DSpace collection: 博碩士論文

優化新型口腔負壓治療阻塞性睡眠呼吸中止症之療效;優化新型口腔負壓治療阻塞性睡眠呼吸中止症之療效

Tue, 20 Oct 2026 08:34:45 GMT

title: 優化新型口腔負壓治療阻塞性睡眠呼吸中止症之療效;優化新型口腔負壓治療阻塞性睡眠呼吸中止症之療效 abstract: 中文摘要一、研究背景與動機阻塞性睡眠呼吸中止症（OSA）是全球盛行率逐年攀升的慢性睡眠呼吸障礙，患者因睡眠時咽喉部上呼吸道反覆塌陷與閉塞，引發間歇性低氧血症、高碳酸血症及睡眠碎片化，是導致高血壓、心肌梗塞等心血管疾病的獨立危險因子。目前臨床一線黃金標準為連續正壓呼吸器（CPAP），其利用正壓氣流作為「氣動支架」撐開氣道。然而，CPAP 在真實世界的臨床失败或中斷率高達40%至50%，主因在於強迫性正壓氣流帶來的物理不適（如黏膜乾燥、呼氣阻力）、面罩漏氣引起的微覺醒，以及面部強力束縛產生的幽閉恐懼與心理負荷。為克服上述弊端，新興的口腔負壓治療（OPT）系統（如 iNAP® Lite）應運而生。其透過在閉合的口腔內建立微真空環境，產生表面黏附力與力學拉力，主動將鬆弛的軟顎與後墜的舌體向前向上固定，確保咽喉通道暢通。OPT 不遮蓋口鼻、無強迫氣流，具備高依從性潛力。然而，過去早期臨床多採用固定壓力（-40 mmHg）的一體適用經驗性治療，導致療效因個體間的「解剖與病理表型異質性」（如BMI、頸圍、軟組織順應性不同）而表現出較大不穩定性，整體成功率僅約60%。本研究推動動機正奠基於此反思：如同 CPAP 需透過全夜睡眠多項生理檢查（PSG）進行壓力滴定才能找到最佳治療壓力，作用於複雜口腔解剖結構的 OPT，同樣必須經歷個人化的壓力滴定程序，以邁向精準睡眠醫學。二、研究目的本研究之總體目標在於評估並論證 OSA 患者接受個人化、漸進式口腔負壓（OPT）動態滴定後的具體臨床療效。具體子目標包括：驗證其對呼吸暫停事件的初級清除療效（AHI改善與血氧穩定度）；解碼微觀睡眠結構之調整（特別是 N3 深層睡眠之比例）與逆轉睡眠碎片化之機制；評估 14 天居家去敏感適應期之臨床信度與日間嗜睡改善；以及探討局部高負壓干擾之生理中性與臨床安全性邊界（如對週期性肢體抽動指數 PLM Index 的影響與局部黏膜刺激）。三、研究方法本研究共納入 30 名 20 至 70 歲之中重度 OSA 受試者（排除中樞型呼吸中止、嚴重肺病及嚴重牙周病者）。第一階段（臨床篩選與基準值評估）：記錄受試者基本解剖特徵，並填寫愛普沃斯嗜睡量表（ESS）評估基線日間過度嗜睡程度。第二階段（14天居家適應期）：受試者正式滴定前需於家中持續使用 OPT 設備至少 14 天，建立神經習慣化去敏感，降低不適感引發的微覺醒，提升數據信度。第三階段（過夜壓力滴定實驗）：在醫院睡眠中心由專業技師監控執行。受試者同時配戴 PSG 與 OPT 裝置。初始負壓設定為 -40 mmHg，以每 15 分鐘為一評估窗口。若發生大於 2 次呼吸中止、大於 3 次低通氣、大於 5 次 RERA 或持續 3 分鐘鼾聲，則遞增 10 mmHg 負壓，壓力上限設定為生理安全閾值 -250 mmHg。最佳壓力判定指受試者處於仰臥位且進入快速動眼期（REM）時，能使呼吸事件降至最低、且血氧飽和度最穩定的最小負壓值。四、研究結果與發現 30 名受試者之平均年齡為 51.27±13.97 歲，平均 BMI 為 25.87±3.42 kg/m2。經過個人化壓力滴定介入後，核心生理指標變化如下：初級療效（呼吸事件與血氧改善）：全體受試者之平均 AHI 從基準值的 39.59 events/hr 大幅降至治療後的 8.17 events/hr，總體 AHI 改善率高達 80.4%，具備極其顯著的統計學差異（p < 0.001）。以治療後 AHI < 5 events/hr 為優異療效標準（Excellent response），共有 25 名受試者成功達標，總體臨床成功率高達 83.3%。夜間最低血氧飽和度（Min SpO2）也由 80.93% 顯著改善至 83.70%（p = 0.023）。次級療效（睡眠結構與日間症狀優化）：受試者的第三期非快速動眼睡眠（N3 深層睡眠）比例顯著由 14.89% 提升至 23.23%（p = 0.003），顯示睡眠品質得到實質質性轉變。同時，全夜微覺醒指數（Arousal Index）由 48.47 次/小時顯著降至 26.23 次/小時（p < 0.001），有效逆轉微觀睡眠碎片化。客觀指標的改善成功轉譯為日間生活品質的提升，愛普沃斯嗜睡量表（ESS）評分由 10.61 顯著降至 5.52（p < 0.001）。安全性與生理中性：治療前後之週期性肢體抽動指數（PLM Index）並無統計學上的顯著差異（p = 0.816），證實當負壓逐步遞增至生理安全閾值上限時，局部物理牽引機制具備高度的解剖區域侷限性與生理中性，排除誘發治療誘導型肢體運動障礙之風險。部分受試者在高負壓下雖有短暫口腔黏膜刺激、酸痛或唾液過多等副作用，但通常在移除設備或適應後即獲得實質改善。五、結論與貢獻本研究成功證實了口腔負壓治療（OPT）具備顯著的「劑量效應（Dose-response relationship）」，透過增加口腔內負壓的個人化滴定程序，能有效克服因個體組織順應性不同而導致的治療失敗。在流體力學上，OPT 在個人化最佳壓力下運作時，負壓完全侷限於口腔封閉艙內，口咽與鼻咽部達成力學隔離，不會像 CPAP 般增加呼吸道的呼氣阻力與呼吸功（WOB），兼具局部精準性與生理中性。本研究的臨床實證成功將 OPT 從過去的經驗性嘗試正式提升為「精準睡眠醫學」的標準臨床實務，不僅為廣大無法耐受黃金標準 CPAP 的 OSA 患者開闢了一條高依從性、高舒適度且客觀有效的全新一线治療希望，亦能為下一代智慧化、自動調壓型（Auto-OPT）醫療器材的密閉迴路控制演算法研發，提供關鍵的生物力學核心參數矩陣。關鍵詞：阻塞性睡眠呼吸中止症（OSA）、口腔負壓治療（OPT）、壓力滴定、睡眠多項生理檢查（PSG）、深層睡眠（N3）;Optimizing treatment of the novel intra-oral negative pressure for obstructive sleep apnea Abstract 1. Background and Motivation Obstructive Sleep Apnea (OSA) is a highly prevalent chronic sleep-disordered breathing condition characterized by the recurrent collapse and occlusion of the upper airway during sleep. This anatomical obstruction induces intermittent hypoxemia， hypercapnia， and profound sleep fragmentation， serving as an independent risk factor for major cardiovascular co-morbidities such as hypertension and myocardial infarction. While Continuous Positive Airway Pressure (CPAP) remains the first-line clinical gold standard by acting as a "pneumatic splint" to hold the airway open， its real-world effectiveness is heavily compromised by suboptimal patient compliance， exhibiting long-term failure or discontinuation rates between 40% and 50%. This poor adherence stems primarily from the sensory discomfort of forced airflow (e.g.， mucosal dryness， increased work of breathing)， mask leakages that trigger cortical arousals， and psychological burdens such as claustrophobia. To address these critical limitations， Intra-oral Negative Pressure Therapy (OPT) systems (such as the iNAP® Lite) have emerged as an innovative alternative. By establishing a localized， sealed micro-vacuum inside the closed oral cavity， OPT utilizes surface adhesion and mechanical traction to stabilize the soft palate anteriorly and anchor the tongue superiorly and forward. This effectively clears the pharyngeal airway while keeping the nose unmasked and free of forced airflow， offering immense potential for high long-term compliance. However， early clinical implementations of OPT heavily relied on a fixed empirical pressure (typically -40 mmHg) under a "one-size-fits-all" paradigm. Due to the vast "anatomical and pathological phenotypic heterogeneity" among OSA patients—including stark variations in body mass index (BMI)， neck circumference， and upper airway tissue compliance—this fixed-pressure model resulted in unstable clinical response rates hovering around only 60%. For individuals with severe obstructions or extensive pharyngeal fat infiltration， a low fixed pressure is simply insufficient to overcome the mechanical resistance of the surrounding soft tissues， leading to treatment failure. The motivation of this doctoral research is directly rooted in this clinical gap: just as CPAP requires an overnight titration procedure during polysomnography (PSG) to determine the ideal therapeutic pressure， OPT—acting on highly complex oral anatomy—must also undergo a systematic， personalized pressure titration protocol to transition the technology into the realm of precision sleep medicine. 2. Research Objectives The primary objective of this study is to evaluate and validate the clinical and physiological efficacy of a personalized， incremental OPT dynamic titration protocol in patients with moderate-to-severe OSA. The specific sub-objectives are: to verify its primary ventilatory efficacy in eliminating obstructive events and stabilizing blood oxygen levels; to decode the qualitative shifts in macro and micro sleep architecture (particularly the restoration of N3 deep sleep) and elucidate the mechanisms underlying the reversal of sleep fragmentation; to assess the clinical reliability of a 14-day home-based desensitization/habituation period alongside improvements in daytime symptoms; and to investigate the physiological neutrality and clinical safety boundaries of high localized negative pressure， specifically regarding its impact on the Periodic Limb Movement (PLM) Index and local mucosal irritation. 3. Methodology This study prospectively enrolled 30 adult participants aged 20 to 70 years diagnosed with moderate-to-severe OSA， excluding individuals presenting with central sleep apnea， severe pulmonary disease， or advanced periodontal conditions. Phase I (Clinical Screening & Baseline Assessment): Comprehensive baseline anatomical characteristics were documented， and the Epworth Sleepiness Scale (ESS) was administered to quantify the severity of baseline excessive daytime sleepiness. Phase II (14-Day Home Habituation): Prior to the formal overnight laboratory experiment， participants utilized the OPT device at home for at least 14 consecutive days. This phase establishes neural habituation and peripheral sensory desensitization， reducing transient device-induced micro-arousals to ensure high clinical data reliability. Phase III (Overnight Lab Titration Experiment): Executed in a specialized sleep center under continuous monitoring by a registered sleep technologist， participants underwent full-night PSG concurrently with the OPT system. The vacuum pressure started at a baseline of -40 mmHg， using a 15-minute evaluation window. If obstructive criteria were met—defined as >2 apneas， >3 hypopneas， >5 respiratory effort-related arousals (RERAs)， or continuous snoring for 3 minutes—the negative pressure was incrementally increased by 10 mmHg， up to a maximum safety threshold of -250 mmHg. The "Optimal Pressure" was defined as the minimum negative pressure that minimized respiratory events and stabilized oxygen saturation while the participant was in the supine position during Rapid Eye Movement (REM) sleep. 4. Research Results and Findings The 30 participants had a mean age of 51.27±13.97 years and a mean BMI of 25.87±3.42 kg/m2 . Following the personalized OPT titration protocol， the core physiological parameters demonstrated remarkable shifts: Primary Ventilatory Efficacy (Respiratory Events and Oxygenation): The cohort′s mean Apnea-Hypopnea Index (AHI) dropped dramatically from a baseline of 39.59 events/hr to a post-treatment titrated level of 8.17 events/hr， representing a highly statistically significant overall AHI reduction rate of 80.4% (p < 0.001). Defining an "Excellent Response" as a post-treatment AHI < 5 events/hr， a total of 25 participants successfully achieved this clinical benchmark， yielding an overall clinical success rate of 83.3%. Concurrently， the mean minimum oxygen saturation (Min SpO2) significantly improved from 80.93% to 83.70% (p = 0.023). Secondary Efficacy (Sleep Architecture and Daytime Symptoms): The proportion of stage N3 non-rapid eye movement sleep (deep slow-wave sleep) increased significantly from 14.89% to 23.23% (p = 0.003)， demonstrating a substantial structural optimization of sleep quality. Furthermore， the overnight Arousal Index fell significantly from 48.47 events/hr to 26.23 events/hr (p < 0.001)， successfully reversing micro-sleep fragmentation. This robust objective improvement directly translated into an enhanced daytime quality of life， as evidenced by a substantial decrease in ESS scores from 10.61 to 5.52 (p < 0.001). Safety and Physiological Neutrality: No statistically significant difference was observed between the pre- and post-treatment Periodic Limb Movement (PLM) Index (p = 0.816). This confirms that even when the negative pressure is titrated up to the upper physiological safety limit (-250 mmHg)， the mechanical traction remains highly localized to the specific upper airway structures and retains excellent physiological neutrality， eliminating the risk of treatment-induced movement disorders. Minor side effects， such as transient mucosal irritation， local soreness， or initial hypersalivation under high negative pressures， were recorded in a small subset of patients but resolved completely upon device removal or continued adaptation. 5. Conclusions and Contributions This doctoral thesis successfully demonstrates a distinct "dose-response relationship" inherent to Intra-oral Negative Pressure Therapy; executing a personalized titration protocol capable of scaling negative pressures allows clinicians to effectively overcome the therapeutic failures previously caused by individual variations in tissue compliance. Biomechanically， when operating at an individualized optimal pressure， the negative pressure is completely contained within the sealed oral cavity. By pneumatically isolating the oropharynx from the nasopharynx， OPT avoids introducing any external expiratory resistance or adding to the patient′s work of breathing (WOB)—a massive physiological advantage over positive-pressure systems. Ultimately， this empirical research bridges a major translational gap， transitioning OPT from an empirical alternative into a standardized， evidence-based clinical modality within precision sleep medicine. It establishes an objective， high-compliance， and comfortable frontline clinical pathway for CPAP-intolerant patients， while providing the foundational biomechanical parameter matrices necessary for developing next-generation intelligent， closed-loop auto-adjusting (Auto-OPT) medical devices. Keywords: Obstructive Sleep Apnea， Intra-oral Negative Pressure Therapy， Personalized Titration， Sleep Fragmentation， Precision Sleep Medicine.

以深度學習為基礎之鼓膜穿孔影像分析架構：術前傳導性聽力損失評估與第一型鼓膜成形術後聽力預後預測;A Deep Learning–Based Framework for Tympanic Membrane Perforation Image Analysis: Preoperative Assessment of Conductive Hearing Loss and Prediction of Postoperative Hearing Outcomes After Type I Tympanoplasty

Tue, 20 Oct 2026 08:34:07 GMT

title: 以深度學習為基礎之鼓膜穿孔影像分析架構：術前傳導性聽力損失評估與第一型鼓膜成形術後聽力預後預測;A Deep Learning–Based Framework for Tympanic Membrane Perforation Image Analysis: Preoperative Assessment of Conductive Hearing Loss and Prediction of Postoperative Hearing Outcomes After Type I Tympanoplasty abstract: 研究背景：聽力損失為全球重要公共衛生議題，其中慢性中耳炎所造成之鼓膜穿孔為臨床常見且可治療之病因。鼓膜穿孔會破壞外耳與中耳之聲壓傳遞機制，造成傳導性聽力損失，臨床上多以氣骨導差（air-bone gap， ABG）量化。純音聽力檢查雖為評估聽力之標準方法，但其執行仰賴病人配合、專業設備與適當檢查環境；在幼童、高齡者、認知功能受損者、基層醫療或資源受限場域中，常無法即時取得可靠聽力資料。相較之下，耳鏡與耳內視鏡影像為臨床例行取得之資料，能直接呈現鼓膜穿孔之大小、位置與形態，具有轉化為功能性評估資料來源之潛力。研究目的：本論文旨在建立一套以深度學習為基礎之鼓膜穿孔影像分析架構，將耳鏡影像中可觀察之鼓膜與穿孔形態轉化為可量化且可重現之幾何特徵，並進一步應用於兩個臨床問題：其一為術前傳導性聽力損失之 ABG 評估，其二為第一型鼓膜成形術後聽力預後之預測。透過此一架構，本研究嘗試將傳統以質性判讀為主之鼓膜影像，延伸為可支援聽力功能評估與臨床決策輔助之量化資料來源。研究方法：本研究採回溯性設計，整體資料包括三個主要部分。第一，納入 1，164 張門診鼓膜內視鏡影像，包括 1，014 張完整鼓膜影像與 150 張鼓膜穿孔影像，用於建立 Mask R-CNN 深度學習分割模型，以自動分割鼓膜與穿孔區域。第二，納入具有鼓膜穿孔與聽力資料之 225 張影像，包括 150 張門診穿孔影像與 75 張手術室穿孔影像，用於萃取穿孔面積比例、穿孔質心距離、穿孔至聽小骨參考點相對距離及穿孔角度等幾何特徵，並建立術前 ABG 預測模型。第三，納入 2020 至 2024 年間接受經耳道內視鏡第一型鼓膜成形術且術後達解剖成功之單純性鼓膜穿孔病人 121 例，結合年齡、術前 ABG 與術前影像幾何特徵，建立 ABG closure 與術後 ABG 之預後預測模型。模型表現分別以分割效能指標、相關性分析、迴歸模型效能、內部驗證及獨立測試資料進行評估。研究結果：深度學習分割模型於鼓膜與穿孔區域分割皆展現良好表現，並可產生與人工標註相近之影像幾何特徵。術前 ABG 分析顯示，穿孔面積比例為影響傳導性聽力損失之主要影像因子，與 ABG 呈正相關；位置相關特徵則提供次要但具解釋意義之空間資訊。在術前 ABG 預測方面，二次迴歸模型較單純理論模型具有較佳表現，顯示影像幾何特徵可有效連結鼓膜穿孔形態與傳導性聽力損失。在術後聽力預後模型中，Elastic Net 模型於獨立測試資料集中預測 ABG closure 之 R² 為 0.418、RMSE 為 4.39 dB；使用 Mask R-CNN 自動特徵時，R² 為 0.421、RMSE 為 4.38 dB。術後 ABG 預測之 R² 為 0.363、RMSE 為 4.36 dB；使用自動特徵時，R² 為 0.366、RMSE 為 4.35 dB，顯示自動化影像特徵可達與人工標註相近之預測表現。結論：本論文建立一套由鼓膜影像取得、深度學習自動分割、幾何特徵萃取、術前 ABG 預測至術後聽力預後評估之整合性影像分析架構。研究結果顯示，鼓膜穿孔影像中所蘊含之幾何資訊，尤其是穿孔面積比例，與傳導性聽力損失具有明確關聯；結合術前聽力資料與影像幾何特徵後，亦可合理預測第一型鼓膜成形術後之聽力改善與殘餘 ABG。本研究證實耳鏡影像可由傳統質性判讀工具，進一步轉化為可量化、可重現且具臨床決策輔助潛力之功能性資料來源，並為人工智慧於耳科聽力功能評估與手術預後預測之轉譯應用提供實證基礎。;Background: Hearing loss is a major global public health issue. Tympanic membrane perforation caused by chronic otitis media is a common and treatable cause of conductive hearing loss. Tympanic membrane perforation disrupts the sound pressure transmission mechanism between the external and middle ear， leading to conductive hearing loss， which is commonly quantified using the air-bone gap (ABG). Although pure-tone audiometry is the standard method for hearing assessment， its performance depends on patient cooperation， specialized equipment， and an appropriate testing environment. In children， older adults， patients with cognitive impairment， primary care settings， or resource-limited environments， reliable audiometric data may not always be immediately available. In contrast， otoscopic and otoendoscopic images are routinely obtained in clinical practice and directly demonstrate the size， location， and morphology of tympanic membrane perforations， suggesting their potential as a functional source of information for hearing assessment. Objective: This dissertation aimed to develop a deep learning–based image analysis framework for tympanic membrane perforations. The framework was designed to transform otoscopic image findings into quantifiable and reproducible geometric features and to apply these features to two clinically relevant tasks: preoperative assessment of conductive hearing loss using ABG prediction and prediction of postoperative hearing outcomes after type I tympanoplasty. Through this integrated framework， this dissertation sought to extend otoscopic images from qualitative visual interpretation to quantitative data that may support functional hearing assessment and clinical decision-making. Methods: This retrospective study consisted of three main datasets and analytical components. First， 1，164 outpatient otoendoscopic images， including 1，014 intact tympanic membrane images and 150 perforated tympanic membrane images， were used to develop a Mask R-CNN deep learning segmentation model for automatic segmentation of the tympanic membrane and perforation regions. Second， 225 perforation images with corresponding audiometric data， including 150 outpatient perforation images and 75 intraoperative perforation images， were used to extract geometric features， including perforation area ratio， perforation centroid distance， relative distance from the perforation to the ossicular reference point， and perforation angle， for preoperative ABG prediction. Third， 121 patients with simple tympanic membrane perforations who underwent transcanal endoscopic type I tympanoplasty between 2020 and 2024 and achieved anatomical success were included for postoperative prognostic modeling. Age， preoperative ABG， and preoperative image-derived geometric features were integrated to predict ABG closure and postoperative ABG. Model performance was evaluated using segmentation metrics， correlation analyses， regression model performance， internal validation， and independent test datasets. Results: The deep learning segmentation model demonstrated good performance in segmenting both the tympanic membrane and perforation regions， and the automated image-derived geometric features were comparable to those obtained from manual annotations. In the preoperative ABG analysis， perforation area ratio was the primary image-derived factor associated with conductive hearing loss and showed a positive correlation with ABG， whereas location-related features provided secondary but interpretable spatial information. For preoperative ABG prediction， the quadratic regression model outperformed the theoretical model， indicating that image-derived geometric features can effectively link tympanic membrane morphology with conductive hearing loss. In the postoperative prognostic model， the Elastic Net model achieved an R² of 0.418 and an RMSE of 4.39 dB for predicting ABG closure in the independent test dataset. When Mask R-CNN–derived automated features were used， the R² was 0.421 and the RMSE was 4.38 dB. For postoperative ABG prediction， the Elastic Net model achieved an R² of 0.363 and an RMSE of 4.36 dB， whereas the model using automated features achieved an R² of 0.366 and an RMSE of 4.35 dB. These findings indicate that automated image-derived features can achieve predictive performance comparable to manual annotations. Conclusions: This dissertation established an integrated image analysis framework encompassing tympanic membrane image acquisition， deep learning–based automated segmentation， geometric feature extraction， preoperative ABG prediction， and postoperative hearing outcome assessment. The results demonstrate that geometric information embedded in tympanic membrane perforation images， particularly perforation area ratio， is clearly associated with conductive hearing loss. When combined with preoperative audiometric data， image-derived geometric features can also reasonably predict hearing improvement and residual ABG after type I tympanoplasty. This study supports the transformation of otoscopic images from conventional qualitative interpretation into quantifiable， reproducible， and clinically meaningful functional data with potential value for decision support， and provides empirical evidence for the translational application of artificial intelligence in otologic functional assessment and surgical prognostication.

基於向量增量編碼之高效能 NoSQL 文件版本控制系統研究與實現;基於向量增量編碼之高效能 NoSQL 文件版本控制系統研究與實現

Tue, 20 Oct 2026 08:33:44 GMT

title: 基於向量增量編碼之高效能 NoSQL 文件版本控制系統研究與實現;基於向量增量編碼之高效能 NoSQL 文件版本控制系統研究與實現 abstract: 在大數據環境下，醫療照護與工業製造領域對於資料全生命週期管理的需求日益增加，特別是需具備法律效力與可追溯性的歷史紀錄。現有資料庫解決方案處理高頻率且微幅修改的半結構化資料時，仍存在顯著瓶頸：傳統關聯式資料庫因正規化設計導致歷史查詢產生大量關聯（Join）開銷；而 NoSQL 資料庫多採全量快照策略，導致儲存成本隨版本數量增加而大幅上升。本研究提出一種「基於向量的增量儲存架構」（Vector-Based Delta Storage Architecture），將文件演變過程抽象化為有序的結構化差異向量，並結合 Amazon DynamoDB 的單一資料表設計，實現高效的儲存模型。透過導入確定性的正向重建與反向回滾演算法，並配合自適應基準重置策略，確保查詢延遲不會隨歷史深度線性增長。實驗結果顯示，在百萬級紀錄規模下，本架構較傳統系統降低 60% 的查詢延遲（平均 0.29 秒），並節省約 80% 的儲存空間。本架構一方面解決了 NoSQL 版本控制的效能權衡問題，並透過 PC 製造業工程變更管理與 HL7 FHIR 醫療報告案例，驗證了其在滿足 FDA 21 CFR Part 11 合規性要求上的實用性。;In the era of big data， domains such as healthcare and industrial manufacturing increasingly demand precise lifecycle management of data， particularly for legally binding and traceable historical records. However， existing database solutions face significant bottlenecks when handling high-frequency， fine-grained updates in semi-structured data. Traditional Relational Database Management Systems (RDBMS) suffer from excessive join overhead during historical queries due to high normalization， while mainstream NoSQL document databases typically rely on full-snapshot strategies， leading to exponential growth in storage costs as versions accumulate. To address these challenges， this dissertation proposes a Vector-Based Delta Storage Architecture. We abstract the document evolution process into a sequence of ordered， structured delta vectors and integrate them with a Single-Table Design in Amazon DynamoDB to achieve a schema-agnostic and efficient storage model. The architecture introduces deterministic forward reconstruction and reverse rollback algorithms， coupled with an adaptive baseline reset policy， ensuring that query latency does not grow linearly with historical depth. Experimental results on a synthetic dataset of one million records demonstrate that， compared to traditional relational systems， the proposed architecture reduces query latency for historical version reconstruction by 60% (average 0.29 seconds) and decreases storage space requirements by approximately 80%. Furthermore， the system exhibits superior scalability and stability under high-concurrency write workloads. This research not only resolves the trade-off between storage efficiency and performance in NoSQL version control but also validates the architecture′s practical value through case studies in PC manufacturing Engineering Change Orders (ECO) and HL7 FHIR medical reports. It effectively supports strict compliance requirements， such as FDA 21 CFR Part 11， and robust audit trails.

成人與青少年皮肌炎轉錄體途徑的生物資訊分析及潛在藥物預測;Bioinformatic Analysis of Transcriptomic Pathways in Adult and Juvenile Dermatomyositis with Potential Drug Prediction

Fri, 06 Mar 2026 10:53:08 GMT

title: 成人與青少年皮肌炎轉錄體途徑的生物資訊分析及潛在藥物預測;Bioinformatic Analysis of Transcriptomic Pathways in Adult and Juvenile Dermatomyositis with Potential Drug Prediction abstract: 皮肌炎（Dermatomyositis, DM）與幼年型皮肌炎（Juvenile Dermatomyositis, JDM）為一類罕見的特發性發炎性肌肉病變（Idiopathic Inflammatory Myopathies, IIMs），其特徵為骨骼肌慢性發炎與免疫反應導致的組織損傷。雖然成人與幼年型皮肌炎在臨床表現及病理組織學上有許多相似之處，但近年研究顯示兩者在病生理機轉上也存在差異。本研究目的為使用生物資訊學方法，系統性分析成人與幼年型皮肌炎的共同與特異性轉錄體特徵，並利用藥物篩選平台(drug screening platform)預測潛在的治療候選化合物。本研究自 Gene Expression Omnibus (GEO) 資料庫取得肌肉切片微陣列資料集，包括成人皮肌炎 GSE128470、GSE5370、GSE2044 及幼年型皮肌炎 GSE11971、GSE3307。利用 TAC軟體進行批次校正與差異表達基因分析（DEG analysis）。經由 Gene Ontology (GO) 與 Kyoto Encyclopedia of Genes and Genomes (KEGG) 進行功能富集分析，並透過 STRING 與 Cytoscape 建立蛋白質交互作用網路（PPI network）及篩選樞紐基因（hub genes）。候選藥物則以 LINCS L1000CDS2 平台進行預測。結果顯示，成人皮肌炎共鑑定出 850 個差異表達基因，主要富集於抗病毒、發炎與免疫相關路徑，包括第一型干擾素反應、MAPK 及 NOD-like receptor 訊號途徑。幼年型皮肌炎共鑑定出 2,544 個差異表達基因，顯示出干擾素驅動的先天免疫活化、細胞黏附、血管新生及 PI3K–Akt 訊號上調。兩組比較後共有 437 個重疊基因，顯示兩者皆具有干擾素介導之「類病毒免疫表現型」；然而，幼年型皮肌炎呈現更明顯的血管重塑與細胞黏附。樞紐基因分析指出 STAT1, DDX58, IFIH1, CXCL10, ISG15, MX1, GBP1, EIF2AK2 為疾病關鍵調控基因。藥物預測結果中，XMD-1150 與 KM 00927 同時被預測為成人與幼年型皮肌炎的共同候選治療化合物。綜合而言，本研究以整合轉錄體分析闡明了成人與幼年型皮肌炎之共同與差異的分子機制，指出干擾素介導的免疫活化為兩者共同核心特徵，而血管與細胞黏附訊號的上調則以幼年型皮肌炎較為顯著。所預測之候選藥物提供了後續實驗驗證與精準治療開發的潛在方向，對於難治型皮肌炎具有重要臨床應用價值。 ;Dermatomyositis (DM) and juvenile dermatomyositis (JDM) are rare idiopathic inflammatory myopathies characterized by chronic muscle inflammation and immune-mediated tissue injury. Despite overlapping clinical and histopathological features, increasing evidence suggests that adult and juvenile DM also possess distinct features of pathogenesis. This study aimed to systematically elucidate the shared and unique transcriptomic landscapes of DM and JDM and to predict potential therapeutic candidates through integrative bioinformatics and drug repurposing approaches. Publicly available microarray datasets of muscle biopsy specimens (GSE128470, GSE5370, and GSE2044 for DM; GSE11971 and GSE3307 for JDM) were obtained from the Gene Expression Omnibus (GEO) database. Batch correction and differential expression analyses were conducted using the Transcriptome Analysis Console (TAC) software. Functional enrichment analyses based on Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were performed to identify key biological pathways, while protein–protein interaction (PPI) networks and hub gene analyses were conducted using STRING and Cytoscape. Candidate compounds were predicted via the LINCS L1000CDS2 platform. In DM, 850 DEGs were identified, predominantly enriched in antiviral, inflammatory, and immune signaling pathways, including type I interferon response, MAPK, and NOD-like receptor signaling. In JDM, 2,544 DEGs were identified, highlighting interferon-driven innate immunity, cell adhesion, angiogenesis, and PI3K–Akt signaling. Comparative analyses revealed 437 overlapping DEGs, underscoring a shared interferon-mediated viral-like immune signature across age groups, while JDM exhibited enhanced vascular remodeling and adhesion-related processes. Hub gene analysis identified STAT1, DDX58, IFIH1, CXCL10, ISG15, MX1, GBP1, EIF2AK2 as central regulators. Drug repurposing predicted XMD-1150 and KM 00927 as potential shared therapeutic candidates for DM and JDM. In conclusion, this integrative transcriptomic study delineates both shared and distinct molecular pathways in DM and JDM, highlighting interferon-driven immune activation as a common hallmark and vascular-adhesion signaling as a distinguishing feature in JDM. The predicted compounds provide a foundation for future experimental validation and development of targeted therapies in refractory dermatomyositis.