TWI812135B - Pneumonia remote detection system and method thereof - Google Patents

A pneumonia remote detection method includes: detecting a respiratory sample with a gas sensor array to obtain at least one gas feature pattern; capturing a secretion liquid sample with an image-capturing unit to obtain at least one secretion microscopic image; processing the secretion microscopic image to obtain at least one processed microscopic feature; executing a machine-learning procedure with the gas feature pattern and the processed microscopic feature to obtain at least one correlation of gas feature pattern and microscopic feature; and identifying an abnormal result of pneumonia with the correlation of gas feature pattern and microscopic feature.

肺炎遠端檢測系統及其方法 Pneumonia remote detection system and method thereof

本發明係關於一種肺炎〔pneumonia〕或肺部呼出氣體異常狀態之遠端〔remote〕檢測〔detection〕系統及其方法；特別是關於一種肺炎〔例如：新冠肺炎〕遠端呼出氣體〔exhalation〕檢測系統及其方法；更特別是關於一種肺炎遠端檢測系統及其方法採用機器學習方式〔machine learning〕進行遠端檢測一呼出氣體〔exhaling air〕。 The present invention relates to a remote detection system and method for detecting abnormal conditions of pneumonia or exhaled air in the lungs; in particular, it relates to remote exhalation detection of pneumonia (for example, COVID-19). Systems and methods thereof; more particularly, a remote detection system for pneumonia and a method thereof using machine learning to remotely detect exhaled air.

舉例而言，習用辨識肺炎種類之呼吸器裝置、其氣體辨識晶片及其氣體辨識方法，例如：中華民國專利公告第TW-I458464號之〝可早期偵測及辨識肺炎種類之呼吸器、其氣體辨識晶片、及其氣體辨識方法〞發明專利，其揭示一種氣體辨識裝置及其呼吸器，且該氣體辨識裝置包含一感測器陣列、一感測器介面電路、一隨機類神經網路晶片、一記憶體及一微控制器。 For example, respirator devices that are commonly used to identify pneumonia types, their gas identification chips and gas identification methods, such as: "Respirators that can early detect and identify pneumonia types, their gas "Identification Chip and Gas Identification Method" invention patent discloses a gas identification device and its respirator, and the gas identification device includes a sensor array, a sensor interface circuit, a random neural network chip, a memory and a microcontroller.

承上，前述第TW-I458464號之該感測器陣列包含數個感測器〔sensor〕及一感測薄膜，而該感測薄膜用以吸附數種待測氣體，以便經由該感測器產生一氣味訊號〔odor signal〕，且該氣味訊號可對應於該待測氣體之至少一種或數種。 Following the above, the sensor array of the aforementioned No. TW-I458464 includes several sensors and a sensing film, and the sensing film is used to adsorb several types of gases to be measured so as to pass through the sensor. An odor signal is generated, and the odor signal may correspond to at least one or several types of gases to be detected.

承上，前述第TW-I458464號之該感測器介面電路用以讀取及適當分析各個該待測氣體之氣味訊號，以便於該感測器介面電路可適當產生一氣體圖案訊號〔gas pattern signal〕，且該氣體圖案訊號可對應於其所屬之該待測氣體。 Following the above, the sensor interface circuit of the aforementioned No. TW-I458464 is used to read and appropriately analyze the odor signals of each gas to be measured, so that the sensor interface circuit can appropriately generate a gas pattern signal [gas pattern signal), and the gas pattern signal can correspond to the gas to be measured to which it belongs.

承上，前述第TW-I458464號之該隨機類神經網路晶片用以放大各個該氣體圖案訊號之間的差異，並用以降低各個該氣體圖案訊號的維度〔dimensional reduction〕，以便產生一分析結果，且該記憶體用以儲存一氣體訓練資料。 Following the above, the aforementioned stochastic neural network chip No. TW-I458464 is used to amplify the difference between each gas pattern signal, and to reduce the dimensionality of each gas pattern signal to produce an analysis result. , and the memory is used to store gas training data.

承上，前述第TW-I458464號之該微控制器用以接收該分析結果，並依該分析結果執行一混合氣體辨識演算法，以便進行辨識該待測氣體之所屬種類。倘若該分析結果為一不存在氣體時，將該不存在氣體歸類為不存在於該氣體訓練資料的一未知氣體分類，再依該氣體訓練資料產生一辨識結果。 Following the above, the aforementioned microcontroller No. TW-I458464 is used to receive the analysis result and execute a mixed gas identification algorithm according to the analysis result in order to identify the type of the gas to be measured. If the analysis result is that a gas does not exist, the non-existent gas is classified as an unknown gas classification that does not exist in the gas training data, and an identification result is generated based on the gas training data.

承上，前述第TW-I458464號之該呼吸器包含一吐氣端管路及一氣體辨識裝置，而該氣體辨識裝置利用一氣體辨識晶片進行分析一病人及其呼氣，且該病人於該吐氣端管路中進行呼出一氣體，以便利用該氣體辨識晶片辨識肺炎的種類。另外，該氣體辨識裝置係直接連結位於該病人之體外的該吐氣端管路，並擷取該病人所呼出的該氣體，以便進行辨識該氣體。 Following the above, the aforementioned respirator No. TW-I458464 includes an exhalation end pipe and a gas identification device, and the gas identification device uses a gas identification chip to analyze a patient and his exhalation, and the patient exhales A gas is exhaled in the end pipe, so that the gas identification chip can be used to identify the type of pneumonia. In addition, the gas identification device is directly connected to the expiration end pipe located outside the patient's body, and captures the gas exhaled by the patient in order to identify the gas.

然而，前述第TW-I458464號之該氣體辨識裝置及其呼吸器僅採用該感測器介面電路及隨機類神經網路晶片分析該待測氣體，且利用該混合氣體辨識演算法進行演算及辨識該待測氣體之所屬種類而已，其並非用以進行遠端〔例如：偏鄉地區〕檢測肺部呼吸異常狀態〔例如：新冠肺炎及其變種病毒感染肺炎〕之呼出氣體。 However, the aforementioned gas identification device and its respirator No. TW-I458464 only use the sensor interface circuit and the random neural network chip to analyze the gas to be measured, and use the mixed gas identification algorithm for calculation and identification. The gas to be measured is of a specific type, and it is not used to detect exhaled gas in remote areas (such as rural areas) for abnormal pulmonary breathing conditions (such as pneumonia caused by COVID-19 and its mutant viruses).

另一習用肺炎偵測裝置，例如：美國專利公開第US-20210315481號之〝Pneumonia detection device〞發明專利申請案，其揭示一種肺炎偵測裝置，而該肺炎偵測 裝置用以偵測一患者之一呼出氣體，以便判讀該患者的肺部是否染上一種或數種病菌。 Another commonly used pneumonia detection device, for example: U.S. Patent Publication No. US-20210315481 "Pneumonia detection device" invention patent application, which discloses a pneumonia detection device, and the pneumonia detection device The device is used to detect the exhaled air of a patient to determine whether the patient's lungs are infected with one or several germs.

承上，前述第US-20210315481號之該肺炎偵測裝置包含一多氣體感測模組、一溫濕度控制模組及一運算控制單元，且該多氣體感測模組包含一腔室、一氣體感測器陣列、一進氣管路、一進氣閥門、一出氣閥門及一氣體通道。 Following the above, the pneumonia detection device in the aforementioned No. US-20210315481 includes a multi-gas sensing module, a temperature and humidity control module and a computing control unit, and the multi-gas sensing module includes a chamber, a A gas sensor array, an air inlet pipeline, an air inlet valve, an air outlet valve and a gas channel.

承上，前述第US-20210315481號之該氣體感測器陣列設置於該腔室，而該氣體感測器陣列對一病菌代謝物所產生一氣體反應產生數個特徵訊號。該進氣管路用以導引該呼出氣體進入該腔室，而該進氣閥門設置於該進氣管路上，並可決定該呼出氣體是否通過該進氣管路而進入該腔室，且該出氣閥門設置在該腔室上，並可決定該呼出氣體是否離開該腔室，且該氣體通道連通於該腔室及進氣管路之間。 Following the above, the gas sensor array in the aforementioned US-20210315481 is disposed in the chamber, and the gas sensor array generates several characteristic signals in response to a gas produced by a bacterial metabolite. The air inlet pipe is used to guide the exhaled gas into the chamber, and the air inlet valve is disposed on the air inlet pipe and can determine whether the exhaled gas enters the chamber through the air inlet pipe, and The air outlet valve is arranged on the chamber and can determine whether the exhaled gas leaves the chamber, and the gas channel is connected between the chamber and the air inlet pipeline.

承上，前述第US-20210315481號之該溫濕度控制模組用以控制及量測該氣體通道內之一溫度及一溼度，而該運算控制單元連接於該多氣體感測模組及溫濕度控制模組，以便自該多氣體感測模組及溫濕度控制模組獲得該溫度及溼度。 Following the above, the temperature and humidity control module in the aforementioned No. US-20210315481 is used to control and measure a temperature and a humidity in the gas channel, and the computing control unit is connected to the multi-gas sensing module and the temperature and humidity Control module to obtain the temperature and humidity from the multi-gas sensing module and the temperature and humidity control module.

承上，前述第US-20210315481號之該腔室於偵測時處於一溫度〔介於45℃至60℃〕及一濕度〔介於7%至20%之間〕，而該氣體感測器陣列接觸該呼出氣體後，產生數個量測訊號，且數個該量測訊號對應於該氣體感測器陣列，且該運算控制單元依該量測訊號及病菌之特徵訊號進行比對後，產生一比對結果，以判讀該患者是否染上該病菌。 Following the above, the chamber of the aforementioned US-20210315481 is at a temperature [between 45°C and 60°C] and a humidity [between 7% and 20%] at the time of detection, and the gas sensor After the array contacts the exhaled gas, a plurality of measurement signals are generated, and the plurality of measurement signals correspond to the gas sensor array, and the computing control unit compares the measurement signals with the characteristic signals of germs, A comparison result is generated to determine whether the patient is infected with the bacteria.

然而，前述第US-20210315481號之該氣體辨識裝置及其呼吸器僅用以辨識呼吸器相關肺炎〔VAP， ventilator-associated pneumonia〕，其為加護病房〔intensive care unit，ICU，最常見的死因，例如克雷伯氏肺炎桿菌毒性菌株K1/K2〔Klebsiella pneumoniae virulent strain K1/K2〕感染而已，其並非用以進行遠端〔例如：偏鄉地區〕檢測肺部呼吸異常狀態〔例如：新冠肺炎及其變種病毒感染肺炎〕之呼出氣體。 However, the aforementioned gas identification device and its respirator No. US-20210315481 are only used to identify respirator-associated pneumonia [VAP, ventilator-associated pneumonia], which is the most common cause of death in the intensive care unit (ICU), such as Klebsiella pneumoniae virulent strain K1/K2 [Klebsiella pneumoniae virulent strain K1/K2] infection, it is not used for Detect exhaled air from remote locations (for example, in rural areas) for abnormal lung breathing conditions (for example, pneumonia caused by COVID-19 and its variant viruses).

另一習用特發性間質性肺炎檢測方法及其標記，例如：美國專利公開第US-20110086360號之〝Method for Detection of Idiopathic Interstitial Pneumonia〞發明專利申請案，其亦揭示一種特發性間質性肺炎〔idiopathic interstitial pneumonia，IIP〕檢測方法及其標記。該特發性間質性肺炎檢測方法選擇採用一生物樣本〔biological sample〕並進行量測。 Another commonly used detection method and marker for idiopathic interstitial pneumonia, such as the invention patent application "Method for Detection of Idiopathic Interstitial Pneumonia" in U.S. Patent Publication No. US-20110086360, which also discloses an idiopathic interstitial pneumonia Detection methods and markers for idiopathic interstitial pneumonia (IIP). This idiopathic interstitial pneumonia detection method selects a biological sample and performs measurement.

承上，前述第US-20110086360號之該特發性間質性肺炎檢測方法於該生物樣本進行量測一骨膜素基因〔periostin gene〕之超標量〔expression level〕或一骨膜素蛋白〔periostin protein〕數量。另外，該特發性間質性肺炎檢測方法可選擇採用一標記〔marker〕。 Following the above, the aforementioned idiopathic interstitial pneumonia detection method No. US-20110086360 is used to measure the super-standard expression level of a periostin gene or a periostin protein in the biological sample. 〕quantity. In addition, this idiopathic interstitial pneumonia detection method can optionally use a marker.

然而，前述第US-20110086360號之該檢測方法及其標記僅用以檢測特發性間質性肺炎〔IIP〕而已，其並非用以進行遠端〔例如：偏鄉地區〕檢測肺部呼吸異常狀態〔例如：新冠肺炎及其變種病毒感染肺炎〕之呼出氣體。 However, the aforementioned detection method and its markers in US-20110086360 are only used to detect idiopathic interstitial pneumonia [IIP], and are not used to detect lung respiratory abnormalities in remote areas (such as rural areas). Exhaled gas due to conditions [for example: pneumonia caused by COVID-19 and its variant viruses].

另一習用特發性間質性肺炎預防劑或治療劑之篩選方法及其篩選或檢測套組，例如：美國專利公開第US-20130230861號之〝Method for Detection of Idiopathic Interstitial Pneumonia〞發明專利申請案，其亦揭示一種特發性間質性肺炎預防劑或治療劑之篩選方法及其篩選或檢測套組。該特發性間質性肺炎預防劑或治療劑之篩選方法 進行量測一骨膜素基因之超標量或一骨膜素蛋白數量。 Another commonly used screening method for preventive or therapeutic agents for idiopathic interstitial pneumonia and its screening or detection kit, for example: "Method for Detection of Idiopathic Interstitial Pneumonia" invention patent application in U.S. Patent Publication No. US-20130230861 , which also discloses a screening method for preventive or therapeutic agents for idiopathic interstitial pneumonia and its screening or detection kit. Screening method for preventive or therapeutic agents for idiopathic interstitial pneumonia A superscalar amount of a periostin gene or a periostin protein amount is measured.

承上，前述第US-20130230861號之該特發性間質性肺炎預防劑或治療劑之篩選方法為於候選物質〔candidate substance〕中培養細胞〔cell〕能產生骨膜素，且該篩選套組〔kit〕可選擇包含該細胞，因而其具有能力產生骨膜素。 Following on from the above, the screening method for the preventive or therapeutic agent for idiopathic interstitial pneumonia in the aforementioned US-20130230861 is to culture cells in the candidate substance [candidate substance] that can produce periostin, and the screening kit The [kit] may optionally contain the cells so that they have the ability to produce periostin.

承上，前述第US-20130230861號之該檢測套組可選擇包含一多核苷酸〔polynucleotide〕，且該多核苷酸包含至少21個核苷酸。另外，該檢測套組可選擇包含一抗體〔antibody〕，而該抗體可辨認一肽〔peptide〕物質，且該肽物質具有骨膜素蛋白之胺基酸序列〔amino acid sequence〕。 Following the above, the test kit of the aforementioned US-20130230861 may optionally include a polynucleotide, and the polynucleotide includes at least 21 nucleotides. In addition, the detection kit may optionally include an antibody, and the antibody can recognize a peptide substance, and the peptide substance has the amino acid sequence of periostin protein.

然而，前述第US-20130230861號之該篩選方法及其篩選或檢測套組僅用以篩選特發性間質性肺炎〔IIP〕而已，其並非用以進行遠端〔例如：偏鄉地區〕檢測肺部呼吸異常狀態〔例如：新冠肺炎及其變種病毒感染肺炎〕之呼出氣體。 However, the screening method and its screening or detection kit in the aforementioned No. US-20130230861 are only used to screen for idiopathic interstitial pneumonia [IIP], and are not used for remote (such as: rural areas) detection. Exhaled air caused by abnormal pulmonary breathing conditions (for example: pneumonia caused by COVID-19 and its variant viruses).

雖然前述專利公告第TW-I458464號、公開第US-20210315481號、公開第US-20110086360號及公開第US-20130230861號已揭示肺炎相關辨識技術，但其並未提供如何進一步遠端檢測之相關技術，因此習用肺炎相關辨識必然存在進一步遠端〔例如：偏鄉地區〕檢測的潛在需求。 Although the aforementioned Patent Publication No. TW-I458464, Publication No. US-20210315481, Publication No. US-20110086360 and Publication No. US-20130230861 have disclosed pneumonia-related identification technologies, they have not provided relevant technologies for further remote detection. , so the commonly used pneumonia-related identification must have potential demand for further remote (for example: rural areas) testing.

顯然，前述中華民國專利公告第TW-I458464號、美國專利公開第US-20210315481號、公開第US-20110086360號及公開第US-20130230861號發明專利申請案僅為本發明技術背景之參考及說明目前技術發展狀態而已，其並非用以限制本發明之範圍。 Obviously, the aforementioned Republic of China Patent Publication No. TW-I458464, United States Patent Publication No. US-20210315481, Publication No. US-20110086360 and Publication No. US-20130230861 Invention Patent Application are only for reference and explanation of the technical background of the present invention. This is merely a state of technological development and is not intended to limit the scope of the present invention.

有鑑於此，本發明為了滿足上述需求，其提供 一種肺炎遠端檢測系統及其方法，其利用一氣體感測器陣列偵測一呼吸道氣體樣本，以便獲得一氣體特徵圖樣，並利用一影像攝取單元攝取一分泌物液體樣本，以便獲得一分泌物液體顯微影像，且將該分泌物液體顯微影像進行影像處理，以便獲得一已處理影像特徵，且利用該氣體特徵圖樣及已處理影像特徵進行一機器學習作業，以便獲得一氣體特徵與影像特徵相關性，且利用該氣體特徵與影像特徵相關性判斷產生一肺部呼出氣體異常狀態結果，以便大幅提升檢測肺部呼出氣體異常狀態之可靠度。 In view of this, in order to meet the above needs, the present invention provides A remote detection system for pneumonia and its method, which uses a gas sensor array to detect a respiratory gas sample to obtain a gas characteristic pattern, and uses an image acquisition unit to take in a secretion liquid sample to obtain a secretion liquid microscopic image, and image processing is performed on the secretion liquid microscopic image to obtain a processed image feature, and a machine learning operation is performed using the gas feature pattern and the processed image feature to obtain a gas feature and image Feature correlation, and use the correlation between the gas feature and the image feature to determine and generate an abnormal status result of exhaled lung gas, so as to greatly improve the reliability of detecting abnormal status of exhaled lung gas.

本發明較佳實施例之主要目的係提供一種肺炎遠端檢測系統及其方法，其利用一氣體感測器陣列偵測一呼吸道氣體樣本，以便獲得一氣體特徵圖樣，並利用一影像攝取單元攝取一分泌物液體樣本，以便獲得一分泌物液體顯微影像，且將該分泌物液體顯微影像進行影像處理，以便獲得一已處理影像特徵，且利用該氣體特徵圖樣及已處理影像特徵進行一機器學習作業，以便獲得一氣體特徵與影像特徵相關性，且利用該氣體特徵與影像特徵相關性判斷產生一肺部呼出氣體異常狀態結果，以達成提升遠端檢測肺部呼出氣體異常狀態之可靠度之目的。 The main purpose of the preferred embodiment of the present invention is to provide a remote detection system and method for pneumonia, which uses a gas sensor array to detect a respiratory gas sample to obtain a gas characteristic pattern, and uses an image capture unit to capture A secretion liquid sample is used to obtain a secretion liquid microscopic image, and the secretion liquid microscopic image is image-processed to obtain a processed image feature, and the gas feature pattern and the processed image feature are used to perform a Machine learning operations are used to obtain the correlation between a gas feature and an image feature, and use the correlation between the gas feature and the image feature to determine and generate an abnormal status result of exhaled lung air, in order to achieve improved remote detection of abnormal status of exhaled lung air. purpose.

為了達成上述目的，本發明較佳實施例之肺炎遠端檢測方法包含： In order to achieve the above objectives, the remote detection method of pneumonia in a preferred embodiment of the present invention includes:

利用一氣體感測器陣列偵測一呼吸道氣體樣本，以便獲得一氣體特徵圖樣； using a gas sensor array to detect a respiratory gas sample to obtain a gas characteristic pattern;

利用一影像攝取單元攝取一分泌物液體樣本，以便獲得一分泌物液體顯微影像； Using an image acquisition unit to acquire a secretion liquid sample to obtain a secretion liquid microscopic image;

將該分泌物液體顯微影像進行影像處理，以便獲得一已處理影像特徵； Perform image processing on the secretion liquid microscopic image to obtain a processed image feature;

利用該氣體特徵圖樣及已處理影像特徵進行一機 器學習作業，以便獲得一氣體特徵與影像特徵相關性；及 Using the gas characteristic pattern and the processed image characteristics, a machine Machine learning operation to obtain the correlation between a gas feature and an image feature; and

利用該氣體特徵與影像特徵相關性判斷產生一肺部呼出氣體異常狀態結果。 The correlation between the gas characteristics and the image characteristics is used to determine the abnormal state of exhaled gas in the lungs.

本發明較佳實施例之該呼吸道氣體樣本包含一揮發性有機物氣體、一無機物氣體或其組合體。 In a preferred embodiment of the present invention, the respiratory gas sample includes a volatile organic gas, an inorganic gas or a combination thereof.

本發明較佳實施例之該分泌物液體樣本包含一唾液、一痰液、一鼻咽液或其任意組合體。 In a preferred embodiment of the present invention, the secretion liquid sample includes saliva, sputum, nasopharyngeal fluid or any combination thereof.

本發明較佳實施例之該呼吸道氣體樣本、分泌物液體樣本或兩者經由一抽吸裝置取得。 In a preferred embodiment of the present invention, the respiratory gas sample, secretion liquid sample, or both are obtained through a suction device.

本發明較佳實施例之利用一體溫感測器量測一體溫資料，且該體溫資料為一正常體溫資料及一發燒體溫資料，以便利用該體溫資料判斷是否為一有症狀感染或一無症狀感染。 A preferred embodiment of the present invention uses a body temperature sensor to measure body temperature data, and the body temperature data is a normal body temperature data and a fever body temperature data, so that the body temperature data can be used to determine whether it is a symptomatic infection or an asymptomatic infection. Infect.

為了達成上述目的，本發明較佳實施例之肺炎遠端檢測系統包含： In order to achieve the above objectives, the pneumonia remote detection system in a preferred embodiment of the present invention includes:

一採樣裝置單元，其用以採樣於一呼吸道，以便獲得一呼吸道氣體樣本及一分泌物液體樣本； A sampling device unit for sampling a respiratory tract to obtain a respiratory gas sample and a secretion liquid sample;

一氣體感測器陣列，其連接於該採樣裝置，並利用該氣體感測器陣列偵測該呼吸道氣體樣本，以便獲得一氣體特徵圖樣； A gas sensor array, which is connected to the sampling device and uses the gas sensor array to detect the respiratory gas sample to obtain a gas characteristic pattern;

一影像攝取單元，其連接於該採樣裝置，並利用該影像攝取單元攝取該分泌物液體樣本，以便獲得一分泌物液體顯微影像；及 An image capture unit is connected to the sampling device and uses the image capture unit to capture the secretion liquid sample to obtain a secretion liquid microscopic image; and

一演算及處理單元，其將該分泌物液體顯微影像進行影像處理，以便獲得一已處理影像特徵； A calculation and processing unit that performs image processing on the secretion liquid microscopic image to obtain a processed image feature;

其中利用該氣體特徵圖樣及已處理影像特徵進行一機器學習作業，以便獲得一氣體特徵與影像特徵相關性，且利用該氣體特徵與影像特徵相關性判斷產生一肺部呼出氣體異常狀態結果。 The gas feature pattern and the processed image feature are used to perform a machine learning operation to obtain a correlation between the gas feature and the image feature, and the correlation between the gas feature and the image feature is used to determine and generate a result of abnormal lung exhaled gas status.

本發明較佳實施例之該氣體感測器陣列包含數個場效電晶體〔MOSFET〕元件。 The gas sensor array according to the preferred embodiment of the present invention includes a plurality of field effect transistor (MOSFET) elements.

本發明較佳實施例之該氣體感測器陣列配置於一電子鼻裝置。 In a preferred embodiment of the present invention, the gas sensor array is configured in an electronic nose device.

本發明較佳實施例之該影像攝取單元包含一電子顯微鏡單元、一數位顯微鏡單元或其它顯微鏡單元。 The image capturing unit in a preferred embodiment of the present invention includes an electron microscope unit, a digital microscope unit or other microscope units.

本發明較佳實施例另包含一體溫感測器單元，且該體溫感測器單元用以量測一體溫資料。 A preferred embodiment of the present invention further includes a body temperature sensor unit, and the body temperature sensor unit is used to measure body temperature data.

本發明較佳實施例之該體溫感測器單元選自一貼紙體溫計、一奶嘴體溫計、一紅外線體溫感測器、一紅外線影像體溫偵測器或其任意組合。 The body temperature sensor unit of the preferred embodiment of the present invention is selected from a sticker thermometer, a pacifier thermometer, an infrared body temperature sensor, an infrared imaging body temperature detector or any combination thereof.

1:採樣裝置單元 1: Sampling device unit

11:呼吸道氣體樣本 11:Respiratory gas sample

12:分泌物液體樣本 12: Secretion liquid sample

2:氣體感測器陣列 2: Gas sensor array

3:影像攝取單元 3:Image capture unit

4:資料傳輸單元 4: Data transmission unit

5:演算及處理單元 5: Calculation and processing unit

6:體溫感測器單元 6: Body temperature sensor unit

第1圖：本發明第一較佳實施例之肺炎遠端檢測系統之架構示意圖。 Figure 1: Schematic structural diagram of the pneumonia remote detection system according to the first preferred embodiment of the present invention.

第2圖：本發明較佳實施例之肺炎遠端檢測方法之流程示意圖。 Figure 2: A schematic flow chart of the remote detection method of pneumonia according to a preferred embodiment of the present invention.

第3圖：本發明較佳實施例之肺炎遠端檢測系統及其方法獲得原始數據及其氣體特徵圖樣之示意圖。 Figure 3: A schematic diagram of the pneumonia remote detection system and its method in the preferred embodiment of the present invention to obtain raw data and its gas characteristic pattern.

第4圖：本發明較佳實施例之肺炎遠端檢測系統及其方法利用氣體感測器陣列量測獲得呼吸道氣體樣本及其氣體特徵資料之示意圖。 Figure 4: A schematic diagram of the pneumonia remote detection system and its method using a gas sensor array to measure and obtain respiratory gas samples and gas characteristic data according to the preferred embodiment of the present invention.

第5圖：本發明較佳實施例之肺炎遠端檢測系統及其方法獲得顯微影像之示意圖。 Figure 5: Schematic diagram of microscopic images obtained by the remote pneumonia detection system and its method according to the preferred embodiment of the present invention.

第6圖：本發明較佳實施例之肺炎遠端檢測系統及其方法採用通訊架構之示意圖。 Figure 6: A schematic diagram of the communication architecture adopted by the pneumonia remote detection system and its method according to the preferred embodiment of the present invention.

第7圖：本發明第二較佳實施例之肺炎遠端檢測系統之示意圖。 Figure 7: Schematic diagram of the pneumonia remote detection system according to the second preferred embodiment of the present invention.

為了充分瞭解本發明，於下文將舉例較佳實施例並配合所附圖式作詳細說明，且其並非用以限定本發明。 In order to fully understand the present invention, preferred embodiments will be exemplified and described in detail below with the accompanying drawings, which are not intended to limit the present invention.

本發明較佳實施例之肺炎遠端檢測系統及其方法為呼吸道氣體樣本之分析及分泌物液體樣本影像辨識作業之前處理，其適用於各種肺部呼出氣體異常狀態之辨識裝置及其相關應用設備，例如：各類型電腦系統、電腦網路系統〔如物聯網〕、自動化遠端檢測系統或醫療照護系統，但其並非用以限定本發明之應用範圍。 The pneumonia remote detection system and its method in the preferred embodiment of the present invention are pre-processing for the analysis of respiratory gas samples and image recognition of secretion liquid samples, and are suitable for identification devices and related application equipment for various abnormal conditions of exhaled lung gases. , for example: various types of computer systems, computer network systems (such as the Internet of Things), automated remote detection systems or medical care systems, but they are not used to limit the scope of application of the present invention.

第1圖揭示本發明第一較佳實施例之肺炎遠端檢測系統之架構示意圖。請參照第1圖所示，舉例而言，本發明第一較佳實施例之肺炎遠端檢測系統包含一採樣裝置單元〔sampling device unit〕1、一氣體感測器陣列〔gas sensor array〕2、一影像攝取單元〔image capturing unit〕3、一資料傳輸單元〔data transmitting unit〕4及一演算及處理單元〔data calculation and processing unit〕5，且可依不同需求更換該資料傳輸單元4及演算及處理單元5或增加其它相關周邊設備。 Figure 1 shows a schematic structural diagram of a pneumonia remote detection system according to the first preferred embodiment of the present invention. Please refer to Figure 1. For example, the pneumonia remote detection system according to the first preferred embodiment of the present invention includes a sampling device unit 1 and a gas sensor array 2. , an image capturing unit [image capturing unit] 3, a data transmitting unit [data transmitting unit] 4 and a calculation and processing unit [data calculation and processing unit] 5, and the data transmission unit 4 and the calculation unit can be replaced according to different needs. and processing unit 5 or add other related peripheral equipment.

請參照第1圖所示，舉例而言，該採樣裝置單元1包含一抽吸裝置〔未繪示〕，而該抽吸裝置可選自一氣壓差抽吸裝置、一機械式抽吸裝置、一幫浦〔pump〕抽吸裝置或其它電動抽吸裝置，且該採樣裝置單元1可選自一可拋式〔disposable〕或單次使用〔single use〕採樣裝置單元。 Please refer to Figure 1. For example, the sampling device unit 1 includes a suction device (not shown), and the suction device can be selected from a pressure difference suction device, a mechanical suction device, A pump suction device or other electric suction device, and the sampling device unit 1 can be selected from a disposable (disposable) or single use (single use) sampling device unit.

請再參照第1圖所示，舉例而言，該採樣裝置單元1具有一採樣管體〔例如：中空式塑膠管體或其它材質管體〕、一閥門〔未標示〕及一採樣袋體〔例如：密封式塑膠袋體或其它透明材質袋體〕，而該閥門配置於該採樣管體及採樣袋體之間，且該採樣管體經由該閥門適當連接於該採樣袋體，以便一使用者〔未標示〕經由該採樣管 體適當提供一樣本至該採樣袋體，並將該閥門由開啟切換至關閉。 Please refer to Figure 1 again. For example, the sampling device unit 1 has a sampling pipe body [for example: a hollow plastic pipe body or other material pipe body], a valve [not labeled] and a sampling bag body [ For example: a sealed plastic bag or other transparent material bag), and the valve is disposed between the sampling tube and the sampling bag, and the sampling tube is appropriately connected to the sampling bag through the valve for easy use. or (unmarked) passes through the sampling tube The body appropriately provides a sample to the sampling bag body and switches the valve from open to closed.

請再參照第1圖所示，舉例而言，該採樣裝置單元1自該使用者取得至少一呼吸道氣體樣本11及至少一分泌物液體樣本12，且該呼吸道氣體樣本11、分泌物液體樣本12或兩者經由該抽吸裝置以適當技術手段〔例如：自動化方式、半自動化方式或手動方式〕自該使用者〔例如：受測者或受測病患〕取得。 Please refer to Figure 1 again. For example, the sampling device unit 1 obtains at least one respiratory gas sample 11 and at least one secretion liquid sample 12 from the user, and the respiratory gas sample 11 and the secretion liquid sample 12 Or both are obtained from the user (for example: the subject or the patient being tested) through the suction device using appropriate technical means (for example: automated, semi-automatic or manual).

請再參照第1圖所示，舉例而言，該氣體感測器陣列2包含數個氣體感測器〔未繪示〕，並將數個該氣體感測器適當排列形成一陣列單元，且該氣體感測器陣列包含數個金屬氧化物半導體場效電晶體元件〔MOSFET，metal-oxide-semiconductor field-effect transistor〕或其它氣體感測器。 Please refer to Figure 1 again. For example, the gas sensor array 2 includes several gas sensors (not shown), and the several gas sensors are appropriately arranged to form an array unit, and The gas sensor array includes several metal-oxide-semiconductor field-effect transistor elements [MOSFET, metal-oxide-semiconductor field-effect transistor] or other gas sensors.

請再參照第1圖所示，舉例而言，該氣體感測器陣列2之數個氣體感測器可選擇包含甲烷〔methane〕感測器、丁烷〔butane〕感測器、乙醇〔ethanol〕感測器、一氧化碳〔carbon monoxide〕感測器或其它功能氣體感測器〔例如：濕度感測器〕。 Please refer to Figure 1 again. For example, the gas sensors of the gas sensor array 2 may include a methane sensor, a butane sensor, and an ethanol sensor. 〕sensor, carbon monoxide sensor or other functional gas sensor 〔for example: humidity sensor〕.

請再參照第1圖所示，舉例而言，該氣體感測器陣列2可選擇適當連接於該採樣裝置1及其周邊設備，且該氣體感測器陣列2可選擇適當連接於該資料傳輸單元4。另外，本發明另一較佳實施例之該氣體感測器陣列2可選擇適當配置於一電子鼻裝置〔electronic nose〕、一醫療檢測裝置或其它具類似功能之裝置。 Please refer to Figure 1 again. For example, the gas sensor array 2 can be appropriately connected to the sampling device 1 and its peripheral equipment, and the gas sensor array 2 can be appropriately connected to the data transmission Unit 4. In addition, the gas sensor array 2 of another preferred embodiment of the present invention can be appropriately configured in an electronic nose device, a medical detection device, or other devices with similar functions.

請再參照第1圖所示，舉例而言，本發明另一較佳實施例之該採樣裝置單元1或氣體感測器陣列2可選擇適當技術手段〔例如：手動、半自動或全自動連接方式〕連接於一檢驗裝置，以便進一步適當檢驗該呼吸道氣體樣 本11。 Please refer to Figure 1 again. For example, the sampling device unit 1 or the gas sensor array 2 of another preferred embodiment of the present invention can select appropriate technical means [for example: manual, semi-automatic or fully automatic connection methods 〕Connected to a testing device for further appropriate testing of the respiratory gas sample Ben 11.

請再參照第1圖所示，舉例而言，該影像攝取單元3可選擇包含一電子顯微鏡單元、一數位顯微鏡單元、其它顯微鏡單元〔microscope unit〕或其它具類似顯微鏡功能之單元、構件或裝置，以便自該樣本適當取得一樣本顯微影像。 Please refer to Figure 1 again. For example, the image capture unit 3 may optionally include an electron microscope unit, a digital microscope unit, other microscope units, or other units, components or devices with similar microscope functions. , in order to properly obtain a sample microscopic image from the sample.

請再參照第1圖所示，舉例而言，該影像攝取單元3可選擇適當連接於該採樣裝置1及其周邊設備，且該影像攝取單元3可選擇適當連接於該資料傳輸單元4。另外，本發明另一較佳實施例之該影像攝取單元3可選擇適當配置於一醫療檢測裝置或其它具類似功能之裝置。 Please refer to Figure 1 again. For example, the image capture unit 3 can be appropriately connected to the sampling device 1 and its peripheral equipment, and the image capture unit 3 can be appropriately connected to the data transmission unit 4 . In addition, the image capturing unit 3 of another preferred embodiment of the present invention can be appropriately configured in a medical testing device or other devices with similar functions.

請再參照第1圖所示，舉例而言，該影像攝取單元3可選擇為一攝影單元〔video unit〕、一照相機單元〔camera unit〕或一影像攝取鏡頭單元〔image-capturing lens unit〕，而該影像攝取單元3用以輸入或攝取至少一個或數個〔一系列〕樣本影像及其相關影像，且該樣本影像包含單一個樣本影像〔即同一使用者〕或不同人之數個樣本影像〔m×n〕或影像方塊。 Please refer to Figure 1 again. For example, the image capturing unit 3 can be selected as a video unit, a camera unit or an image-capturing lens unit. The image capture unit 3 is used to input or capture at least one or several (series) sample images and related images, and the sample images include a single sample image (that is, the same user) or several sample images of different people. [ m × n ] or image square.

請再參照第1圖所示，舉例而言，該資料傳輸單元4選自一計算機裝置，且該計算機裝置可選自一工作站電腦〔workstation computer〕、一桌上型電腦〔desktop computer〕、一筆記型電腦〔notebook或laptop computer〕、一平板電腦〔tablet personal computer〕、一行動通訊裝置〔mobile communication device〕、一智慧型手機〔smart phone〕或其它具計算機功能之裝置，但其並非用以限定本發明之範圍。 Please refer to Figure 1 again. For example, the data transmission unit 4 is selected from a computer device, and the computer device can be selected from a workstation computer, a desktop computer, a Notebook or laptop computer, a tablet personal computer, a mobile communication device, a smart phone or other device with computer functions, but it is not used for limit the scope of the invention.

請再參照第1圖所示，舉例而言，該演算及處理單元5可選擇適當配置於一伺服器裝置〔server device〕或一雲端資料庫裝置〔cloud database device〕，且該演算 及處理單元5可選擇具有至少一個或數個數學演算法〔mathematical algorithm〕，且該數學演算法可選自一線性迴歸〔linear regression，REG〕模型、一人工神經網路〔artificial neural network，ANN〕模型或一基因表達規劃〔gene-expression programming，GEP〕模型。 Please refer to Figure 1 again. For example, the calculation and processing unit 5 can be appropriately configured in a server device [server device] or a cloud database device [cloud database device], and the calculation And the processing unit 5 can optionally have at least one or several mathematical algorithms, and the mathematical algorithm can be selected from a linear regression (REG) model, an artificial neural network (ANN) 〕 model or a gene expression programming [gene-expression programming, GEP] model.

第2圖揭示本發明較佳實施例之肺炎遠端檢測方法之流程示意圖。請參照第2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測系統及其方法之執行方式係屬可利用電腦執行之程序步驟〔computer-executable process step〕，其可執行於各種電腦設備，例如：工作站電腦、桌上型電腦、筆記型電腦、平板電腦、行動通訊裝置、智慧型手機或其它具計算機功能之裝置，但其並非用以限定本發明之範圍。 Figure 2 shows a schematic flow chart of a remote detection method for pneumonia according to a preferred embodiment of the present invention. Please refer to Figure 2. For example, the execution method of the pneumonia remote detection system and its method in the preferred embodiment of the present invention is a computer-executable process step, which can be executed In various computer equipment, such as: workstation computers, desktop computers, notebook computers, tablet computers, mobile communication devices, smart phones or other devices with computer functions, this is not intended to limit the scope of the present invention.

請再參照第1及2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測方法包含步驟S1：首先，舉例而言，以適當技術手段〔例如：自動化方式、半自動化方式或手動方式〕於該採樣裝置單元1及其周邊設備上利用至少一個或數個該氣體感測器陣列2偵測一呼吸道氣體樣本11，以便自該呼吸道氣體樣本11獲得至少一個或數個原始數據〔raw data〕及至少一個或數個氣體特徵圖樣〔feature pattern〕。 Please refer to Figures 1 and 2 again. For example, the pneumonia remote detection method in the preferred embodiment of the present invention includes step S1: First, for example, using appropriate technical means [for example: automated method, semi-automatic method or manual method) using at least one or several gas sensor arrays 2 on the sampling device unit 1 and its peripheral equipment to detect a respiratory gas sample 11, so as to obtain at least one or several original samples from the respiratory gas sample 11 Data (raw data) and at least one or several gas characteristic patterns (feature pattern).

第3圖揭示本發明較佳實施例之肺炎遠端檢測系統及其方法獲得原始數據及其氣體特徵圖樣之示意圖。 請參照第3圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測系統及其方法獲得原始數據，如第3圖之上半部所示，並獲得其氣體特徵圖樣，如第3圖之下半部所示，例如：類比轉換數位或特徵極值擷取。 Figure 3 shows a schematic diagram of the pneumonia remote detection system and its method for obtaining raw data and its gas characteristic pattern according to the preferred embodiment of the present invention. Please refer to Figure 3. For example, the pneumonia remote detection system and its method according to the preferred embodiment of the present invention obtain raw data, as shown in the upper half of Figure 3, and obtain its gas characteristic pattern, as shown in the upper half of Figure 3. As shown in the lower half of Figure 3, for example: analog conversion to digital or feature extreme value extraction.

請再參照第1及2圖所示，舉例而言，該呼吸道氣體樣本11包含一揮發性有機物氣體〔volatile organic compound，VOC〕、一無機物氣體或其組合體之代謝物，而該揮發性有機物氣體包含飽和碳氫化合物〔例如：乙烷、戊烷、乙醛或其它烴類〕、不飽和碳氫化合物〔例如：異戊二烯或其它烴類〕、丙酮〔脂質代謝引起〕、乙硫醇〔蛋氨酸代謝不完全引起〕、二甲硫醚〔蛋氨酸代謝不完全引起〕、二甲胺〔肺部組織受損引起〕及氨〔肺部組織受損引起〕。 Please refer to Figures 1 and 2 again. For example, the respiratory gas sample 11 contains a volatile organic compound gas. compound, VOC], a metabolite of an inorganic gas or a combination thereof, and the volatile organic gas contains saturated hydrocarbons [for example: ethane, pentane, acetaldehyde or other hydrocarbons], unsaturated hydrocarbons [ For example: isoprene or other hydrocarbons], acetone [caused by lipid metabolism], ethyl mercaptan [caused by incomplete methionine metabolism], dimethyl sulfide [caused by incomplete methionine metabolism], dimethylamine [caused by lung tissue damage] [caused by damage to lung tissue] and ammonia [caused by damage to lung tissue].

請再參照第1及2圖所示，舉例而言，將該呼吸道氣體樣本11之原始數據及氣體特徵圖樣可選擇於一雲端裝置或其周邊設備適當進行數據處理〔例如：極值演算特徵值、特徵擷取或其它統計處理〕。 Please refer to Figures 1 and 2 again. For example, the original data and gas characteristic pattern of the respiratory gas sample 11 can be selected to be appropriately processed in a cloud device or its peripheral equipment [for example: extreme value calculation characteristic value] , feature extraction or other statistical processing].

第4圖揭示本發明較佳實施例之肺炎遠端檢測系統及其方法利用氣體感測器陣列量測獲得呼吸道氣體樣本及其氣體特徵資料之示意圖。請參照第1及4圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測系統及其方法自該呼吸道氣體樣本11獲得該氣體特徵圖樣，並利用該氣體特徵圖樣取出其特徵值〔eigenvalue〕進行一機器學習作業，以便獲得一氣體特徵值模型。 Figure 4 shows a schematic diagram of a pneumonia remote detection system and method using a gas sensor array to measure and obtain respiratory gas samples and gas characteristic data according to a preferred embodiment of the present invention. Please refer to Figures 1 and 4. For example, the pneumonia remote detection system and its method according to the preferred embodiment of the present invention obtain the gas characteristic pattern from the respiratory gas sample 11, and use the gas characteristic pattern to extract its characteristics. Value [eigenvalue] performs a machine learning operation to obtain a gas eigenvalue model.

請再參照第1及2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測方法包含步驟S2：接著，舉例而言，以適當技術手段〔例如：自動化方式、半自動化方式或手動方式〕於該採樣裝置單元1及其周邊設備上利用該影像攝取單元3適當攝取一分泌物液體樣本12，以便獲得一分泌物液體顯微影像〔例如：影像〔m×n〕或其它適當擷取影像方塊〕。 Please refer to Figures 1 and 2 again. For example, the pneumonia remote detection method in the preferred embodiment of the present invention includes step S2: Next, for example, using appropriate technical means [for example: automated method, semi-automatic method Or manually) use the image capture unit 3 on the sampling device unit 1 and its peripheral equipment to properly capture a secretion liquid sample 12, so as to obtain a secretion liquid microscopic image [for example: image [ m × n ] or other Capture image squares appropriately].

請再參照第1及2圖所示，舉例而言，該分泌物液體樣本12包含一唾液〔例如：傳染性氣溶膠〕、一痰液、一鼻咽液或其任意組合體，且該分泌物液體樣本12適當取自該採樣裝置單元1及其周邊設備。 Please refer to Figures 1 and 2 again. For example, the secretion liquid sample 12 includes a saliva (for example: infectious aerosol), a sputum, a nasopharyngeal fluid or any combination thereof, and the secretion The physical liquid sample 12 is suitably taken from the sampling device unit 1 and its peripheral equipment.

第5圖揭示本發明較佳實施例之肺炎遠端檢測系統及其方法獲得顯微影像之示意圖。請參照第1及5圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測系統及其方法自該分泌物液體樣本12獲得該分泌物液體顯微影像，且該分泌物液體顯微影像顯示至少一異常特徵〔例如：冠狀病毒輪廓〔SARS、MERS、Covid-19〕或其它特徵〕，如第5圖之箭頭所示。 Figure 5 shows a schematic diagram of the pneumonia remote detection system and its method for obtaining microscopic images according to the preferred embodiment of the present invention. Please refer to Figures 1 and 5. For example, the pneumonia remote detection system and its method in the preferred embodiment of the present invention obtain the secretion liquid microscopic image from the secretion liquid sample 12, and the secretion liquid The microscopic image shows at least one abnormal feature (for example: coronavirus outline [SARS, MERS, Covid-19] or other features), as shown by the arrow in Figure 5.

請再參照第5圖所示，舉例而言，本發明較佳實施例之該分泌物液體樣本12可選擇添加一螢光染料，以便該分泌物液體樣本12可顯示至少一螢光影像，以便該分泌物液體顯微影像可顯示至少一螢光特徵或至少一異常螢光特徵〔例如：螢光染色冠狀病毒輪廓或其它螢光染色特徵〕。 Please refer to Figure 5 again. For example, the secretion liquid sample 12 in the preferred embodiment of the present invention can optionally add a fluorescent dye, so that the secretion liquid sample 12 can display at least one fluorescent image, so that The microscopic image of the secretion liquid can show at least one fluorescent feature or at least one abnormal fluorescent feature [for example: fluorescent staining coronavirus outline or other fluorescent staining features].

第6圖揭示本發明較佳實施例之肺炎遠端檢測系統及其方法採用通訊架構之示意圖。請參照第6圖所示，舉例而言，將該採樣裝置單元1、氣體感測器陣列2、影像攝取單元3〔包含螢光染色藥劑或其它配件〕及資料傳輸單元4〔包含網路設備或其它配置設備〕可選擇配置於一遠端位置〔例如：偏鄉地區、工業區、住宅區或其它區域〕，以便達成節省檢驗設備資源、節省檢驗人力資源及提升檢驗效率。 Figure 6 shows a schematic diagram of the communication architecture adopted by the pneumonia remote detection system and its method according to the preferred embodiment of the present invention. Please refer to Figure 6, for example, the sampling device unit 1, the gas sensor array 2, the image capture unit 3 (including fluorescent dyeing agents or other accessories) and the data transmission unit 4 (including network equipment) Or other configuration equipment) can be optionally configured in a remote location (such as rural areas, industrial areas, residential areas or other areas) to save inspection equipment resources, save inspection human resources and improve inspection efficiency.

請再參照第6圖所示，舉例而言，將該演算及處理單元5及其周邊設備〔例如：判斷演算模型、人工智慧軟體、雲端資料庫或其它〕配置於一近端位置〔例如：醫療檢驗中心或其它機關單位〕，並經由網際網路或其它通訊方式連結傳輸於該資料傳輸單元4。 Please refer to Figure 6 again. For example, the calculation and processing unit 5 and its peripheral devices (such as: judgment calculation model, artificial intelligence software, cloud database or others) are configured in a near-end location (for example: Medical examination center or other institutional units], and transmit it to the data transmission unit 4 through the Internet or other communication methods.

請再參照第1及2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測方法包含步驟S3：接著，舉例而言，以適當技術手段〔例如：自動化方式、半自動化方 式或手動方式〕將該分泌物液體顯微影像進行影像處理，以便獲得至少一個或數個已處理影像特徵，例如：去除雜訊、二值化、投影、特徵擷取或其它處理。 Please refer to Figures 1 and 2 again. For example, the pneumonia remote detection method in the preferred embodiment of the present invention includes step S3: Next, for example, using appropriate technical means [for example: automated methods, semi-automated methods [Form or manual method] perform image processing on the secretion liquid microscopic image to obtain at least one or several processed image features, such as: noise removal, binarization, projection, feature extraction or other processing.

請再參照第1及2圖所示，舉例而言，將該分泌物液體顯微影像可選擇於一雲端裝置或其周邊設備適當進行影像處理〔例如：去除雜訊、二值化、投影、特徵擷取或其它處理〕。 Please refer to Figures 1 and 2 again. For example, the secretion liquid microscopic image can be appropriately processed on a cloud device or its peripheral equipment (for example: noise removal, binarization, projection, Feature extraction or other processing].

請再參照第1及2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測方法包含步驟S4：接著，舉例而言，以適當技術手段〔例如：自動化方式、半自動化方式或手動方式〕利用該氣體特徵圖樣及已處理影像特徵進行一機器學習作業，以便獲得至少一個或數個氣體特徵與影像特徵相關性〔correlation〕。 Please refer to Figures 1 and 2 again. For example, the pneumonia remote detection method in the preferred embodiment of the present invention includes step S4: Next, for example, using appropriate technical means [for example: automated method, semi-automatic method or manual method) using the gas feature pattern and the processed image features to perform a machine learning operation in order to obtain the correlation between at least one or several gas features and image features.

請再參照第1及2圖所示，舉例而言，將該氣體特徵圖樣及已處理影像特徵之機器學習作業可選擇於一雲端裝置或其周邊設備適當進行影像處理〔例如：機器學習處理、線性迴歸處理或其它統計演算處理〕。 Please refer to Figures 1 and 2 again. For example, the machine learning operation of the gas characteristic pattern and the processed image characteristics can be appropriately performed on a cloud device or its peripheral equipment for appropriate image processing [for example: machine learning processing, Linear regression processing or other statistical calculation processing].

請再參照第1及2圖所示，舉例而言，本發明較佳實施例之肺炎遠端檢測方法包含步驟S5：接著，舉例而言，以適當技術手段〔例如：自動化方式、半自動化方式或手動方式〕利用該氣體特徵與影像特徵相關性判斷產生一肺部呼出氣體異常狀態結果〔即陰性或陽性〕，以降低產生假陰性或假陽性之機率。 Please refer to Figures 1 and 2 again. For example, the pneumonia remote detection method in the preferred embodiment of the present invention includes step S5: Next, for example, using appropriate technical means [for example: automated method, semi-automatic method Or manual method) use the correlation between the gas characteristics and the image characteristics to determine the abnormal status result of exhaled lung gas (i.e. negative or positive), so as to reduce the probability of false negative or false positive.

請再參照第1及2圖所示，舉例而言，將該該氣體特徵與影像特徵相關性及其判斷所產生該肺部呼出氣體異常狀態結果可選擇於一雲端裝置或其周邊設備適當進行影像處理〔例如：相關性判斷處理或其它判斷演算處理〕。 Please refer to Figures 1 and 2 again. For example, the correlation between the gas characteristics and the image characteristics and the abnormal status of the exhaled lung gas generated by the judgment can be appropriately performed on a cloud device or its peripheral equipment. Image processing [for example: correlation judgment processing or other judgment calculation processing].

第7圖揭示本發明第二較佳實施例之肺炎遠端 檢測系統之示意圖。請參照第7圖所示，相對於第一實施例，本發明第二較佳實施例之肺炎遠端檢測系統另包含一體溫感測器單元6，且該體溫感測器單元6用以量測一體溫資料〔例如：受測者或受測病患〕。 Figure 7 shows the distal end of pneumonia according to the second preferred embodiment of the present invention. Schematic diagram of the detection system. Please refer to Figure 7. Compared with the first embodiment, the pneumonia remote detection system of the second preferred embodiment of the present invention further includes a body temperature sensor unit 6, and the body temperature sensor unit 6 is used to measure Measure body temperature information [for example: the subject or the patient being tested].

請再參照第7圖所示，舉例而言，該體溫感測器單元6可選自一長期監控體溫感測器單元或一遠端監控體溫感測器單元，而該體溫感測器單元6可選自一貼紙體溫計、一奶嘴體溫計、一紅外線體溫感測器、一紅外線影像體溫偵測器或其任意組合。 Please refer to Figure 7 again. For example, the body temperature sensor unit 6 can be selected from a long-term monitoring body temperature sensor unit or a remote monitoring body temperature sensor unit, and the body temperature sensor unit 6 You can choose from a sticker thermometer, a pacifier thermometer, an infrared body temperature sensor, an infrared imaging body temperature detector or any combination thereof.

請再參照第7圖所示，舉例而言，該體溫資料為一正常體溫資料及一發燒體溫資料〔例如：個別人體發燒統計溫度資料或其正常體溫資料〕，且配合該氣體特徵與影像特徵相關性及其相關該肺部呼出氣體異常狀態結果，以便利用該體溫資料判斷是否為一有症狀感染或一無症狀感染。 Please refer to Figure 7 again. For example, the body temperature data is a normal body temperature data and a fever body temperature data [for example: individual body fever statistical temperature data or its normal body temperature data], and matches the gas characteristics and image characteristics. Correlation and related results of the abnormal state of exhaled air from the lungs, in order to use the body temperature data to determine whether it is a symptomatic infection or an asymptomatic infection.

請再參照第7圖所示，舉例而言，將該該氣體特徵與影像特徵相關性及其判斷所產生該肺部呼出氣體異常狀態結果結合該體溫資料可選擇於一雲端裝置或其周邊設備適當進行影像處理〔例如：相關性判斷處理或其它判斷演算處理〕。 Please refer to Figure 7 again. For example, the correlation between the gas characteristics and the image characteristics and the result of judging the abnormal status of the exhaled lung gas combined with the body temperature data can be selected in a cloud device or its peripheral equipment. Perform image processing appropriately (for example: correlation judgment processing or other judgment calculation processing).

前述較佳實施例僅舉例說明本發明及其技術特徵，該實施例之技術仍可適當進行各種實質等效修飾及/或替換方式予以實施；因此，本發明之權利範圍須視後附申請專利範圍所界定之範圍為準。本案著作權限制使用於中華民國專利申請用途。 The foregoing preferred embodiments only illustrate the present invention and its technical features. The technology of this embodiment can still be appropriately implemented with various substantially equivalent modifications and/or substitutions; therefore, the scope of rights of the present invention shall depend on the appended patent application. The scope defined shall prevail. The copyright in this case is restricted to use for patent applications in the Republic of China.

1:採樣裝置單元 1: Sampling device unit

11:呼吸道氣體樣本 11:Respiratory gas sample

12:分泌物液體樣本 12: Secretion liquid sample

2:氣體感測器陣列 2: Gas sensor array

3:影像攝取單元 3:Image capture unit

4:資料傳輸單元 4: Data transmission unit

5:演算及處理單元 5: Calculation and processing unit