類別:

遲發性運動障礙辨識之多模態機器學習框架：基於語音與臉部動作分析

Thu, 01 Jan 2026 00:00:00 GMT

標題: 遲發性運動障礙辨識之多模態機器學習框架：基於語音與臉部動作分析; A Multimodal Machine Learning Framework for Distinguishing Tardive Dyskinesia via Voice and Facial Movement Analysis 作者: 許漢強; Dwayne Reinaldy 摘要: 遲發性運動障礙(Tardive Dyskinesia, TD)是一種由藥物引起的運動障礙,估計影響20%至30%接受長期抗精神病藥物治療的患者,但在臨床實踐中仍普遍存在診斷不足的問題。TD患者的言語障礙已被臨床記錄超過三十年,但目前尚無自動化系統用於分析TD患者的言語特徵。本論文提出了一種多模態機器學習框架,透過整合來自簡短、真實場景視訊錄製中的語音和面部分析,實現TD的自動化篩檢。該框架結合了基於解耦機制的音訊模組與3D面部運動學分析模組,並透過跨模態注意力機制將兩者連接,使兩種模態能夠動態互動。我們在包含80名受試者的資料集上評估了該框架,受試者分為三類:TD、帕金森病(Parkinson's Disease)和健康對照組,並採用受試者獨立的交叉驗證方法。該多模態模型在健康對照組與TD的二分類任務中取得了0.93的AUROC值,在三分類任務中取得了0.89的AUROC值,在所有任務中均持續優於僅使用面部或僅使用音訊的基線模型。我們還採用了可解釋人工智慧(Explainable AI, XAI)方法,以提供可解釋的預測結果,將模型的注意力機制與已知TD言語障礙相關的語音片段聯繫起來。總體而言,我們的研究填補了一個持續三十餘年的診斷空白,並朝著在真實臨床環境中實現客觀TD篩檢邁出了重要一步。; Tardive Dyskinesia (TD) is a medication-induced movement disorder affecting an estimated 20-30% of patients on long-term antipsychotic treatment, yet remains widely underdiagnosed in clinical practice. Speech impairment in TD has been clinically documented for over three decades, yet no automated system has been used for analyzing speech in TD. This thesis presents a multimodal machine learning framework for automated TD screening by integrating voice and facial analysis from brief, in-the-wild video recordings. The proposed framework combines a disentanglement mechanism-based audio module with a 3D facial kinematic analysis module connected via cross-modal attention, enabling the two modalities to interact dynamically. We evaluated this framework on a dataset of 80 subjects across three classes: TD, Parkinson's Disease, and healthy controls under subject-independent cross-validation. The multimodal model achieved an AUROC of 0.93 for healthy control versus TD discrimination and 0.89 for the three-class classification task, consistently outperforming the face and audio only baseline across all tasks. We also employed Explainable AI (XAI) to provide interpretable predictions, linking model attention to speech segments consistent with known speech impairments in TD. Overall, our work addresses a diagnostic gap that has persisted for over three decades and represents a step toward objective TD screening in real-world clinical settings.

透過可解釋的認知評估，研究轉移、少樣本和元學習在基於靜息態和聽覺穩定狀態反應腦電圖的神經退化性疾病診斷之生物標記識別

Mon, 01 Jan 2024 00:00:00 GMT

標題: 透過可解釋的認知評估，研究轉移、少樣本和元學習在基於靜息態和聽覺穩定狀態反應腦電圖的神經退化性疾病診斷之生物標記識別; Investigating Transfer, Few-Shot, and Meta Learning for Biomarker Identification in Resting-State and Auditory Steady-State Response EEG-Based Diagnosis of Neurodegenerative Diseases with Interpretable Cognitive Assessments 作者: 楊美美; DANIELLE PENELLA P. YU 摘要: 失智症患者的靜息態和聽覺穩定狀態反應 (ASSR) 相關腦電圖 (EEG) 訊號，如腦電圖減慢、伽瑪相位鎖定和功率調變改變，通常能揭露患者的認知異常。傳統上，腦電圖的分析涉及使用功率譜密度(PSD) 地形圖、試驗間一致性(ITC) 計算和事件相關頻譜擾動(ERSP) 分析等技術將訊號轉換為空間頻率和時間頻率之影像表示。這樣的方法可以凸顯關鍵的細微差別以進行準確的診斷和解釋。隨著深度學習 (DL) 在推進診斷的關鍵解決方案展露頭角，透過可解釋的人工智慧 (xAI) 進行，它為傳統方法提供了一種可行、高效且重要的替代方案。而深度學習也揭示了腦電圖作為篩檢工具並協助找出神經退化性疾病個體生物標記的能力。在這項研究中，我們專注於整合遷移學習、小樣本學習 (FSL) 和元學習技術，以最大化深度學習的潛力。我們收集了阿茲海默症 (AD) 、罕見疾病額顳葉失智症 (FTD) 和精神分裂症 (SZ)患者的腦電圖數據，以研究與失智症相關的差異特徵。閉眼時的靜息態腦電圖資料收集自 15 位AD患者、15 位 FTD 患者、16位SZ 患者以及 15位健康對照組 (HC)。此外，針對 ASSR 數據集我們收集了以40 Hz 刺激的數據，其中包括 10 名SZ 患者和 15 名 HC。我們的方法採用預訓練的捲積神經網路 (CNN) 以及 miniImageNet 資料集來有效地分析不同腦電圖資料集。我們結合了FSL 和元學習，以增強模型在2-way 5-shot分類中的泛用性和適應性，其中每個episode 涉及從兩個類別中隨機採樣五個樣例，並在100 多個episode的穩健測試中對三個查詢進行測試。儘管神經退化性疾病研究中的數據可用性是個重大挑戰，而且我們的數據集規模相對較小，但相似的人類上皮細胞 2 型 (HEp-2 Cell) 數據集和血細胞計數和檢測 (BCCD）白血球數據集達到顯著的進步。其中，HEp-2 細胞資料集包含13,596 張影像，分為六類（著絲粒、高爾基體、均質、核仁、核膜和斑點型），而BCCD 白血球資料集包括五類（嗜中性球、淋巴細胞、單核細胞、嗜酸性粒細胞和嗜鹼性粒細胞），總共12,436張影像，達到了55.1%的準確率。相對 α 波段功率是靜息態腦電圖最顯著的特徵。使用含有殘差網路18 (ResNet-18) 的原型網路(prototypical network)和帶有ResNet-34 編碼器的匹配網路(matching network)，我們分別實現了82.83% 和76.33% 的平均準確度。其中最具辨別能力的類別對是AD 和SZ（94.44%和 92.22% 的準確率）。相較之下，我們的結果表明，P4 和 Fz 通道在 ASSR 間相干性方面實現了最高性能，平均準確度分別為 68.83% 和 70.00%。另一方面，與事件相關的光譜擾動發現T5 和 Fz 通道是有區別性的，兩個模型的平均表現為 71.17% 和 69.33%。為了更深入研究神經動力學，我們實作了梯度加權類活化映射 (Grad-CAM)，這是 xAI 中的一個突出方法。這種方法大大增強了我們對疾病狀態和病理機制中振盪行為的理解，並可能有助於生物標記的辨識。 Grad-CAM 顯示頂葉的相對 α 功率降低，以及與靜止狀態下認知障礙的潛在關聯性。此外，基於ASSR 的Grad-CAM視覺化揭示了誘發功率和鎖相中明顯的伽馬帶振盪，表明N-甲基-D-天冬氨酸受體(NMDAR) 拮抗劑可能引起伽馬氨基丁酸(GABA) 中間神經元功能障礙。證據顯示，在 ASSR 任務期間，Fz 通道 ITC和ERSP在Grad-CAM 上減少，顯示 SZ 和 HC 受試者存在異常的神經同步和頻譜動態模式。這些神經損傷的模式會影響神經網路的完整性，而能揭示特定的大腦區域如何導致疾病。總之，結合傳統腦電圖分析方法和新穎的深度學習技術使我們能夠解決神經退化性疾病研究中的關鍵挑戰。我們的方法不僅提高了其穩健性和有效性，還提高了複雜神經數據的可解釋性。因此，我們的方法能啟發新的臨床方案發展，以及加速新藥標的的發現。; Resting-state and auditory steady-state response (ASSR)-related electroencephalography (EEG) signals in individuals with dementia commonly reveal cognitive abnormalities, such as EEG slowing, gamma phase locking, and power modulation alterations. Traditionally, EEG analysis involves transforming signals into spatial frequency and time-frequency image representations using techniques via power spectral density (PSD) topographic maps, intertrial coherence (ITC) calculations, and event-related spectral perturbation (ERSP) analysis, highlighting crucial nuances essential for accurate diagnosis and interpretation. With the emergence of deep learning (DL) as a pivotal solution in advancing diagnostics, it offers a feasible, efficient, and crucial alternative to conventional methods for accurate differentiation, complemented by interpretation through explainable artificial intelligence (xAI). DL methods, in particular, reveal the ability of EEG to serve as a screening tool and assist in identifying biomarkers in individuals with neurodegenerative diseases. In this study, we focus on integrating transfer learning, few-shot learning (FSL), and meta-learning techniques to maximize the potential of DL. EEG data from individuals with conditions such as Alzheimer’s disease (AD), a rare disease frontotemporal dementia (FTD), and schizophrenia (SZ) were collected to investigate differential features associated with dementia. Resting-state EEG data from the eyes-closed protocol were collected from 15 patients diagnosed with AD, 15 with FTD, 16 with SZ, as well as from 15 healthy controls (HCs). In addition, 40-Hz stimulus data for the ASSR dataset, consisting of 10 individuals with SZ and 15 HCs, were obtained. Our approach adopted pretrained convolutional neural networks (CNNs) along with the miniImageNet dataset benchmark to analyze diverse EEG datasets effectively. We incorporated FSL and meta-learning to enhance the models’ generalizability and rapid adaptability in 2-way 5-shot classification, where each episode involved randomly sampling five support examples from two classes and robust testing on three queries over 100 episodes. Despite major challenges with data availability in neurodegenerative disease research and the relatively small size of our datasets, significant improvements have been observed in similar test scenarios involving the Human Epithelial Cells type 2 (HEp-2 Cell) dataset and the Blood Cell Count and Detection (BCCD) White Blood Cell dataset. The HEp-2 Cell dataset comprises 13,596 images divided into six classes (Centromere, Golgi, Homogeneous, Nucleolar, Nuclear Membrane, and Speckled). In contrast, the BCCD White Blood Cell dataset includes five classes (Neutrophil, Lymphocyte, Monocyte, Eosinophil, and Basophil), totaling 12,436 images, achieving its highest accuracy of 55.1%. Relative alpha band power is the most distinguishing feature of resting-state EEG. Using a prototypical network with residual network-18 (ResNet-18) and a matching network with ResNet-34 encoders, we achieved mean accuracies of 82.83% and 76.33%, respectively, with the most discriminative class pair being AD and SZ (94.44% and 92.22% accuracies). In contrast, our results indicate that the P4 and Fz channels achieve the highest performance in ASSR intertrial coherence, with average accuracies of 68.83% and 70.00%, respectively. On the other hand, event-related spectral perturbations identify the T5 and Fz channels as discriminative, averaging 71.17% and 69.33% for both models. To delve deeper into neural dynamics, we implemented gradient-weighted class activation mapping (Grad-CAM), a prominent feature attribution in xAI. This method has greatly enhanced our understanding of oscillatory behavior in disease states and pathological mechanisms, potentially aiding in biomarker identification. The Grad-CAMs showed decreased relative alpha power in the parietal lobes and its possible association with cognitive impairment in the resting state. Moreover, ASSR-based Grad-CAM visualizations revealed distinct gamma band oscillations in evoked power and phase locking, suggesting probable gamma-aminobutyric acid (GABA)ergic interneuron dysfunction caused by N-methyl-D-aspartate receptor (NMDAR) antagonists. The evidence showed a reduction in Fz channel ITC and ERSP Grad-CAMs during the ASSR task, indicating aberrant neural synchrony and spectral dynamics patterns across SZ and HC subjects. These patterns underlying neurological impairments affect neural network integrity, thereby revealing how specific brain regions contribute to diseases. In conclusion, combining traditional EEG analysis methods and innovative DL techniques has allowed us to address critical challenges in neurodegenerative disease research. Our methodology improves its robustness and effectiveness and the interpretability of complex neural data. Thus, our approach may inspire the development of new clinical practices and expedite the discovery of new drug targets.

解析 CAM 異種移植模型中 EMT 異質性的多模態分析平台：空間體學與計算驗證

Thu, 01 Jan 2026 00:00:00 GMT

標題: 解析 CAM 異種移植模型中 EMT 異質性的多模態分析平台：空間體學與計算驗證; A Multi-Modal Platform for Resolving EMT Heterogeneity in CAM Xenografts: Spatial Omics and Computational Validation 作者: Ifrah Taqdees; Ifrah Taqdees 摘要: 摘要背景：SNAI1 是一種上皮-間質轉化（EMT）的主控轉錄因子（TF），具有雙重轉錄作用；其中，賴胺酸殘基 K146/K187 的乙醯化會使其功能從抑制因子轉變為激活因子，從而調整 EMT 在不同情境下的分子執行機轉。目前學界對於此切換下游機制理解仍相當有限。本研究以高惡性度漿液性卵巢癌（HGSOC）作為疾病模型，建立一個整合空間轉錄組學（ST）、計算分析與多重免疫螢光（mIF）之平台，以解析由 SNAI1 驅動的上皮-間質轉化異質性。方法：將鉑類耐藥性高分化漿液性卵巢癌（HGSOC）細胞系 PEO4 經基因工程改造，使其分別表達 SNAI1-2R（乙醯化缺陷型）及空載向量（EV），再於雞胚絨毛尿囊膜（CAM）上建立腫瘤異種移植瘤。針對 PEO4-SNAI1-2R 與 PEO4-EV 異種移植瘤（每組 n=2），在 20 個具潛在影響區域內的 38 個 PanCK+/PanCK- 區段中，進行了 ST 分析（GeoMx DSP，癌症轉錄組圖譜；1.8k 個基因）。透過識別 PEO4-SNAI1-2R 與 PEO4-EV 之間的差異表達基因（DEGs），並將其輸入 MOSAIC（主力驗證已知上皮-間質轉化轉錄因子（EMT-TFs）直接靶標的整合式計算流程）進行分析。同時建立了一套微免疫螢光（mIF）工作流程，用以繪製 CAM 異種移植模型的細胞上皮-間質轉化狀態及腫瘤-宿主相互作用。結果：透過空間定量分析共鑑定出 15 個差異表達基因（DEGs）。其中，有 4 個 DEGs 經由整合式計算分析流程 MOSAIC 評定為 SNAI1 的直接靶基因。研究發現，相較於 EV 腫瘤，ALCAM 是 SNAI1-2R 腫瘤中 PanCK+ 區段上調程度最高的基因（log2FC = 0.93，FDR < 0.001），且經 MOSAIC 評定為信度最高的 SNAI1 靶基因。隔室層級分析則進一步解釋 SNAI1-2R 異種移植瘤中的腫瘤內之異質性，而在 EV 異種移植瘤中則未有見此現象。此外，多重成像技術識別出上皮-間質轉化（EMT）的異質性，其中腫瘤區域同時共表達上皮（EpCAM）與間質（Vimentin）標記物，暗示由 SNAI1 驅動的混合型 EMT 狀態。結論：綜上所述，本研究建立了一個概念驗證平台，將空間組學與計算證據相結合，以繪製高分化漿液性卵巢癌（HGSOC）中由 SNAI1 驅動的上皮-間質轉化（EMT）異質性圖譜。此框架為未來不同的 EMT 誘導模型及腫瘤背景提供了嶄新的方向。; Abstract Background: SNAI1 is a master epithelial-mesenchymal transition (EMT) transcription factor (TF) with dual transcriptional roles where the acetylation at lysine residue K146/ K187 switches its function from a repressor to an activator, contributing to the molecular execution of EMT in different contexts. The understanding of mechanistic programs downstream of this switch remains limited. Using high-grade serous ovarian carcinoma (HGSOC) as the disease model, this study establishes an integrated spatial transcriptomics (ST), computational and multiplex immunofluorescence (mIF) platform to resolve SNAI1-driven EMT heterogeneity. Methods: A platinum-resistant HGSOC cell line PEO4 was engineered to express SNAI1-2R (acetylation-deficient) and empty vector (EV). Their tumor xenografts were established in the chick chorioallantoic membrane (CAM). ST profiling (GeoMx DSP, Cancer Transcriptome Atlas; 1.8k genes) of the PEO4-SNAI1-2R and PEO4-EV xenografts (n=2 per group) was performed in 38 PanCK+/PanCK- segments across 20 regions of interest. Differentially expressed genes (DEGs) between PEO4-SNAI1-2R and PEO4-EV were identified and subjected to MOSAIC, an integrated computational pipeline for the verification of direct targets of known EMT-TFs. An mIF workflow was established to map the cellular EMT states and tumor-host interactions of the CAM xenografts. Results: A total of 15 DEGs were identified from spatial profiling. Among them, 4 DEGs were scored as direct SNAI1 targets by MOSAIC, an integrated computational pipeline. ALCAM was found to be the top upregulated gene in PanCK+ segments of SNAI1-2R versus EV tumors (log2FC = 0.93, FDR < 0.001) and scored as a SNAI1 target with highest confidence by MOSAIC. Compartment-level analysis further resolved intra-tumoral heterogeneity in SNAI1-2R but not EV xenografts. In addition, multiplex imaging identified EMT heterogeneity with tumor regions co-expressing epithelial (EpCAM) and mesenchymal (Vimentin) markers, suggestive of hybrid EMT states driven by SNAI1. Conclusion: Together, this study established a proof-of-concept platform linking spatial omics and computational evidence to map SNAI1-driven EMT heterogeneity in HGSOC. This framework provides the future roadmap to generalize across different EMT induction models and tumor contexts.

穿戴式裝置活動訊號在死亡風險分層的角色

Thu, 01 Jan 2026 00:00:00 GMT

標題: 穿戴式裝置活動訊號在死亡風險分層的角色; The Role of Wearable-Derived Activity Signals in Mortality Risk Stratification 作者: 陳萱如; Shiuan-Ju Chen 摘要: 穿戴式裝置可追蹤真實生活中的活動狀態，並可能提供傳統臨床評估較難捕捉的行為資訊。然而，穿戴式裝置衍生之活動模式在死亡風險分層中的角色仍未被充分了解。本研究旨在探討穿戴式裝置衍生之活動訊號於死亡風險分層與行為解釋中的應用價值。本分析使用美國國家健康與營養調查（National Health and Nutrition Examination Survey, NHANES）2011–2014 年資料，納入 4,531 名年齡 40 歲以上、具臨床資料、穿戴式裝置資料與死亡追蹤資料之成人。研究自七天逐小時之活動及睡眠相關訊號中萃取關於活動強度、變異性、破碎化及節律相關的行為特徵。風險分層模型分別以臨床變數、穿戴式裝置衍生特徵及兩者結合作為三種輸入設定，並以 XGBoost 作為主要模型。模型解釋透過 SHAP 值、部分依賴圖（partial dependence plots）及群體層級行為模式分析，評估特徵重要性與風險相關模式。結果顯示，整合穿戴式裝置衍生特徵與臨床變數可提供死亡風險分層的補充資訊，並能在預測風險較高之群體中辨識出較高比例的死亡個案。模型解釋辨識出數個重要的穿戴式裝置衍生行為訊號，包括峰值活動量（PAM_max）與活動破碎化程度（PAM_IV）。進一步的群體分析顯示，同時具有較低峰值活動能力與較高活動破碎化特徵的個體，其觀察到的死亡率最高。本研究指出，穿戴式裝置衍生之活動訊號可捕捉與死亡風險分層相關的行為面向。除傳統模型預測表現外，穿戴式裝置衍生之行為模式亦可提供具解釋性的行為資訊，並可能有助於族群層級健康風險評估方法之發展。; Wearable devices enable continuous assessment of real-world activity behavior and may provide behavioral information not captured by traditional clinical evaluation. However, the role of wearable-derived activity patterns in mortality risk stratification remains incompletely understood. This study investigated the utility of wearable-derived activity signals for mortality risk stratification and behavioral interpretation. We analyzed data from 4,531 adults aged ≥40 years from the 2011–2014 U.S. National Health and Nutrition Examination Survey (NHANES) with clinical data, linked mortality records, and accelerometer data. Wearable-derived features were extracted from seven-day hourly activity and sleep-related signals to characterize activity intensity, variability, fragmentation, and rhythm-related patterns. Risk stratification models were developed using clinical variables, wearable-derived features, and their combination, with XGBoost as the primary model. Model interpretation was assessed using SHAP values, partial dependence plots, and cohort-level behavioral pattern analyses. The integration of wearable-derived features with clinical variables provided complementary information for mortality risk stratification, particularly through AUROC-based discrimination and enrichment of observed mortality events in top-ranked risk groups. Interpretability analyses identified several wearable-derived features, including peak activity level (PAM_max) and activity fragmentation (PAM_IV), as important behavioral signals. Cohort-level analysis further showed that individuals characterized by both lower peak activity capacity and higher activity fragmentation exhibited the highest observed mortality rate. These findings suggest that wearable-derived activity signals capture a complementary behavioral dimension relevant to mortality risk stratification. Beyond conventional model performance metrics, wearable-derived behavioral patterns may provide interpretable information for population-level health risk assessment.

類別:

遲發性運動障礙辨識之多模態機器學習框架：基於語 音與臉部動作分析

透過可解釋的認知評估，研究轉移、少樣本和元學習在基於靜息態和聽覺穩定狀態反應腦電圖的神經退化性疾病診斷之生物標記識別

解析 CAM 異種移植模型中 EMT 異質性的多模態分析平台：空間體學與計算驗證

穿戴式裝置活動訊號在死亡風險分層的角色

遲發性運動障礙辨識之多模態機器學習框架：基於語音與臉部動作分析