使用頦下超音波影像評分與應變量測評估阻塞性睡眠呼吸中止症

阻塞性睡眠呼吸中止症（Obstructive sleep apnea, OSA）是現今常見的睡眠疾病，主要發生原因包括伴隨老化與肥胖而來之口腔周圍的肌肉張力下降，導致OSA患者容易於睡眠平躺時舌根後墜、呼吸道受到壓迫而缺氧。目前臨床主要的診斷方法為使用睡眠多項生理功能檢查（Polysomnography, PSG），但PSG對於患者的病理生理學無法明確解釋，只能從 PSG 得知OSA嚴重程度。研究口腔周遭組織的物理參數與OSA之間的相關性，有助於臨床醫師判斷受測者是否有潛在OSA的可能性，因此陸續有研究使用磁振造影、電腦斷層掃描分析組織的物理參數。相較於上述方法，超音波具有無輻射與即時成像的優點，能更容易且無害地觀察舌頭與口腔組織。本研究使用超音波舌頭功能影像與應變量測，分析有別於舌頭結構（如：厚度、面積）的功能參數，應變彈性影像只需對灰階超音波影像做分析，即可計算舌頭應變與彈性，相較於剪切波彈性影像更容易實現。應變彈性影像亦可搭配口腔動作，動態分析動作期間的舌頭位移、計算分析舌頭彈性。本研究將舌頭分為三個部分，分別是最大舌根尖（Maximum tongue base apex, TBA）、TBA於舌骨端（Hyoid bone side, TBAH）與TBA於下顎骨端（Mandible side, TBAM），並使用舌壓測定器（Maximum tongue pressure measurement, TP）與倒氣測試（Müller's maneuver, MM）兩個動作，搭配斑點追蹤觀察正常呼吸（Normal breathing, NB）到MM或TP期間TBA、TBAH、TBAM的變化，收案的受測者包括8位健康、30位輕度、30位中度、30位重度OSA患者。然而，斑點追蹤對影片中連續幀的品質要求較高，研究過程我們發現部分影像因舌頭拍攝不清楚或受測者動作過大導致舌頭無法穩定地被追蹤，因此我們使用深度學習與類激活映射（Class activation mapping, CAM）計算影像品質分數（Image quality score, IQS）以定量超音波影像品質，並以適合的IQS截止點篩選超音波影像，研究結果顯示較高的IQS可獲得較高的斑點追蹤穩定度。結果表示中重度與健康輕度OSA在MM動作下的TBA應變存在相關性，且預測中重度與非中重度OSA的操作特徵曲線下面積為0.75、敏感度75%、特異度71.4%，說明了使用應變分析OSA存在其研究的價值。

關鍵字

阻塞性睡眠呼吸中止症；舌頭功能影像；應變分析；頦下超音波影像；影像品質分數；倒氣測試；斑點追蹤

並列摘要

Obstructive sleep apnea (OSA) is a common sleep disease mainly seen in the elderly and obese. The leading cause is the low tension around the oral muscle. As a result, patients with OSA are prone to airway compression and hypoxia when sleeping supine. Currently, the principal clinical diagnosis method is based on polysomnography (PSG). The PSG indicates the severity of OSA, but it does not explain the pathophysiology. Studying the correlation between the functions of the oral tissue and the OSA would help clinicians judge the potential risk. Therefore, various studies have been conducted by using MRI and CT to analyze the physical parameters of tissues. Ultrasound has the advantages of non-radiation and imaging in real-time, making it efficient to observe the tongue and oral tissues. This study uses ultrasound tongue functional imaging based on strain measurements instead of morphological information (e.g., thickness, area). Strain elastography is based on speckle tracking of the B-mode images, which is more straightforward than shear wave elastography. In this study, we use submental ultrasound imaging to observe the patient's tongue, which consists of three parts: maximum tongue base apex (TBA), TBA on the hyoid bone side (TBAH), and TBA on the mandible side (TBAM). The measured parameters include maximum tongue pres-sure measurement (TP) and müller's maneuver (MM), and observing the difference of TBA, TBAH, TBAM during the period from normal breathing (NB) to TP or MM. We recruited eight healthy volunteers, 30 mild OSA patients, 30 moderate OSA patients, and 30 severe OSA patients for the study. Good image quality is needed for reliable speckle tracking results. In this study, however, we found the robustness of tongue movement limits the speckle tracking performance. To circumvent this issue, we use Deep Learning and class activation map (CAM) to quantify the image quality score (IQS) and filter out ultrasound images with poor image quality by a suitable IQS cut-off point. The results show that ultrasound images with higher IQS can obtain higher stability of speckle tracking. The results also show that TBA strain in MM cor-relates reasonably well with OSA. The area under the receiver operating characteristic curve is 0.75, sensitivity is 75%, and specificity is 71.4%. Thus, it shows that strain analysis of the tongue can be a valuable tool for OSA diagnosis.

並列關鍵字

obstructive sleep apnea ； tongue functional imaging ； strain measurements ； submental ultrasound ； image quality score ； müller's maneuver ； speckle tracking

參考文獻

[1] A. M. Kim et al., "Tongue Fat and its Relationship to Obstructive Sleep Apnea," Sleep, vol. 37, no. 10, pp. 1639-U227, Oct 2014, doi: 10.5665/sleep.4072

Google Scholar

[2] W. Banhiran, A. Junlapan, P. Assanasen, and C. Chongkolwatana, "Physical Predictors for Moderate to Severe Obstructive Sleep Apnea in Snoring Patients," Sleep and Breathing, vol. 18, no. 1, pp. 151-158, Mar 2014, doi: 10.1007/s11325-013-0863-y

Google Scholar

[3] S. Redline et al., "Obstructive Sleep Apnea-Hypopnea and Incident Stroke The Sleep Heart Health Study," Am. J. Respir. Crit. Care Med., vol. 182, no. 2, pp. 269-277, Jul 2010, doi: 10.1164/rccm.200911-1746OC

Google Scholar

[4] A. S. Gami et al., "Obstructive Sleep Apnea and the Risk of Sudden Cardiac Death A Longitudinal Study of 10,701 Adults," J. Am. Coll. Cardiol., vol. 62, no. 7, pp. 610-616, Aug 2013, doi: 10.1016/j.jacc.2013.04.080

Google Scholar

[5] M. J. Sateia, "International Classification of Sleep Disorders," Chest, vol. 146, no. 5, pp. 1387-1394, 2014, doi: 10.1378/chest.14-0970

Google Scholar

國際替代計量

使用頦下超音波影像評分與應變量測評估阻塞性睡眠呼吸中止症

未授權

主題瀏覽