201034629 六、發明說明： 【發明所屬之技術領域】 本發明係有關於痰音偵 藉由多通道呼吸音量測裝置 模組及警示裝置之整合系統 關處置與衛教工作。 【先前技術】 测、辨識與衛教系統，為一種 、圖形化人機介面、痰音辨識 ，以偵測辨識異常痰音進行相 ❹ ❹ 呼吸是維持生命最重要的.動作。人體細胞需要 給才能維持正常的運作，透過呼❹統不斷的進氣血 互換才能轉人體各部位的機能。#呼吸系統產 2礙時，不僅會影響到其它器官的運作，嚴重時可能會 二生命安全。人員在替病患做呼吸系統評估檢查 寺：除了要有專業的醫療知識外，還需要依靠一些醫療； 辅助診斷’一般最常見的為機械式聽診器，其它較為 ，密的醫療器具則有x_ray、咖、CT、謝等。但在一般 養2或居豕¥境，並未設置這些昂貴的精密醫療儀 :’如果可以先進行基本的聽診檢查，不但減少不必要的 麻煩，也可以避免醫療資源的浪費。 目“檢查是一種非侵入式、方便也是最常見的檢查項 聽病患與醫師直接做面對面的會診，醫師可以透過 >曰二°旱握病患的身體健康狀況，而患者也能藉由會診中 安慰。呼吸系統的聽診的目較在於空氣在氣 泡區氣s的流通情形找出擴張或阻塞部位、評估週邊肺 品K張的凊形、肋膜腔的狀況，並確認這些呼吸音是源 201034629 υ::個部位，這對於在肺部疾病的診 前先進國家的醫療資源已相當充足，但: 的技術，因此義器過度的依賴而忽略了最基本聽診 鹼士处“ 員必須擁有專業的知識以及豐富的經 ο 病徵’但是文獻指出，有許多醫護人員 員從判斷出病徵，其原因料於醫護人 致聽i技又術退1並沒有持續進行聽診技能的教育訓練而導 正常的氣管支氣管每日約可產 :自健;:人或一般病患若是，及道中產生黏 說：為床或氣管插管之病患來 —_:===…易造 這些病患清除肺部分泌物’維持病患的呼吸=暢定= ο 。隻理人貝必須猎由聽診器輔助， 位’給予適當的姿勢擺位和胸部 冑阻塞的部 管或接近喉嚷的部位，然後利用二引流到氣 慢慢將瘦排出。然而，在長期照護機構或 於直酱嗜畑，丄π σ . 1 π人豕照護中，由 π專業濩理人力不足，大都由非專辈 顧病人，因此無法正確判斷疫阻=服;或;屬來照 的照護工作。另外，文獻上指出護 進仃有效排痰 事項並不清楚，也未考慮其安全性，通常==由痰的注意 依常規每二小時抽痰一次，且執行抽痰時:護 4 201034629 ’個人決策模式及抽痰指標的影響报大，過度與不適 痰’可能會導致合併症的產生，如氣管損傷、支氣管=攀、 缺氧、心律不整、低血壓、心跳停止、死亡。 立目前臨床醫師可使用機械式或電子式聽診器來聽診肺 曰，雖然:目前㈣肺音研究包括單通道或多通道肺音榻取 系統、肺音訊號處理等相關研究，但是在臨床上 師自行判斷肺音是否正常。如果夠將病患的肺音完整記= 下來並且傳送至電腦端進行數位訊號處理分析，自動判 異常呼吸音’輔助醫護人員做診斷與正確㈣吸照護衛教 知識’這樣-來可以讓醫護人員對於病患的病情更能夠有 ^的掌握，不但可以減低病患進行抽痰時的風險，也能夠 助濩理人員對於抽痰時機的掌握，並教導看護工與家庭 照護者正確的照護知識。 Ο 八習用呼吸音聽診系統，係應用於非侵入式之聽診診斷 析其採用集音放大等原理，由聲音的種類與層次變化 作相關的診斷判斷，但必須佐以醫療診斷經驗、複雜的系 又備與繁雜的操作過程，始能精準地獲致正確的資訊， 2較直覺與單純的操作模式與受專業知識與技術限制之 ’學習與應用層次方面效果不彰，在呼吸音聽 技術開=產業應用上缺乏專屬性。 此 β目則數位化電子聽診器的發展技術已成熟，總類繁多 疋般量測型、遠距離傳輸型、多通道監測型等，中華 民國專利公告就字第〇834〇1即時多工電子聽診協判系統， 由至多組聽診量測系統組成可量測多種生理音、無線 5 201034629 傳輸電路、PC端訊號分析、資料 129428η -τ^ -4- 、 。中華民國專利公告 ^ _28G可攜式氣喘肺音監喝統，係利用 糸統、無線傳輸電路、pc氣喘音分析 〜 述技術均為常見的肺 所、、且成。上 炎、主p火楚統其受測病徵為氣喘、肺 插其吸相病’但對於長期臥財床、氣管 插吕病患其肺部組織常有黏液孔吕 交总、生々γ 大寻刀，必物堆積在肺中， 谷易w成呼吸道嚴重阻塞而窒息死亡。 ❹ 上述習用呼吸音聽診系統之最大的缺點在於： 1.必須具備相關醫療照護之診斷經 2·複雜㈣統設備在操作過財顯得繁複與困難。 3.擷取及分析所得之相關資訊缺乏完整性。、 、、前述所提及關於習用之啤吸音聽診系統，儘管能夠達 成在聽診診斷分析方面所應具備一般基本要求盘成效，作 ί實際應㈣之直覺性與單純的操作模式等產業應用專屬 ❹性^皆存在諸多缺點與不足的情況下，無法發揮更具體 之產業應用性。 綜上所述，由於習用之呼吸音聽診系統，存在上述之 缺失與不足，基於產業進步之未來趨勢前提下，實在有必 要提出具體的改善方案，以符合產業進步之所需，'更進一 步提供業界更多的技術性選擇。 【發明内容】 目前電子電路的發展和數位訊號處理的技術已有相當 成熟，本發明之痰音偵測、辨識與衛教系統，可提供教師 201034629 ，輔助教學使用以及提供學生、臨床照護者對於呼吸音聽診 之教育訓練，進而提昇醫療照護人員的聽診能力和醫療照 護的品質，且更可以對於抽痰時機的掌握，不但能防止患 者因呼吸道嚴重阻塞導致死亡，並且可以減低病患進行抽 痰時的風險，也能夠教導看護工與家庭照護者正確的呼吸 照護常識。 根據歐洲經濟政策委員會（ElJ Economic Commit tee)在2001年的報告中指出，至2〇2〇年，曰本老 年人口約佔全人口的42%，歐洲老年人口約佔全人口的36 % ’美國老年人口約佔全人口的26%。人口老化已成為全 世界現象，我國亦不例外，根據經濟部的分析，95年國内 65歲以上老人高達226萬人，共212萬（94%)的老人需仰 '賴社區或居家式照護。從老年人平常生活起居健康照護乃 至於醫療照護的各種需求，其中薇藏了魔大的商機，為相 關產業、創造相當可觀的發展契機。 〇 經濟部次長謝發達於96年3月1日指出：『全球健康照護 產業2009年產值將高達新台幣1〇兆元，政府未來將透過醫 療E化及醫療服務國際化的發展，出口國内醫療昭 搶佔全球照護產業商機』。在健康照護方面，呼吸監控為其 :之一 ’隨著健康照護的主軸從機構式服務像社區式、’居 家式服務延伸，商機的觸角也將深入至每個高齡人口的家 庭及所在的社區中。本專利不僅針對老人的啤吸監控肺中 積痰的發生，避免長期臥病在床、肺部慢性病、氣管插管 之病患因為黏液分泌過多所造成之呼吸道嚴重阻塞而窒息 201034629201034629 VI. Description of the Invention: [Technical Field of the Invention] The present invention relates to an integrated system for handling and teaching of voice detection by a multi-channel breathing volume measuring device module and a warning device. [Prior Art] The measurement, identification and education system is a kind of graphical human-machine interface, voice recognition, and detection of abnormal voices. 呼吸 Breathing is the most important action to sustain life. The human body cells need to be able to maintain normal operation, and the functions of various parts of the human body can be transferred through the continuous exchange of blood in the respiratory system. #呼吸系统的产生2, not only will affect the operation of other organs, in case of serious life may be second. The staff is doing a respiratory assessment for the patient. In addition to professional medical knowledge, some medical treatments are needed. Auxiliary diagnosis is generally the most common mechanical stethoscope. Other relatively dense medical devices have x_ray. Coffee, CT, Xie, etc. However, these expensive precision medical instruments are not set up in general or in the environment: 'If basic auscultation can be performed first, not only unnecessary troubles can be reduced, but also waste of medical resources can be avoided. "The examination is a non-invasive, convenient and most common examination item. The patient can directly face-to-face consultation with the doctor. The doctor can grasp the health condition of the patient through the 曰2° drought, and the patient can also Consolation in the consultation. The auscultation of the respiratory system is based on the circulation of air in the bubble area to find the expansion or obstruction, evaluate the shape of the sacral cavity of the peripheral lungs, and the condition of the pleural cavity, and confirm that these breath sounds are the source. 201034629 υ:: A part of the medical resources in the advanced countries before the diagnosis of lung diseases is quite sufficient, but: the technology, therefore the excessive dependence of the righteous neglected the most basic auscultation The knowledge and the richness of the disease's disease's but the literature points out that there are many medical staff members who judge the symptoms. The reason is that the medical staff listens to the i skills and retreats 1 and does not continue the education training of the auscultation skills. Tracheobronchial can be produced daily: self-healing;: If the person or the general patient is, and the sticky in the road is said: the patient who is bed or tracheal intubation comes __:===...easy to make these patients In addition to pulmonary secretions' to maintain a given patient respiratory smooth = = ο. Only the person must be assisted by a stethoscope, and the position should be given to the appropriate position and the part of the chest that is blocked or close to the throat, and then the second drainage can be used to slowly drain the skin. However, in the long-term care institutions or in the straight sputum, 丄π σ. 1 π human care, π professional caremanship is insufficient, mostly by non-professional patients, so it is impossible to correctly determine the disease resistance = service; or The care work that comes with the photo. In addition, the literature points out that the effective drainage of sputum sputum is not clear, and does not consider its safety, usually == 痰 每 每 依 依 依 , , , , , , , , , 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 The influence of decision-making patterns and convulsions is large, and excessive and uncomfortable 'may lead to complications such as tracheal injury, bronchi = climbing, hypoxia, arrhythmia, hypotension, heartbeat, and death. Currently, clinicians can use mechanical or electronic stethoscopes to auscultate pulmonary sputum, although: (4) lung sound research includes single-channel or multi-channel lung sound treading system, lung sound signal processing and other related research, but in clinical practice Determine if the lung sound is normal. If it is enough to completely record the patient's lung sounds and send them to the computer for digital signal processing analysis, automatically determine the abnormal breath sounds 'assisted medical staff to do the diagnosis and correct (four) illuminating the guardian knowledge 'this way - to allow medical staff It is better to have a good grasp of the patient's condition, not only to reduce the risk of convulsions, but also to help the caregiver grasp the timing of convulsions and to teach the caregivers and home caregivers the correct care knowledge.八 Eight-study breath sound auscultation system is applied to non-invasive auscultation diagnosis and analysis. It uses the principles of sound amplification and so on. It is related to the diagnosis and judgment of the type and level of sound, but it must be accompanied by medical diagnosis experience and complex system. With the complicated and complicated operation process, the accurate information can be accurately obtained. 2 The more intuitive and simple operation mode and the limitation of the learning and application level by the professional knowledge and technical restrictions, the breathing sound listening technology is open = There is a lack of specificity in industrial applications. The development technology of this digital target electronic stethoscope has matured, and there are many kinds of measurement types, long-distance transmission type, multi-channel monitoring type, etc., the Republic of China patent announcement on the word 〇834〇1 instant multiplex electronic auscultation Co-judging system, consisting of up to a group of auscultation measurement systems can measure a variety of physiological sounds, wireless 5 201034629 transmission circuit, PC-side signal analysis, data 129428η -τ^ -4-,. Republic of China Patent Announcement ^ _28G Portable asthmatic lung sound monitoring system, using the system, wireless transmission circuit, pc asthmatic analysis ~ The techniques are common lungs, and into. Shangyan, the main p fire Chu Tong Tong measured the symptoms for asthma, lung insertion of its aspiration disease 'but for the long-term bed, tracheal disease, the lung tissue often has a mucus hole Lujia total, oyster γ big search Knife, must accumulate in the lungs, Gu Yi w into the respiratory tract is severely blocked and suffocated to death.最大 The biggest disadvantages of the above-mentioned conventional respiratory sound auscultation system are: 1. It must be diagnosed with relevant medical care. 2. Complex (four) system equipment is complicated and difficult to operate. 3. The information obtained from the analysis and analysis lacks completeness. The above-mentioned beer aspiration auscultation system mentioned in the above mentioned, although it can achieve the general basic requirements for the diagnosis of auscultation diagnosis, the practical application of (4) intuition and simple operation mode is exclusive. In the case that there are many shortcomings and deficiencies in the sex ^, it is impossible to exert more specific industrial applicability. In summary, due to the above-mentioned shortcomings and shortcomings of the respiratory audio auscultation system, it is necessary to propose specific improvement plans based on the future trend of industrial progress, in order to meet the needs of industrial progress, and further provide More technical choices in the industry. SUMMARY OF THE INVENTION At present, the development of electronic circuits and the technology of digital signal processing are quite mature. The voice detection, identification and teaching system of the present invention can provide teachers 201034629, assisting teaching use and providing students and clinical caregivers with The educational training of respiratory auscultation improves the auscultation ability and medical care quality of medical caregivers, and can better control the timing of convulsions, not only prevent patients from dying due to severe obstruction of the respiratory tract, but also reduce the convulsions of patients. The risk of time can also teach the correct breathing care common sense of caregivers and home caregivers. According to the 2001 report of the European Economic Policy Committee (ElJ Economic Commit tee), by the year of 2002, the elderly population accounted for 42% of the total population, and the elderly in Europe accounted for 36% of the total population. The elderly population accounts for about 26% of the entire population. The ageing of the population has become a worldwide phenomenon. China is no exception. According to the analysis of the Ministry of Economic Affairs, in the past 95 years, there were 2.26 million elderly people over the age of 65, and a total of 2.12 million (94%) of the elderly need to rely on community or home care. . From the ordinary life and health care of the elderly, it is the various needs of medical care. Among them, Wei has hidden the business opportunities of the Magic City and created considerable opportunities for related industries. 〇 Economic Development Minister Xie Development pointed out on March 1, 1996: "The global health care industry will have a production value of NT$1 trillion in 2009. The government will export domestically through the development of medical E and medical services. Medical Zhao grabs the business opportunities of the global care industry. In terms of health care, respiratory monitoring is one of them: 'With the main axis of health care from institutional services like community-based, 'home-based services, the business opportunities will reach into the families and communities of every elderly population. in. This patent not only monitors the occurrence of accumulated sputum in the lungs of the elderly, but also avoids the suffocation of the respiratory tract caused by excessive mucus secretion in patients with chronic diseases, chronic diseases of the lungs, and tracheal intubation. 201034629
•，亡同時也提供照護者適當的呼吸照護衛教知識，未來 在健康照護產章部份·，—T- IQ 業包括. Μ ^ 了^咼市场的佔有率。健康照護產 1. 居家呼吸照護與衛生教育。 2. 長期照護機構之呼吸照護。 3 ·醫療院所之呼吸照護。 每天護理人力短缺的問題-直存在，而護理人員 〇逐=广：事項包括：發藥、打針、紀錄工作曰誌，益法 = 二痰、抽痰、翻身、擦洗，所以目前傾向 .镬父家二ΓΓ上反而會增加其危險性，因為無論是看 〇 -技能，像是基本的聽玲能力2的4護人貝该有的醫療 位引流知識。因此在執行正確的妥 工或家庭照護者進行衛教教學，=他=要對看護 照護技術。傳統的聽診器中雖然可f 確的 是必須擁有專業的聽診 月楚的I、見肺音，但 果能利用數位1味沾步 才有辦法判斷出病徵，在此如 -些科學量二 =1 支術將呼吸音數位化，並且透過 是做基本聽蜂檢杳戈3廂辅助醫護人員判斷病徵’無論 故，本發:係 單純的操作模式受限等方 系統在直覺性與 術等方面受到限制之問題，點’以及在實用化技 直覺層次與使用效能等方面之表現為=== 8 201034629 :::二用上之便利性等方面之提升，以達成所應具備之 it、ί之聽料斷分析之―般功料，並使其兼具痰音 要^。識與術教技術開發設備產業應用性之實際發展與 ❹ ❹ 立店=以不""由主客觀條件觀之，利用具有多點彳貞測之痰 μ姑1 *辨識與俯教系'统’在國内外專利中目前確實無相 ==用於呼吸音診斷領域，且根據内政部社會司於 :006年的統計，我國65歲以上老年人口佔總人口的9 87 超過聯合國兩齡化社會的標準，並估計於別25年將 、％ 乂上，而根據經濟合作發展組織（〇ecd)的研究， :歲以上的老人每人平均醫療支出，約為π歲以下者的四 /、、;'而根據美國退休人士協會的研究顯示，當中高齡 3 =醫療崎f求時，約以成的人都偏好尋求在家 ^ '—'^服務’顯示居家醫療照護已經成為全球的重 勢’因此具備市場無可取代之技術之優勢’極適合應 =於痰音偵測、辨識與衛教等設備市場，勢必可以帶來疫 曰偵測：辨識與衛教技術開發設備之生產與設計製造產業 相關市場之莫大商機。 方二達成上述目的及功能’其具體採行的技術手段及 一種痰音偵測、辨識與衛教系統，包括： 立至乂呼吸曰夏測裝置，包括機械式聽診集音盤及聲 二感測器’聲音感測器設於機械式聽診集音盤内，用以轉 換所收集之呼吸音為呼吸音類比訊號。 9 201034629 路用⑪合有類比電路、類比多工器及數位電 路’用==音_職轉換為呼吸音數位訊號。 一儲存及分析呼吸音數位訊號。 μ賴心’用㈣識料音數位訊號之瘦音訊 號0 運异單7L ’與痰音辨識模組電性連接，用以接收並 處理痰音訊號，並產生一處理結果。 stf裝置’與運算單元電性連接1以根據運算單 ◎元之處理結果而不動作或產生—相對應之警示訊息。 上述類比電路包含前級放大器、帶拒滤波器、帶通滤 波器及後級放大器。 上述類比夕工器係整合呼吸音量測裝置之呼吸音類比 訊號進入數位電路。 上述數位電路包含類比數位轉換器、微控制器及數位 傳輸介面。 上述痰音辨識模組包含痰音辨識參數萃取及痰音辨識 模梨。 上述運算單元另電性連接一顯示裝置，該顯示裝置連 接多媒體衛教影音模組、呼吸音訊號模組及病歷資料庫， 立該顯示裝置設有一圖形化人機介面。 上述多媒體衛教影音模組，包含多媒體聽診教學之數 位訊息及多媒體呼吸照護衛教之數位訊息。 上述呼吸音訊號模組包含一呼吸音訊號擷取介面及— 啤吸音訊號分析介面。 201034629 - 上述運算單疋根據痰音辨識模組之痰音訊號產生一處 理結果’並根據該處理結果使多媒體衛教影音模組、呼吸 音訊號模組及病歷資料庫中之至少一種模組（或資料庫） 輸出相對應訊息。 ' 本發明之具體特點與功效在於： 1. 藉由本系統偵測出肺臟中產生黏液（痰）的部位，進而 播放出呼吸照護相關衛教影片，引導照護者針對產生黏液 o(痰）的部位進彳丁引流的動作’協助患者清除呼吸道的分泌 物’以維持呼吸道的暢通。 2. 可提供照護人員對於抽痰時機的掌握’不但能防止 患者因呼吸道嚴重阻塞導致死亡，1且可以減低病患進行 抽痰時的風險。。 3. 可提供教師辅助教學使用以及提供學生、臨床照護 5對於呼吸音聽診之教育訓練，也能夠教導看護工與家庭 $遵者正確的呼吸照護常識，進而提昇醫療照護人員的聽 〇診能力和醫療照護的品質。 一 【實施方式】 明參閱第一圖，為本發明實施例之系統架構流程圖， 其中，一種痰音偵測、辨識與衛教系統，包括： 少A呼吸音置測裝置（丨）、轉換電路（2 )、分析裝置（3 ) ^通常為電腦〕、痰音辨識模組⑷、運算單元（5)以及 :不裝置（6)，其中，呼吸音量測裝置〇)接收來自受測 . ^8)之呼吸音（81) ’再送入轉換電路（2)作轉換，轉 奐疋成後以分析震置（3)儲存及分析結果，同時將結果傳 11 201034629 -送至痰音辨識參數萃取（41)以及痰音辨識模型（⑵ 之痰：辨，莫組（4)，以辨識受測者⑴之啤吸音（81) 之痰曰狀恕’再由與痰音辨識模組（4 )電性連接之運算單 元（5一）接收，處理後產生一處理結果。運算單元⑴則 與吕不裂置（6)電性連接，用以根據運算單元⑴之處 理結果=動作或產生一相對應之警示訊息。 運算單兀（5)另電性連接顯示裝置（7)，顯示裝置 ()連接夕媒體衛教影音模組（72)、呼吸音訊號模組⑺） ^歷資料庫（74)，且該顯示裝置⑺設有圖形化人機 ”面（71);其中多媒體衛教影音模組（72 )，包含多媒體 I ^教予（721 )之數位訊息及多媒體呼吸照護衛教（722 ) 之數位訊息’而呼吸音訊號模組（73)則包含呼吸音訊號 操取介^73U及呼吸音訊號分析介面（732 )。 運算單元（5)根據痰音辨識模組⑷之痰音訊號產 生一處理結果，並根據該處理結果使多媒體衛教影音模組 〇 ( 72)、呼吸音訊號模組（73)以及病歷資料庫（μ)中之 至少-種模組〔或資料庫〕輸出相對應訊息；亦即受測者 之痰音辨識結果〔不論正常或異常〕載入病歷資料庫（⑷ 作記錄，若發現異常痰音，則警示裝置⑷會產生擎報訊 息以警示相關醫護或照護人員注意，同時會載入多媒體衛 教々q模組（72 )作為教育訓練之實際參考資訊。 請參閱第二圖，4本發明實施例之呼吸音量測 '路架構圖，其中，呼吸音量測I置⑴〔另請同時^閱第 12 201034629 一圖〕均各包含内置聲音感測器（111)之機械式聽診集音 盤（Π )，複數通道〔通道1、2…〕均各包獨立之類比電路 (21) ’各類比電路（21)内含前級放大器（211)、帶拒滤 波器（212)、帶通濾波器（213)以及後級放大器（214); 各複數通道之類比電路（21)由類比多工器（22)整合後 進入數位電路（23)，數位電路（23)包含類比數位轉換器 (231)、微控制器（232 )以及數位傳輸介面（233 )，最後 ❹輸出數位訊號。機械式聽診集音盤（u)，係置於人體肺臟 各部位〔每一部位即連結一呼吸音量測裝置（丨）及類比電 路以形成獨立之訊號通道〕以收集呼吸音（81)，機械式聽 診集音盤（11)設有聲音感測器（m )，係轉換呼吸音（81) 為呼吸音類比訊號（811 )〔係為原始之類比訊號〕，經由前 級放大器（211)、帶拒濾波器（212)、帶通濾波器（213) 及後級放大器（214)等之類比訊號處理程序後，各通道之 呼吸音類比訊號（811)經類比多工器（22)進入數位電路 ❹（23 )’再經數位電路（23)之類比數位轉換器（231 )、微 控制器（ 232 )以及數位傳輪介面（233 )後成為呼吸音數 位訊號（812)輸出，以連結分析裝置（3)〔通常為電腦〕 以及痰音辨識模組（4)〔另請參閱第一圖〕。 而類比電路（21)類比多工器（22)以及數位電路（23) 共同組成如第一圖所示之轉換電路（2)，亦即經由轉換電 路之轉換，可以將原始經機械式聽診集音盤（1 1 )轉換呼 吸音（81)而成之呼吸音類比訊號（811)，進一步轉換成 為呼吸音類比訊號（812)，以方便後續之分析處理。 13 201034629 - 在類比電路（21)方面，由於原始呼吸音（μ)訊號 振幅非常的微小，為了正確量測到呼吸音（8 1 )訊號，本 系統採用儀表放大器作為前級放大器（211 )，主要由於其 南共模拒斥比值（Common-Mode Rejection Ratio, CMRR)和 訊雜比（Signal to Noise Rate, SNR)及高精確度，相當適 合用於生理δ孔號放大器上。收集到的呼吸音（81 )訊號經 過帶拒滤波益（212 ) (Not ch F i 11 er)與帶通減波5| (213) (Bandpass Filter)，將60Hz雜訊、低頻與高頻干擾濾除 〇後，再進入後級放大器（214)與一類比多工器（22)(Anaj〇g Multiplexer)，其主要目的是系統為一多通道呼吸音（81) 擷取的設計考量。 在數位電路（23)方面，呼吸音（81 )訊號由類比電 —路（21 )傳至類比數位轉換器（231)，將類比訊號轉成數 位訊號，接著進入微控制器（232 )進行運算，微控制器（232 ) 係用來進行韌體設定取樣速率和數位資料傳輸處理。最後 ❹透過數位傳輸介面（ 233 )將收到的數位訊號回傳至分析裝 置（3)〔通常為電腦〕端，進行數位訊號處理。 請參閱第三圖，為本發明實施例之圖形化人機介面架 構圖，其中，顯示裝置（7)可藉由圖形化人機介面（71) 輸出包含多媒體衛教影音模組（72 )、呼吸音訊號模組（73 ) 及病歷資料庫（74);多媒體衛教影音模組（72)則可提供 多媒體聽診教學（721)及多媒體呼吸照護衛教（722 )。 ^媒體聽診教學（721 )包含聽診器擺放位置照片、人 體肺臟聽診位置示意圖、呼吸音波形呈現、各種正常與異 14 201034629 常肺音聲音檔、文字描述各種肺音特性等資訊〔另請 參閱附件3〕，可讓使用者量測和儲存正常人或病患之肺部 呼吸曰（81)訊號，並直接在分析裝置⑺〔通常為電腦 H另閱第—圖〕上做呼吸音（81)訊號之分割、病理 特徵仏己和撥放，藉以提供醫學院學生之課程教學和訓練 =診^之工具。同時，也提供一個學生學習的介面，讓 學生回家學f時可以反覆聽肺音，加深其學習印象和效果。 ❹ 〇 夕媒體呼吸照護衛教（722)包含胸部扣擊與姿位引流 …、片、人體肺臟積痰位置示意圖、胸部扣擊與姿位引流古五 音影片、文字描述各部位姿位引流與胸部扣擊的方法等 ^〔另請同時參_件4〕，#系統辨識出量測部位有瘦音 時’針對肺部不同的積痰部位，配合姿位引流之文字之解 說和圖片’系統就會適時的撥放出由專業醫護人員所指導 =確的姿位引流與胸部扣擊之正確操作處置技術影片，來 助看護工或家庭照護者將病患肺中之痰排出。 ’吸音訊號模組（73)則包含呼吸音訊號擷取介面 機人）及呼吸音訊號分析介面⑽）；前者藉由圖形化人 ⑺）擷取並顯示受測者（8)之料音（81)〔另 圖、第二圖〕之肺音波形顯示晝面、錄音功能、 枝2 r、訊號儲存功能、訊號操取功能等相關資訊〔另 =時參閱附件υ’後者藉由圖形化人機介 並=受測者⑴之呼吸音（81)〔另請參閱第—圖第 _圖]之肺音訊號原始波形圖、傅立葉轉換頻域圖、平均 /功率密度頻譜圖、短時傅立葉轉換時頻圖、平均數頻 15 201034629 率與中位數頻率之參數分析等 件2〕；使用者 關貝訊〔另❺同時參閱附 D m #°進行深人的觀察，可利用啐 吸曰讯號摘取介面（731)中所招视 V -r 美供之訊號擷取功能，將該 奴讯唬擷取下以呼吸音訊號分 ㈣I 刀祈门面（732)進行訊號分析 病歷資料庫（74)則提供病歷新增與修改舊病歷資料 的:能，讓使用者可直接在分析裝置（3)〔通常為電腦〕〔 另明參閱第-圖〕上作肺音訊號之分割合病理特徵標記〔另 請同時參閱附件5〕。 使用者可以分別獨立操控圖形化人機介面（71)包含 $媒體衛教影音模組（72)、呼吸音訊號模、组（73)及病歷 貝料庫（74)等三個介面’當使用者想觀看聽診教學内容 或呼吸照護衛教内容時，可由多媒體衛教影音模組（72) 進入；當使用者想擷取呼吸音（81)訊號〔另請參閱第二 圖〕或分析呼吸音（81)訊號，可由呼吸音訊號模組（73) ❹進入；當使用者想新增、修改、紀錄受測者（8 )〔另請參 閱弟一圖〕之病歷時可由病歷資料庫（74)介面進入。 δ月參閱第四圖’為本發明實施例之呼吸音訊號分析處 理流程圖’其中’呼吸音（81 )訊號先經由數位訊號的前 處理，包括： 步驟一（Α1):去除直流和多項式漂移（detrending)。 步驟二（A2):以數位遽波器操取訊號（truncation)、 加窗（windowing)等。 步驟三（A3):進行訊號的參數分析，為配合臨床聽診 16 201034629 = 意的呼吸音性質’訊號分析經由呼吸音訊 面（732 )之特性解析，包括： ^丨 的時間Olumion)。 振“罝的分析、呼/吸期 ==析⑺22):利用快逮傅立葉轉換（fft) ^^lT0d0gr^ ^ ^ ^頻”曰的平均頻率，中位數頻率，頻寬。 ❹ Ο 此利析（7323 ):因為肺音訊料屬於時頻訊號，因 頻率變化情形。 約法刀析在某段時間上’肺音的 從上述分析進一步可找尋相關特徵參數。 步驟四（A4):痰音辨識參數萃 =性與統計分析技術找出最適當二== 區別：：五（A5):痰音辨識模型（42)利用統計模型中的 像是二模型’區別無疫與有痰之肺音或是其他辨識方法 類神經網路、人工智慧等都會有極佳的效果。 有痰：：六Γ):辨識出量測部位是否有痰音，若辨識出 執行曰夺’系統會進入步驟七（A7)及步驟八（A8)繼續 步驟七（A7):經由警示裝置⑷發出警訊。 步驟八（Α8)·同時經由多媒體呼吸照 另請^第三圖〕播放多媒體衛教影片。 （722)〔 @參閱第五冑’為本發明實施例之瘦音辨識分析處理 201034629 流程圖，其中： 步驟一（B1):針對呼吸音 行债測’確定其為吸氣或呼氣 訊號。 (81)訊號先對此段訊號進 分別定為吸氣訊號或呼氣 步驟二（B2);依楠士政 參閱第四圖〕所辨識的結果^日辨識模型（42)〔另請 情形1:當系統判斷為吸(呼m情#形產生： 〇 ❹ 氣是否異常，若為正t，m二);^會=斷此吸（呼） 為正常，則代表吸氣訊號與呼氣訊號均;:二否異常’若 氣是時，糊斷此吸（呼） ::: ::為正吊，則判斷此呼(吸 異r表此段訊號在吸(呼)氣期為正常，而，(吸) 氣是為吸(呼)氣時，會先判斷此吸(呼) 異，右為異吊’則判斷此呼(吸)氣是否異 ;期:正:一訊號在吸⑷氣期為異常，而呼⑷ 情形4:當系統判斷為吸（啤）氣時，會先判斷 =異常，若為異常’則判斷此呼(吸)氣是否異) 為異吊，則代表吸氣訊號與呼氣訊號均為異常。 右 上述情形2至情形4均係屬異常痰音 ::第-圖〕’則經由警示震置⑷發出警訊ί:;:: 護衛教⑽〔另請參閱第三圖〕播放，； 18 201034629 之應明:系針對瘦音偵測、辨識與衛教系統 " η 、曰種藉由多通道呼吸音量測裝置（丨）' 轉換電路（2)、分析襄置 （ 组（4 )、瞀⑽/ M通中為電月自〕、痰音辨識模 、、 運异早70 (5)以及警示裝置⑷之整合 以，識異常疫音（82)進行相關處置與衛教工二 ^之改良與設計，為本發明對於痰音仙卜 教糸統所作最具體之精進。 ^術 Ο 【圖式簡單說明】 第圖.本發明實施例之系統架構流程圖。 第二圖：本發明實施例之呼吸音量測裝置電路架構圖。 第二圖：本發明實施例之圖形化人機介面架構圖。 第四圖：本發明實施例之呼吸音訊號分析處理流程圖。 第五圖：本發明實施例之痰音辨識分析處理流程圖。 附件1 :呼吸音訊號擷取介面。 附件2:呼吸音訊號分析介面。 Ο 附件3:多媒體聽診教學法。 附件4:多媒體姿位引流呈現方法。 附件5 .病歷標記資料儲存庫。 附件6 :硬體控制電路之實施例。 附件7 :使用者人機介面之實施例。 【主要元件符號說明】 (1 ) 呼吸音量測裝置 (II) 機械式聽診集音盤 (III) 聲音感測器 201034629 Ο• Death also provides caregivers with proper respiratory care knowledge. In the future, in the health care chapter, the T-IT industry includes. Μ ^ The market share. Health care products 1. Home care and health education. 2. Respiratory care for long-term care facilities. 3 · Respiratory care in medical institutions. The problem of shortage of nursing manpower every day - straight, and nursing staff swearing = wide: matters include: medicine, injection, record work, ambition = 痰, convulsions, turning over, scrub, so the current tendency. On the other hand, the second squat will increase its risk, because no matter whether it is watching 〇-skills, it is like the basic listening ability 2 of the 4 Guardian. Therefore, in the implementation of the correct work or home care provider for the teaching of the education, = he = to care for the care technology. Although the traditional stethoscope can be sure that it must have a professional auscultation of the moon I, see the lung sound, but the fruit can use the number of 1 taste step to have a way to judge the symptoms, here - some scientific quantity two = 1 The technique is to digitize the breath sounds, and the basics are to listen to the bee prosthetics, and the health care personnel are judged by the medical staff. No matter what, the hair system is limited in terms of intuition and surgery. The problem of limitation, the point 'and the performance level of the practical technology and the performance of the performance of the performance === 8 201034629 ::: the convenience of the two aspects of the improvement, in order to achieve the desired it, ί Listen to the "analog" of the analysis, and make it both sound. The actual development of the applicability of the knowledge and technology development equipment industry and the development of the equipment industry ❹ 立 立 = = 不 不 "" from the subjective and objective conditions, using the multi-point speculation of the 姑 μ Gu 1 * identification and deeds 'Unification' is currently no phase in domestic and foreign patents == used in the field of respiratory sound diagnosis, and according to the statistics of the Ministry of the Interior's Social Department: 006 years, China's elderly population over 65 years old accounted for 9 87 of the total population. The social standards are estimated to be in the next 25 years, and according to the study of the Organization for Economic Co-operation and Development (〇ecd), the average medical expenditure per person above the age of the elderly is about four per person under π years old. According to a study by the American Association of Retired Persons, middle-aged 3 = medical care, and people who are willing to seek home at home ^ '-'^ service' shows that home medical care has become a global heavyweight' Therefore, it has the advantage of the irreplaceable technology in the market. It is very suitable for the equipment market such as voice detection, identification and education. It is bound to bring about epidemic detection: identification and production and design and manufacture of educational technology development equipment. Industry-related city Great business opportunities. Fang II achieves the above objectives and functions' specific technical means and a voice detection, identification and education system, including: standing to the 乂 breathing 曰 summer measuring device, including mechanical auscultation set and sound two sensing The 'sound sensor' is located in the mechanical auscultation set to convert the collected breath sounds into breath sound analog signals. 9 201034629 The road 11 has an analog circuit, an analog multiplexer and a digital circuit, and is converted into a breath sound digital signal by using == sound. A storage and analysis of breath sound digital signals. The thin sound signal 0 of the (4) sound-sounding digital signal is electrically connected to the voice recognition module for receiving and processing the voice signal and generating a processing result. The stf device' is electrically connected to the arithmetic unit 1 so as not to act or generate a corresponding warning message according to the processing result of the operation unit. The analog circuit described above includes a preamplifier, a band reject filter, a band pass filter, and a post amplifier. The above-mentioned analogy device integrates the breath sound analog signal of the breathing volume measuring device into the digital circuit. The above digital circuit includes an analog digital converter, a microcontroller, and a digital transmission interface. The above-mentioned voice recognition module includes a voice recognition parameter extraction and a voice recognition die pear. The computing unit is further electrically connected to a display device, and the display device is connected to the multimedia audio-visual video module, the respiratory audio signal module and the medical record database, and the display device is provided with a graphical human-machine interface. The above-mentioned multimedia education audio-visual module includes digital information for multimedia auscultation teaching and digital information for multimedia respiratory care. The breath sound signal module includes a breath sound signal capture interface and a beer sound absorption signal analysis interface. 201034629 - The above operation unit generates a processing result according to the voice signal of the voice recognition module, and according to the processing result, at least one of the multimedia education video module, the respiratory sound signal module and the medical record database ( Or database) Output the corresponding message. The specific features and effects of the present invention are as follows: 1. The system detects the part of the lung that produces mucus (痰), and then plays a respiratory care-related health education film to guide the caregiver to the part that produces mucus o (痰). Into the drainage of the sputum 'help the patient clear the secretions of the respiratory tract' to maintain the smooth flow of the respiratory tract. 2. It can provide the caregiver's mastery of the timing of convulsions~ not only prevent patients from dying due to severe obstruction of the respiratory tract, but also reduce the risk of convulsions. . 3. Provide teacher-assisted instructional use and provide education for students, clinical care 5 for respiratory sound auscultation, and can also teach caregivers and families to follow the correct respiratory care knowledge, thereby improving the ability of medical care staff to listen to percussion and The quality of medical care. [Embodiment] Referring to the first figure, a flowchart of a system architecture according to an embodiment of the present invention, wherein a voice detection, identification, and teaching system includes: a less A breath sounding device (丨), conversion The circuit (2), the analysis device (3) ^ is usually a computer], the voice recognition module (4), the arithmetic unit (5), and the: no device (6), wherein the respiratory volume measuring device 〇) receives from the measured. ^8) Breathing sound (81) 'Re-input conversion circuit (2) for conversion, after conversion to analyze the vibration (3) storage and analysis results, and pass the result to 11 201034629 - send to voice recognition parameters Extraction (41) and voice recognition model ((2) 痰: discrimination, Mo group (4), to identify the subject (1) beer absorbing sound (81) 再 恕 再 ' and then with the voice recognition module (4 The electrical connection unit (5) receives and processes a processing result. The operation unit (1) is electrically connected to the Lv (6) for processing according to the processing result of the operation unit (1) = generating or generating a The corresponding warning message. The operation unit 兀 (5) is electrically connected to the display device (7), The device () is connected to the eve media media audio and video module (72), the respiratory sound signal module (7)), and the display device (7) is provided with a graphical human-machine surface (71); The teaching audio and video module (72) includes a digital information of the multimedia I ^ teaching (721) and a digital message of the multimedia respiratory nursing (722), and the breathing sound signal module (73) includes a breathing sound signal operation. ^73U and the breath sound signal analysis interface (732). The arithmetic unit (5) generates a processing result according to the voice signal of the voice recognition module (4), and according to the processing result, the multimedia voice video module 72 (72), At least one module (or database) of the respiratory sound signal module (73) and the medical record database (μ) outputs a corresponding message; that is, the subject's voice recognition result (whether normal or abnormal) is loaded. The medical record database ((4) is recorded. If abnormal noise is found, the warning device (4) will generate a report message to alert the relevant medical care or caregivers, and will also be loaded with the multimedia education module (72) as an educational training. Actual reference information. See The second figure is a diagram of the respiratory volume measurement 'road structure diagram of the embodiment of the present invention, wherein the breathing volume measurement I set (1) [also please also read the 12th 201034629 picture] each includes a built-in sound sensor (111 The mechanical auscultation set (Π), the complex channels [channels 1, 2...] are each independent analog circuit (21) 'The various ratio circuits (21) contain the preamplifier (211), with rejection filter The device (212), the band pass filter (213) and the post amplifier (214); the analog circuit (21) of each complex channel is integrated by the analog multiplexer (22) and enters the digital circuit (23), the digital circuit (23) The analog digital converter (231), the microcontroller (232), and the digital transmission interface (233) are included, and finally the digital signal is output. The mechanical auscultation set (u) is placed in various parts of the human lung (each part is connected to a breathing volume measuring device (丨) and analog circuit to form an independent signal channel) to collect breath sounds (81), mechanical The auscultation set (11) is provided with a sound sensor (m), which converts the breath sound (81) into a breath sound analog signal (811) (which is a primitive analog signal), via a preamplifier (211), with a belt After the analog signal processing program such as the rejection filter (212), the band pass filter (213), and the post amplifier (214), the breath sound analog signal (811) of each channel enters the digital circuit through the analog multiplexer (22). ❹(23)' is converted into a breath sound digital signal (812) output by an analog digital converter (231), a digital controller (232), and a digital transfer interface (233) via a digital circuit (23) to connect the analysis device. (3) [usually a computer] and a voice recognition module (4) [see also the first figure]. The analog circuit (21) analog multiplexer (22) and the digital circuit (23) together constitute a conversion circuit (2) as shown in the first figure, that is, the conversion of the conversion circuit can be used to convert the original mechanical auscultation set. The breather (1 1 ) converts the breath sound (81) into a breath sound analog signal (811), which is further converted into a breath sound analog signal (812) to facilitate subsequent analysis and processing. 13 201034629 - In the analog circuit (21), since the amplitude of the original breath sound (μ) signal is very small, in order to correctly measure the breath sound (8 1 ) signal, the system uses an instrumentation amplifier as the preamplifier (211). Mainly due to its Common-Mode Rejection Ratio (CMRR) and Signal to Noise Rate (SNR) and high accuracy, it is quite suitable for physiological delta aperture amplifiers. The collected breath sound (81) signal passes through the rejection filter (212) (Not ch F i 11 er) and the bandpass subtraction 5| (213) (Bandpass Filter), which will 60Hz noise, low frequency and high frequency interference. After filtering out the enthalpy, it then enters the post-amplifier (214) and an analog multiplexer (22) (Anaj〇g Multiplexer) whose main purpose is to design a multi-channel breath sound (81). In the digital circuit (23), the breath sound (81) signal is transmitted from the analog electric circuit (21) to the analog digital converter (231), the analog signal is converted into a digital signal, and then enters the microcontroller (232) for operation. The microcontroller (232) is used to perform firmware setting sampling rate and digital data transmission processing. Finally, the received digital signal is transmitted back to the analysis device (3) (usually the computer) through the digital transmission interface (233) for digital signal processing. Please refer to the third figure, which is a schematic diagram of a schematic human-machine interface architecture according to an embodiment of the present invention, wherein the display device (7) can output a multimedia educational video module (72) through a graphical human-machine interface (71), The respiratory audio signal module (73) and the medical record database (74); the multimedia education audio-visual module (72) can provide multimedia auscultation teaching (721) and multimedia respiratory care (722). ^Media auscultation teaching (721) contains photographs of position of stethoscope, schematic position of auscultation of human lungs, waveform of breath sounds, various normal and different 14 201034629 sounds of lung sounds, text description of various lung sound characteristics, etc. (Please also refer to the attachment 3), allows the user to measure and store the respiratory sputum (81) signal of the normal person or patient, and directly make a breath sound on the analysis device (7) (usually computer H read the same figure) (81) The division of the signal, the pathological characteristics of the self-discipline and the release, in order to provide medical students with the course teaching and training = diagnostic tools. At the same time, it also provides a student learning interface, so that students can listen to lung sounds when they go home to learn f, and deepen their learning impressions and effects. 〇 〇 媒体 Media Respiratory Care (722) includes chest slamming and posture drainage..., film, human lung accumulation position map, chest slamming and posture drainage ancient five-tone film, text description of various positions of position drainage and chest The method of smashing, etc. ^[Please also refer to _piece 4], #System recognizes that when the measurement part has thin sound, 'for the different accumulation parts of the lungs, the explanation and picture of the text with the posture position drainage It will timely release the correct operation and treatment technology film that is guided by professional medical staff = correct posture drainage and chest smash, to help the caregiver or home caregiver to discharge the sputum in the patient's lungs. 'The sound absorbing signal module (73) includes the breathing sound signal acquisition interface machine) and the respiratory sound signal analysis interface (10); the former captures and displays the sound of the subject (8) by the graphical person (7) ( 81) [other figure, second picture] lung sound waveform display face, recording function, branch 2 r, signal storage function, signal manipulation function and other related information (also = see attachment υ 'the latter by graphical people The machine is the = (1) breath sound (81) [see also Figure - Figure _] lung sound signal original waveform, Fourier transform frequency domain, average / power density spectrum, short-time Fourier transform Time-frequency diagram, average frequency 15 201034629 rate and median frequency parameter analysis and so on 2]; user Guan Beixun (also see also attached D m # ° for deep observation, you can use the suction signal In the extracting interface (731), the signal extraction function of the V-r US is taken, and the slave is removed by the breathing signal (4) I knife praying facade (732) for signal analysis medical record database (74) ) Provide medical records to add and modify old medical records: Connected to the analysis device (3) [usually a computer] (see also Figure - Figure) for the division of the lung sound signal and the pathological feature mark (see also Annex 5). The user can independently control the graphical human-machine interface (71) including three media interfaces: media media audio and video module (72), respiratory audio signal module, group (73) and medical record library (74). When you want to watch the auscultation content or the respiratory care content, you can enter it by the multimedia audio-visual module (72); when the user wants to capture the breath sound (81) signal (see also the second picture) or analyze the breath sound (81) The signal can be entered by the breathing sound signal module (73); when the user wants to add, modify, and record the medical record of the subject (8) (see also the picture of the younger brother), the medical record database (74) ) The interface enters. In the fourth month, reference is made to the fourth figure 'flow chart of the respiratory sound signal analysis process according to the embodiment of the present invention', wherein the 'breath sound (81) signal is pre-processed through the digital signal, including: Step 1 (Α1): removing DC and polynomial drift (detrending). Step 2 (A2): Using a digital chopper to perform truncation, windowing, and the like. Step 3 (A3): Perform parameter analysis of the signal to coordinate with clinical auscultation. 16 201034629 = Meaning of breath sounds' signal analysis through the characteristics of the breath sound surface (732), including: ^丨 time Olumion). Vibration "罝 analysis, call / suction period == analysis (7) 22): using the fast catch Fourier transform (fft) ^ ^ lT0d0gr ^ ^ ^ frequency "曰 average frequency, median frequency, bandwidth. ❹ Ο This analysis (7323): Because the lung audio signal is a time-frequency signal, due to frequency changes. The method of analyzing the lung sounds from a certain period of time can further find relevant characteristic parameters from the above analysis. Step 4 (A4): Voice recognition parameter extraction = sex and statistical analysis techniques to find the most appropriate two == difference:: five (A5): voice recognition model (42) using the statistical model in the image is the two model 'distinction There is no effect on the lungs of the epidemic and the lungs or other identification methods, such as neural networks and artificial intelligence. There are::6Γ): Identifies whether there is a voice in the measurement part. If it is recognized that the system is executed, the system will enter step 7 (A7) and step 8 (A8) to continue step 7 (A7): via the warning device (4) Send a warning. Step 8 (Α8)·At the same time, through the multimedia breathing photo, please ^^3] broadcast the multimedia education video. (722) [@ 胄 胄 胄 胄 胄 瘦 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 (81) The signal is first set as the inspiratory signal or exhalation step 2 (B2) for this segment of the signal; the result identified by the following is the identification model (42) [Please also case 1 : When the system judges to be sucking (when the mood is generated: whether the suffocation is abnormal, if it is positive t, m 2); ^ will = break the suction (call) is normal, it means the inspiratory signal and the exhalation signal Both;: 2 no abnormality 'If the gas is, when the suction (call) ::: :: is the hanging, then judge this call (sucking the r table this signal is normal in the suction (call) gas period, However, when (sucking) gas is for sucking (calling) gas, it will first judge whether the suction (call) is different, and if the right is different hanging, then it is judged whether the call (sucking) gas is different; period: positive: one signal is sucking (4) The gas period is abnormal, and the call (4) Case 4: When the system judges that it is sucking (beer) gas, it will judge = abnormal. If it is abnormal, then it is judged whether the breath is different. Both the air signal and the breath signal are abnormal. The above situation 2 to case 4 are all abnormal voices:: - map] 'is sent a warning via the warning (4) ί:;:: Guardian (10) [see also the third picture] play,; 18 201034629 It should be clear: for the thin tone detection, identification and health education system " η, 藉 by multi-channel breathing volume measurement device (丨) 'conversion circuit (2), analysis device (group (4),瞀(10)/ M-pass is the integration of electric moon from the moon, the voice recognition module, the transport of the early 70 (5) and the warning device (4), the abnormal epidemic sound (82) for related disposal and the improvement of the teaching staff And the design is the most specific advancement of the present invention for the voice syllabus. ^ Ο Ο [Simplified description of the drawings] Figure 1. System architecture flow chart of the embodiment of the present invention. Second figure: embodiment of the present invention The circuit diagram of the breathing volume measuring device. The second figure: the graphical human-machine interface architecture diagram of the embodiment of the invention. The fourth figure: the flow chart of the respiratory sound signal analysis processing according to the embodiment of the invention. Flow chart of the voice recognition analysis processing of the embodiment. Annex 1: Breathing sound signal acquisition Attachment 2: Breathing sound signal analysis interface. 附件 Annex 3: Multimedia auscultation teaching method. Annex 4: Multimedia posture drainage presentation method. Annex 5. Medical record data repository. Annex 6: Example of hardware control circuit. 7: User human interface. [Main component symbol description] (1) Respiratory volume measurement device (II) Mechanical auscultation sound collection disk (III) Sound sensor 201034629 Ο
(2) 轉換電路 (21) 類比電路 (211) 前級放大器 (213) 帶通濾‘波器 (22) 類比多工器 (23) 數位電路 (24) 音量微調器 (231) 類比數位轉換器 (233 ) 數位傳輸介面 (3) 分析裝置 (4) 痰音辨識模組 (41) 痰音辨識參數萃取 (42) 痰音辨識模型 (5) 運算單元 (6) 警示裝置 (7) 顯示裝置 (71) 圖形化人機介面 (72) 多媒體衛教影音模組 (721) 多媒體聽診教學 (722 ) 多媒體呼吸照護衛教 (73) 呼吸音訊號模組 (731) 呼吸音訊號擷取介面 (732 ) 呼吸音訊號分析介面 (7321 ) 時域分析 (212) (214) (232 ) (7322) 帶拒濾波器 後級放大器 微控制器 頻域分析 20 201034629 (7323) 時頻分析 (74) 病歷資料庫 (8) 受測者 (81) 呼吸音 (811) 呼吸音類比訊號 (812) 呼吸音數位訊號 (82) 異常痰音 (9) 耳機 (Al) 步驟一 (A2) 步驟二 (A3) 步驟三 (A4) 步驟四 (A5) 步驟五 (A6) 步驟六 (A7) 步驟七 (A8) 步驟八 (Bl) 步驟一 (B2) 步驟二(2) Conversion circuit (21) Analog circuit (211) Preamplifier (213) Bandpass filter 'waver (22) Analog multiplexer (23) Digital circuit (24) Volume trimmer (231) Analog to digital converter (233) Digital transmission interface (3) Analysis device (4) Voice recognition module (41) Voice recognition parameter extraction (42) Voice recognition model (5) Operation unit (6) Warning device (7) Display device ( 71) Graphical Human Machine Interface (72) Multimedia Education Video Module (721) Multimedia Auscultation Teaching (722) Multimedia Respiratory Education (73) Breathing Sound Signal Module (731) Breathing Sound Signal Capture Interface (732) Breathing Sound Signal Analysis Interface (7321) Time Domain Analysis (212) (214) (232) (7322) Frequency Domain Analysis of Rejection Filter Post-Amplifier Microcontrollers 20 201034629 (7323) Time-Frequency Analysis (74) Medical Record Database (8) Subject (81) Breath sound (811) Breath sound analog signal (812) Breath sound digital signal (82) Abnormal voice (9) Headphone (Al) Step 1 (A2) Step 2 (A3) Step 3 (A4) Step 4 (A5) Step 5 (A6) Step 6 (A7) Step 7 (A8) Steps Eight (Bl) Step One (B2) Step Two
21twenty one