for example, in the experimental group (1) there is middle ear hydrops or no middle ear hydrops, and in the experimental group ( 2) for mild, moderate or severe middle ear hydrops, and in the experimental group (3) serous or mucous middle ear hydrops, and set a threshold to separate the two result categories, and then divide each experiment
  • the weighted probability density function parameters for example
  • Figure 3A and Figure 3B are the test results of the severity of middle ear hydrops in group (2), and Figure 3A is an embodiment of the present disclosure, using the weighted probability density function with the method and device for measurement and interpretation of the present disclosure
  • the AUROC obtained by ROC analysis is as high as 0.87
  • Figure 3B is the control group, only using a single probability density function parameter for mild to moderate effusion.
  • the AUROC obtained by ROC analysis is only 0.53, which shows that the measurement and interpretation method has no ability to predict the severity of middle ear hydrops.
  • the results show that the method and device for measuring and interpreting disclosed in the present disclosure can be used to judge the severity of middle ear hydrops in patients.
  • Figure 4A and Figure 4B are the test results of the fluid properties of middle ear hydrops in group (3).
  • Figure 4A is an embodiment of the present disclosure, using the method and device for measurement and interpretation of the present disclosure, using the weighted probability After the density function parameters are used to interpret the status of patients with serous hydrocephalus and mucinous hydrocephalus, the AUROC obtained by ROC analysis is effectively increased to 0.69, while Figure 4B is the control group, only using a single probability density function After interpreting the status of patients with serous hydrocephalus and mucinous hydrocephalus, the AUROC obtained by ROC analysis is only 0.55, which shows that the measurement and interpretation method can hardly predict the hydrops of the middle ear characteristic. This result shows that using the method and device for measurement and interpretation of the present disclosure can effectively improve the ability to judge the fluid nature (ie, serous or mucous) of the patient's middle ear hydrops.
  • the disclosed array measurement and interpretation device and method for ultrasonic detection of middle ear effusion uses a single-array ultrasonic probe to use the mastoid surface behind the subject's ear as an audio window to send ultrasonic signals, and then receive And analyze the ultrasonic echo signal returned according to the water condition in the middle ear of the subject to perform non-invasive measurement; wherein, due to the large area at the mastoid, and there is no relevant prior art disclosure that can reflect The detection position of the best measurement result. Therefore, if the measurement position cannot be accurately grasped, the result will be inaccurate due to the uncertainty of the measurement position.
  • the new parameters can not only maintain a good ability to judge whether the patient has middle ear hydrops, but also can be used It can be used to judge the severity of middle ear hydrops, and can also effectively improve the ability to judge the fluid nature (ie
  • the disclosed array measurement and interpretation device and method for ultrasonic detection of middle ear hydrops can improve the existing shortcomings of the prior art, and has the advantages of non-invasiveness, no radiation, applicable to any place and time, etc., especially It is suitable for measuring the condition of middle ear hydrops in children, and if the user's middle ear hydrops are measured repeatedly before and after surgery using the present disclosure, it can be used to assess whether the middle ear fluid has cleared before and after surgery, and most importantly, The method of data analysis disclosed in the present invention can also reduce the uncertainty caused by subjective judgment, make the diagnosis of middle ear hydrops more rapid, the operation is more convenient, and the detection result is more accurate.

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)
  • Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)

Abstract

一种超声波侦测中耳积水之数组式测量与判读的方法及装置，该装置包含一超声波探头（11）、一超声波发射接收器（12）、一模拟数字转换器（13）、及一分析单元（14），能够将乳突处表面分隔成复数个测量区，并于使用超声波进行中耳积水的非侵入式检测时，以线性判别分析进行预训练，找出最佳的侦测位置及其加权参数以得到准确的评估数值。

Description

超声波侦测中耳积水之数组式测量与判读装置与方法 技术领域
本公开是关于一种超声波侦测中耳积水之数组式测量与判读装置与方法，尤其是一种以单阵元超声波探头为基础，结合复数个数组测量区域的机率密度函数参数的超声波侦测中耳积水之数组式测量与判读装置与方法。
背景技术
中耳炎是一种常见于孩童的疾病，为目前最常使用抗生素的疾病，也是造成孩童需进行手术的主要原因之一，而中耳炎并发中耳积水的机率高且不是察觉，又中耳积水若不进行治疗会导致诸多并发症的发生，是造成听力损失的潜在杀手，因此诊断中耳积水在临床上有重要的价值。
在临床上，耳镜是中耳积水最基础且直接的检查方式，主要透过检视耳膜的外观或辅以吹气来观察耳膜的振动，以判别中耳是否有积液，但诊断需高度依赖医师的经验，容易落入主观判断。鼓室图则是另一种常用于评估中耳积水的方法，类似对耳膜吹气，并在期间测量耳膜对压力改变之历程中的运动，但需要患者配合对其耳道封密加压进行测试，常常因为孩童害怕或哭闹，且较难沟通，影响耳塞与耳道密合程度，施测时间亦较难掌握。
而计算机断层扫描(computer tomography)以及核磁共振成像(nuclear magnetic resonance imaging)等医学影像系统是目前主要侦测中耳积水状况的标准方法，然而计算机断层扫描有产生辐射的疑虑，且上述两种方法所需的测量时间较长，加上测量环境较为密闭，对于中耳炎病者，尤其是小孩来说，容易产生抗拒的反应；此外，计算机断层扫描以及核磁共振成像还受限于装置的体绩，更有使用时间以及地点的限制。相较于相关技术中的影像系统，超声波则提供一种非侵入式、无辐射、适用于任何地点与任何时间，且可用于评估手术前后中耳积水之判读。
目前美国专利公开第US7131946号以及US20100069752号已揭露使用超声波探头伸入耳道以侦测中耳积水，但此测量方法仍属于侵入性，尤其不适合用于侦测孩童的中耳积水；而中华人民共和国台湾地区专利第1549657号 揭露使用超声波探头条贴于用户耳朵后方对应乳突的位置，以作为音窗来进行非侵入式的量测，并依据超声波回音频号的统计特性，判断受检者的中耳积水状况，此种技术虽然克服先前超声波探头必须申入耳道才能评估中耳积水的困境，但是由于乳突区域大，在无法精准掌握测量位置的情况下，会因为测量位置的不确定性，造成侦测结果的失准。
综上所述，相较于相关技术中的影像系统，研发一种非侵入式、无辐射疑虑、适用于任何地点与任何时间，且可精准地评估中耳积水程度及状况的方法与判读装置，着实有其必要性。
发明内容
缘此，本公开之一目的在提供一种超声波侦测中耳积水状况之数组式测量与判读装置，包含：一超声波探头，是贴于一用户之耳朵后方的乳突表面并发送一超声波，且所述乳突表面分隔成复数个数组式测量区域，并依据所述用户的中耳积水状况，于各数组式测量区域分别产生一超声波回音频号；一超声波发射接收器，是与所述超声波探头连接，并接收各数组式测量区域的该些超声波回音频号；一模拟数字转换器，是与所述超声波发射接收器连接，并将该些超声波回音频号分别转换为一数字讯号；以及一分析单元，是与所述模拟数字转换器连接，且根据该些数字讯号计算出各数组式测量区域的一机率密度函数(probability density)参数，再使用预训练模型找出分别对应于各数组式测量区域之机率密度函数参数的一重要性权重，并将各数组式测量区域的机率密度函数参数与所对应的重要性权重相乘后加总，以产生一经加权的机率密度函数参数，用于量化所述用户之中耳积水程度，据以判读所述用户之中耳疾病的状况。
在本公开之一实施例中，所述超声波探头是为一低频率延迟探头。
在本公开又一实施例中，所述分析单元是为一个人计算机。
本公开之又一目的是提供一种超声波侦测中耳积水状况的数组式测量与判读方法，包含以下步骤：将一超声波探头贴于一用户之耳朵后方的乳突表面并发送一超声波，将所述乳突表面分隔成复数个数组式测量区域，并依据所述用户的中耳积水状况，于各数组式测量区域分别产生一超声波回音频号， 并以一超声波发射接收器接收各数组式测量区域的该些超声波回音频号；将该些超声波回音频号，以一模拟数为转换器分别转换为相应的一数字讯号；以及依据该些数字讯号，以一分析单元计算出各数组式测量区域的一机率密度函数参数，并使用预训练模型找出分别对应各数组式测量区域之机率密度函数参数的一重要性权重，再将各数组式测量区域的机率密度函数参数与所对应的重要性权重相乘后加总，以产生一经加权的机率密度函数参数，用于量化所述用户之中耳积水程度，据以判读所述用户之中耳疾病的状况。
在本公开之一实施例中，对应于各数组式测量区域之机率密度函数参数的重要性权重是使用线性判别分析(Linear discriminant analysis,LDA)进行预训练模型所找出。
在本公开之一实施例中，所述乳突表面分割成12个数组式测量区域。
在本公开之一实施例中，所述超声波侦测中耳积水之数组式测量与判读方法是提升判断所述用户中耳积水严重程度的准确性。
在本公开之一实施例中，所述超声波侦测中耳积水之数组式测量与判读方法是提升判断所述用户中耳积水之积液性质的准确性。
在本公开另一实施例中，所述中耳疾病包含各式性中耳炎、中耳积水、乳突积水、乳突炎以及耳管植入前后追踪。
因此，本公开超声波侦测中耳积水之数组式测量与判读装置及方法，藉由单阵元的超声波探头，以受测者耳朵后方的乳突表面作为音窗并发送超声波讯号，再接收并分析根据所述受测者之中耳积水状况所回传的超声波回音频号，以进行非侵入式的测量；其中，由于乳突处的区域大，且未有相关先前技术揭露可以反应最佳测量结果的侦测位置，因此在无法精准掌握测量位置的情况下，会因为测量位置的不确定性，造成结果失准。
据此，在本公开中，为了优化整体信息的运算及分析方式，并整合至原始操作接口中，以达到方便操作且快速分类中耳积水的状况与性质，来提高以超声波为基础之非侵入式量测方法及其装置的精准性及预期性，而进一步将受测者的乳突处表面分隔成复数个数组式的测量区域，并分别获取各数组式测量区域的超声波回音频号，再分别计算其机率密度函数参数后，于分析单元中搭配数据分析软件，并使用以机器学习为基础的线性判别分析进行预 训练，找出各分隔之测量区域于判断中耳积水状况的重要性权重，并将各测量区域的机率密度函数参数经加权后，以所得的新参数相较于使用单一参数而言，不仅可以维持良好之判断患者是否发生中耳积水的能力，更可以用于判断患者中耳积水的严重程度，且还能有效改善判断患者中耳积水之积液性质(即浆液性或黏液性)的能力。
综上所述，本公开超声波侦测中耳积水之数组式测量与判读装置及方法能够改善先前技术既有的缺点，具有非侵入性、无辐射、适用于任何地点与时间等优点，尤其适合用于测量孩童的中耳积水状况，且若使用本公开重复于手术前后测量用户之中耳积水状况，则可用于评估手术前后之中耳积水是否有清除，最重要的是，本公开数据分析的方式还能够减少因主观判定导致的不确定性，使中耳积水的诊断更加快速、操作更方便，且检测结果更加地准确。
以下将进一步说明本公开的实施方式，下述所列举的实施例是用以阐明本公开，并非用以限定本公开之范围，任何熟习此技艺者，在不脱离本公开之精神和范围内，当可做些许更动与润饰，因此本公开之保护范围当视后附之申请专利范围所界定者为准。
附图说明
图1A是为本公开超声波侦测中耳积水的数组式测量与判读装置的组成组件示意图。
图1B是为将用户的乳突表面定义成复数个数组式测量区域的示意图。
图2A是为使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置，测量与判读正常人与中耳积水患者之中耳积水状况的接收者操作特征曲线。
图2B是为仅使用单一机率密度函数参数，测量与判读正常人与中耳积水患者之中耳积水状况的接收者操作特征曲线。
图3A是为使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置，测量与判读轻中度中耳积水患者与重度中耳积水患者之中耳积水状况的接收者操作特征曲线。
图3B是为仅使用单一机率密度函数参数，测量与判读轻中度中耳积水患者与重度中耳积水患者之中耳积水状况的接收者操作特征曲线。
图4A是为使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置，测量与判读浆液性中耳积水患者与黏液性中耳积水患者之中耳积水状况的接收者操作特征曲线。
图4B是为仅使用单一机率密度函数参数，测量与判读浆液性中耳积水患者与黏液性中耳积水患者之中耳积水状况的接收者操作特征曲线。
附图标记
1:超声波侦测中耳积水之数组式测量与判读装置
11:超声波探头
12:超声波发射接收器
13:模拟数字转换器
14:分析单元
具体实施方式
依据本公开有关超声波探测与接收的操作程序与参数条件等，包含超声波探头及超声波发射接收器，是落在熟习此项技术之人士的专业素养与例行技术范畴内。
依据本公开有关模拟数字转换器的操作程序与参数条件等是落在熟习此项技术之人士的专业素养与例行技术范畴内。
请参见图1A，是为本公开超声波侦测中耳积水之数组式测量与判读装置1的组成组件示意图；本公开超声波侦测中耳积水之数组式测量与判读装置1包含：一超声波探头11、一超声波发射接收器12、一模拟数字转换器13、以及一分析单元14；其中，所述超声波探头11与所述超声波发射接收器12连接，而所述超声波发射接收器12与所述模拟数字转换器13连接，且所述模拟数字转换器13与所述分析单元14连接。
本公开超声波侦测中耳积水之数组式测量与判读装置1的使用方法如下：首先，将所述超声波探头11贴于一用户之耳朵后方乳突处的表面，以向作为音窗的乳突表面发送一超声波，并根据所述用户的中耳积水状况，产生相对 应的超声波回音频号以进行非侵入式的测量，而由于乳突区域较大，容易因测量位置不固定而导致结果失真，因此为了提高讯号测量与判读的准确性，在本公开超声波侦测中耳积水之数组式测量与判读装置1中，会先将所述用户的乳突表面定义成复数个数组式测量区域，综合考虑所述超声波探头11的大小与所述用户乳突的大小，本领域熟悉技艺者可以调整数组式测量区域的数量，而如图1B所示，最佳为定义成12个数组式测量区域，并针对各数组式测量区域分别产生一超声波回音频号；接着，再使用所述超声波发射接收器12接收所有的超声波回音频号，并透过所述模拟数字转换器13将该些超声波回音频号分别转换成相对应的一数字讯号，再将各数组式测量区域的该些数字讯号传送至所述分析单元14，以进一步的分析与计算该些数字讯号。
在本公开超声波侦测中耳积水之数组式测量与判读装置1的分析单元14中，会根据该些数字讯号计算出各数组式测量区域的一机率密度函数(probability density)参数(或称不同阶数的统计动差值)，且所述分析单元14会依据各数组式测量区域的数字讯号及其机率密度函数，使用以机器学习为基础的线性判别分析(Linear discriminant analysis,LDA)(或称费雪线性判别，Fisher Linear Discriminant)进行预训练，以找出各数组式测量区域之机率密度函数参数(不同阶数的统计动差值)分别所对应的重要性权重，接着将各数组式测量区域的机率密度函数参数分别与所对应的重要性权重相乘后再全部加总起来，即可获得一经加权的机率密度函数参数，而各数组式测量区域之机率密度函数参数是透过Nakagami参数进行量化，因此当机率密度函数参数的数值越高时，就表示所述用户的情况越接近所欲侦测的中耳积水状况(包含：中耳是否积水、中耳积水的严重程度、以及中耳积水的积液性质，将于后续进行详细的描述)，而在组合重要性权重后，所述经加权的机率密度函数参数也同样依循此判断标准，故藉由所述经加权的机率密度函数参数可精准地量化所述用户的中耳积水状况，而可据以判读所述用户之中耳疾病的状况。
在本公开之一较佳实施例中，所述分析单元14是为个人计算机。且根据本公开，所述分析单元14可以搭配一般市面流通的数据分析软件，例如：MATLAB，并提供机率密度函数与重要性权重等数值的计算之功能。
在本公开之一较佳实施例中，所述超声波探头11是为低频率延迟探头， 因此可分离量测讯号以及脉冲激励讯号(excitation pulse)，而增加量测讯号与机率密度函数分析的准确度。
接着，为测试本公开超声波侦测中耳积水之数组式测量与判读的方法及装置，确实能够有效提升测量与判读中耳积水状况的精准度，分别进行以下三个实验组测试：(1)中耳是否积水：分别于正常人(即未有中耳积水者)、中耳积水患者、以及经耳管手术后三个月患者(亦为中耳积水患者)，使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置；(2)中耳积水的严重程度：分别于轻中度中耳积水患者、以及重度中耳积水患者，使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置；以及(3)中耳积水的积液性质：分别于浆液性中耳积水患者、以及黏液性中耳积水患者，使用本公开超声波侦测中耳积水之数组式测量与判读的方法及装置；其中，本公开超声波侦测中耳积水之数组式测量与判读的方法及装置，会针对中耳是否积水、中耳积水的严重程度、以及中耳积水的积液性质三种不同侦测目的，而将所述三者所分别对应之各数组式测量区域的数字讯号及其机率密度函数，各自以机器学习为基础的线性判别分析进行预训练，以找出此三种不同侦测目的中，各自所对应之各数组式测量区域之机率密度函数参数分别所对应的重要性权重，再计算其各自之经加权的机率密度函数参数。
此外，在所述三个实验组中，对照组皆为未将乳突处的表面分隔成复数个数组式测量区域，对相同族群进行测试，即仅使用单一机率密度函数参数进行中耳积水状况的测量与判读。
在得到测试结果后，将各实验组别分别以统计软件(例如：sigmaplot)绘制接收者操作特征曲线(receiver operating characteristic curve,ROC，以下称为ROC曲线)以进行分析，并比较各实验组与对照组的ROC曲线下方面积(Area under the Curve of ROC,AUROC)，以评估本公开超声波侦测中耳积水之数组式测量与判读的方法及装置的准确度；其中，ROC曲线常用于分析一检测方法的准确度，其为一种二元分析模型，即输出结果只有二种类别的模型，例如在实验组(1)为有中耳积水或无中耳积水，在实验组(2)为轻中度或重度的中耳积水，以及在实验组(3)为浆液性或黏液性中耳积水，并设定一阈值以分别所述二种结果类别，再将各实验组之经加权的机率密度函数参数、或是将各 对照组之单一机率密度函数参数各自带入，以判断其各自的真第一类别性、伪第一类别性、真第二类别性、及伪第二类别性，并绘制出ROC曲线，而ROC曲线下方面积(Area under the Curve of ROC,AUROC)越接近1则表示带入ROC曲线进行分析的检测方法准确性越高；其中，在本公开实施例中，针对中耳是否积水、中耳积水的严重程度、以及中耳积水的积液性质三种不同侦测目的，分别设定以下阈值：>7.58判定有中耳积水；<0.60判定为重度的中耳积水；以及<-0.84判定为黏液性中耳积水。
所述三个实验组及其对照组的ROC分析曲线分别如图2至图4所示；其中，图2A与图2B为第(1)组之中耳积水与否的测试结果，图2A为本公开之一实施例，使用经加权的机率密度函数参数对正常人与中耳积水患者进行状况判读后，以ROC曲线进行分析所得的AUROC为0.85，而图2B则为对照组，仅使用单一机率密度函数参数对正常人与中耳积水患者进行状况判读，以ROC曲线进行分析所得的AUROC为0.87，二者相似。此结果显示，使用本公开测量与判读的方法及装置可以良好判断患者是否发生中耳积水。
图3A与图3B为第(2)组之中耳积水严重程度的测试结果，图3A为本公开之一实施例，以本公开测量与判读的方法及装置，使用经加权的机率密度函数参数对中度中耳积水患者与重度中耳积水患者进行状况判读后，以ROC进行分析所得的AUROC高达0.87，而图3B则为对照组，仅使用单一机率密度函数参数对轻中度中耳积水患者与重度中耳积水患者进行状况判读后，以ROC进行分析所得的AUROC仅有0.53，显示所述测量与判读方法并无预测中耳积水严重程度的能力。此结果显示，使用本公开测量与判读的方法及装置可以用于判断患者中耳积水的严重程度。
图4A与图4B为第(3)组之中耳积水之积液性质的测试结果，图4A为本公开之一实施例，以本公开测量与判读的方法及装置，使用经加权的机率密度函数参数对浆液性中耳积水患者与黏液性中耳积水患者进行状况判读后，以ROC进行分析所得的AUROC有效提升至0.69，而图4B则为对照组，仅使用单一机率密度函数参数对浆液性中耳积水患者与黏液性中耳积水患者进行状况判读后，以ROC进行分析所得的AUROC仅有0.55，显示所述测量与判读方法几乎无法预测中耳积水的积液特性。此结果显示，使用本公开测量 与判读的方法及装置可以有效地提升判断患者中耳积水之积液性质(即浆液性或黏液性)的能力。
因此，本公开超声波侦测中耳积水之数组式测量与判读装置及方法，藉由单阵元的超声波探头，以受测者耳朵后方的乳突表面作为音窗并发送超声波讯号，再接收并分析根据所述受测者之中耳积水状况所回传的超声波回音频号，以进行非侵入式的测量；其中，由于乳突处的区域大，且未有相关先前技术揭露可以反应最佳测量结果的侦测位置，因此在无法精准掌握测量位置的情况下，会因为测量位置的不确定性，造成结果失准。
据此，在本公开中，为了优化整体信息的运算及分析方式，并整合至原始操作接口中，以达到方便操作且快速分类中耳积水的状况与性质，来提高以超声波为基础之非侵入式量测方法及其装置的精准性及预期性，而进一步将受测者的乳突处表面分隔成复数个数组式的测量区域，并分别获取各数组式测量区域的超声波回音频号，再分别计算其机率密度函数参数后，于分析单元中搭配信息分析软件，并使用以机器学习为基础的线性判别分析进行预训练，找出各分隔之测量区域于判断中耳积水状况的重要性权重，并将各测量区域的机率密度函数参数经加权后，以所得的新参数相较于使用单一参数而言，不仅可以维持良好之判断患者是否发生中耳积水的能力，更可以用于判断患者中耳积水的严重程度，且还能有效改善判断患者中耳积水之积液性质(即浆液性或黏液性)的能力。
综上所述，本公开超声波侦测中耳积水之数组式测量与判读装置及方法能够改善先前技术既有的缺点，具有非侵入性、无辐射、适用于任何地点与时间等优点，尤其适合用于测量孩童的中耳积水状况，且若使用本公开重复于手术前后测量用户之中耳积水状况，则可用于评估手术前后之中耳积水是否有清除，最重要的是，本公开数据分析的方式还能够减少因主观判定导致的不确定性，使中耳积水的诊断更加快速、操作更方便，且检测结果更加准确。

Claims (12)

  1. 一种超声波侦测中耳积水状况之数组式测量与判读装置，包含：
    一超声波探头，是贴于一用户之耳朵后方的乳突表面并发送一超声波，且所述乳突表面分隔成复数个数组式测量区域，并依据所述用户的中耳积水状况，于各数组式测量区域分别产生一超声波回音频号；
    一超声波发射接收器，是与所述超声波探头连接，并接收各数组式测量区域的该些超声波回音频号；
    一模拟数字转换器，是与所述超声波发射接收器连接，并将该些超声波回音频号分别转换为一数字讯号；以及
    一分析单元，是与所述模拟数字转换器连接，且根据该些数字讯号计算出各数组式测量区域的一机率密度函数参数，再使用预训练模型找出分别对应于各数组式测量区域之机率密度函数参数的一重要性权重，并将各数组式测量区域的机率密度函数参数与所对应的重要性权重相乘后加总，以产生一经加权的机率密度函数参数，用于量化所述用户之中耳积水程度，据以判读所述用户之中耳疾病的状况。
  2. 根据权利要求1所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，对应于各数组式测量区域之机率密度函数参数的重要性权重是使用线性判别分析进行预训练模型所找出。
  3. 根据权利要求1所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，所述超声波侦测中耳积水之数组式测量与判读装置是提升判断所述用户中耳积水严重程度的准确性。
  4. 根据权利要求1所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，所述超声波侦测中耳积水之数组式测量与判读装置是提升判断所述用户中耳积水之积液性质的准确性。
  5. 根据权利要求1所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，所述超声波探头是为一低频率延迟探头。
  6. 根据权利要求1或5所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，所述分析单元是为一个人计算机。
  7. 根据权利要求1所述的超声波侦测中耳积水状况之数组式测量与判读装置，其中，所述中耳疾病包含各式性中耳炎、中耳积水、乳突积水、乳突炎以及耳管植入前后追踪。
  8. 一种超声波侦测中耳积水状况的数组式测量与判读方法，包含以下步骤：
    将一超声波探头贴于一用户之耳朵后方的乳突表面并发送一超声波，将所述乳突表面分隔成复数个数组式测量区域，并依据所述用户的中耳积水状况，于各数组式测量区域分别产生一超声波回音频号，并以一超声波发射接收器接收各数组式测量区域的该些超声波回音频号；以及
    将该些超声波回音频号，以一模拟数为转换器分别转换为相应的一数字讯号；
    其中，依据该些数字讯号，以一分析单元计算出各数组式测量区域的一机率密度函数参数，并使用预训练模型找出分别对应各数组式测量区域之机率密度函数参数的一重要性权重，再将各数组式测量区域的机率密度函数参数与所对应的重要性权重相乘后加总，以产生一经加权的机率密度函数参数，用于量化所述用户之中耳积水程度，据以判读所述用户之中耳疾病的状况。
  9. 根据权利要求8所述的超声波侦测中耳积水状况的数组式测量与判读方法，其中，对应于各数组式测量区域之机率密度函数参数的重要性权重是使用线性判别分析进行预训练模型所找出。
  10. 根据权利要求8所述的超声波侦测中耳积水状况的数组式测量与判读方法，其中，所述乳突表面分割成12个数组式测量区域。
  11. 根据权利要求8所述的超声波侦测中耳积水状况的数组式测量与判读方法，其中，所述超声波侦测中耳积水之数组式测量与判读方法是提升判断所述用户中耳积水严重程度的准确性。
  12. 根据权利要求8所述的超声波侦测中耳积水状况的数组式测量与判读方法，其中，所述超声波侦测中耳积水之数组式测量与判读方法是提升判断所述用户中耳积水之积液性质的准确性。
PCT/CN2022/075810 2022-02-10 2022-02-10 超声波侦测中耳积水之数组式测量与判读装置与方法 WO2023150960A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2022/075810 WO2023150960A1 (zh) 2022-02-10 2022-02-10 超声波侦测中耳积水之数组式测量与判读装置与方法
CN202280090598.0A CN118714978A (zh) 2022-02-10 2022-02-10 超声波侦测中耳积水之数组式测量与判读装置与方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/075810 WO2023150960A1 (zh) 2022-02-10 2022-02-10 超声波侦测中耳积水之数组式测量与判读装置与方法

Publications (1)

Publication Number Publication Date
WO2023150960A1 true WO2023150960A1 (zh) 2023-08-17

Family

ID=87563423

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/075810 WO2023150960A1 (zh) 2022-02-10 2022-02-10 超声波侦测中耳积水之数组式测量与判读装置与方法

Country Status (2)

Country Link
CN (1) CN118714978A (zh)
WO (1) WO2023150960A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040167404A1 (en) * 2003-02-20 2004-08-26 Robert Bessler Characterization of middle ear effusion
TW200718399A (en) * 2005-11-14 2007-05-16 Spentech Inc Ultrasound tympanoscope
KR20090128613A (ko) * 2008-06-11 2009-12-16 을지대학교 산학협력단 삼출성 중이염 진단장치
US20100069752A1 (en) * 2002-12-06 2010-03-18 Otosonics, Inc. Ultrasonic detection of ear disorders
TW201507701A (zh) * 2013-08-30 2015-03-01 Univ Chang Gung 可用於偵測中耳積水之判讀裝置
TW201641936A (zh) * 2015-05-22 2016-12-01 Univ Chang Gung 超音波定量組織特性靈敏度的提昇方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100069752A1 (en) * 2002-12-06 2010-03-18 Otosonics, Inc. Ultrasonic detection of ear disorders
US20040167404A1 (en) * 2003-02-20 2004-08-26 Robert Bessler Characterization of middle ear effusion
TW200718399A (en) * 2005-11-14 2007-05-16 Spentech Inc Ultrasound tympanoscope
KR20090128613A (ko) * 2008-06-11 2009-12-16 을지대학교 산학협력단 삼출성 중이염 진단장치
TW201507701A (zh) * 2013-08-30 2015-03-01 Univ Chang Gung 可用於偵測中耳積水之判讀裝置
TW201641936A (zh) * 2015-05-22 2016-12-01 Univ Chang Gung 超音波定量組織特性靈敏度的提昇方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CHEN CHIN-KUO, WAN YUNG-LIANG, TSUI PO-HSIANG, CHIU WEN-TA, JUI FANG: "Using 1 MHz pulse-echo ultrasound externally applied to detect mastoid effusion: Cadaver experiments", ULTRASONICS., IPC SCIENCE AND TECHNOLOGY PRESS LTD. GUILDFORD., GB, vol. 52, no. 5, 1 July 2012 (2012-07-01), GB , pages 663 - 667, XP093083131, ISSN: 0041-624X, DOI: 10.1016/j.ultras.2012.01.012 *
O. CLADE ; G. PALCZEWSKA ; J. J. LEWANDOWSKI ; P. KRAKOVITZ ; D. DINET: "P3Q-2 Development and Evaluation of a 20MHz Array for Ultrasonic Detection of Middle Ear Effusion", ULTRASONICS SYMPOSIUM, 2006. IEEE, IEEE, PI, 1 October 2006 (2006-10-01), Pi , pages 2357 - 2360, XP031076799, ISBN: 978-1-4244-0201-4, DOI: 10.1109/ULTSYM.2006.595 *

Also Published As

Publication number Publication date
CN118714978A (zh) 2024-09-27

Similar Documents

Publication Publication Date Title
Rao et al. Acoustic methods for pulmonary diagnosis
TW201519872A (zh) 非侵入式肝纖維化程度評估裝置與方法
Tsui et al. The effect of transducer characteristics on the estimation of Nakagami paramater as a function of scatterer concentration
JP2013519455A (ja) 患者の組織を特徴づける方法
US20110125061A1 (en) Monitoring Conditions Of A Patient&#39;s Urinary System
JP2007195984A (ja) 弾性映像形成のための超音波診断システム及び方法
US20210186426A1 (en) System and method for detection of middle ear fluids
KR100235170B1 (ko) 귀의 음향 반사율 곡선의 형상을 발생 및측정하는 장치 및 방법
TWI773472B (zh) 超音波偵測中耳積水之陣列式測量與判讀裝置與方法
Strahan et al. Effects of dolphin hearing bandwidth on biosonar click emissions
Gangadhar et al. Impact of image orientation on measurements of thyroid nodule stiffness using shear wave elastography
WO2023150960A1 (zh) 超声波侦测中耳积水之数组式测量与判读装置与方法
KR101014558B1 (ko) 스트레인 정규화를 통한 탄성영상 형성 방법 및 그를 위한초음파 시스템
Ayodele et al. A medical percussion instrument using a wavelet-based method for archivable output and automatic classification
CN112075955B (zh) 粘弹性介质超声波参数测量方法与装置
WO2020047805A1 (zh) 一种剪切波弹性成像方法和装置、及计算机存储介质
Hewener et al. A mobile ultrasound system for majority detection
TWI777553B (zh) 肝臟纖維化之非侵入式超音波檢測裝置
Shaw Comparison of wideband energy reflectance and tympanometric measures obtained with Reflwin Interacoustics, Mimosa Acoustics and GSI Tympstar systems
Rus et al. Effect of Contact Conditions of Torsional Wave Elastographic Probe on Human Cervix
Paunksnis et al. Ultrasound quantitative evaluation of human eye cataract
CN111938704B (zh) 膀胱容积检测方法、装置及电子设备
Shahid et al. Simulation-Based Approach to Detecting Pulmonary Embolism Using Capacitive Micromachined Ultrasonic Transducers
EP4018935A1 (en) Method for determining a geometry of an ear canal or a portion of an ear of a person
WO2024107904A1 (en) Detecting lung dysfunction using automated ultrasound monitoring

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22925322

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE