![\"Source:](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/Operation-Smile-Rwanda-1024x681.jpg\")
Traditionally, speech-language pathologists assess hypernasality, but their limited availability and concentration in certain hospitals necessitate extensive, costly cross-regional patient travel. An automated hypernasality assessment method would not only aid pathologists in making accurate evaluations but also facilitate remote patient diagnosis and treatment, significantly reducing costs.<\/p>\n\n\n\n
Leveraging transfer learning technology, Microsoft Research Asia has developed an innovative approach using an automatic speech recognition (ASR) model to enhance hypernasality assessment. This innovative model excels in extracting acoustic features and demonstrates robust generalization capabilities. Comparative studies on two cleft palate datasets reveal that it surpasses existing methods, significantly enhancing the precision of pathologists\u2019 diagnostic processes.<\/p>\n\n\n\n
Following hypernasality evaluations, physicians devise tailored speech therapy regimens for patients. Microsoft Research Asia has advanced this process by developing the Masked Pre-training Pronunciation Assessment (MPA) model. This model supports end-to-end training and adapts to both unsupervised and supervised learning environment, enabling user-friendly remote deployment. Utilizing reference texts and integrating masking with prediction tactics, the MPA model adeptly circumvents issues of misalignment or misrecognition in pronunciation assessments, offering more precise speech correction support for individuals with cleft palate.<\/p>\n\n\n\n
Microsoft Research Asia is actively collaborating with healthcare providers to assess the feasibility of deploying this cutting-edge speech assessment technology. The goal is to enhance the efficiency of medical diagnoses and treatments, lower the financial burden on patients, and extend the benefits of this technology to numerous cleft palate sufferers in isolated regions.<\/p>\n\n\n\n
Related papers:<\/p>\n\n\n\n
- \n
- Paper on hypernasality detection:\u00a0<\/strong>Improving Hypernasality Estimation with Automatic Speech Recognition in Cleft Palate Speech (opens in new tab)<\/span><\/a>. INTERSPEECH 2022.<\/li>\n\n\n\n
- Paper on speech assessment:\u00a0End-to-End Word-Level Pronunciation Assessment with MASK Pre-training (opens in new tab)<\/span><\/a>. INTERSPEECH 2023.<\/li>\n<\/ul>\n\n\n\n
Voice analysis model enhances Alzheimer\u2019s disease screening<\/h4>\n\n\n\n
Alzheimer\u2019s disease, a prevalent neurodegenerative condition primarily affecting the elderly, leads to progressive cognitive decline, including memory loss, language difficulties, and impaired reasoning. While there\u2019s no cure for Alzheimer\u2019s, early detection and intervention are key to decelerating its progression.<\/p>\n\n\n\n
![\"pexels.com\"](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-2-1024x576.jpg\")
Source: pexels.com<\/em><\/figcaption><\/figure>\n\n\n\n Traditional diagnostic methods, such as brain scans, blood tests, and cognitive assessments, are extensive and expensive. However, research indicates that early Alzheimer\u2019s can be detected through speech analysis, identifying symptoms like fluent aphasia and word retrieval challenges.<\/p>\n\n\n\n
Capitalizing on this insight, Microsoft Research Asia has pioneered speech and language analysis technologies to detect Alzheimer\u2019s indicators from sophisticated acoustic and linguistic data. A novel task-oriented model has also been introduced, correlating language descriptions with cognitive tasks.<\/p>\n\n\n\n
In the ADReSS dataset\u2019s subtask involving \u201cCookie Theft\u201d picture descriptions (Figure 1), these methods attained a 91.4% accuracy rate. This innovative approach, merging speech and semantic analysis, significantly increases disease detection accuracy. The model\u2019s high efficiency and performance on new test sets offer promising prospects for Alzheimer\u2019s screening at scale.<\/p>\n\n\n\n
![\"Figure](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-3.png\")
Figure 1: \u201cCookie Theft\u201d used for the descriptive task of detecting Alzheimer\u2019s disease, presented by DementiaBank Pitt Corpus, Becker et al., in 1994.<\/em><\/figcaption><\/figure>\n\n\n\n Related paper:<\/p>\n\n\n\n
- \n
- Leveraging Pretrained Representations with Task-related Keywords for Alzheimer’s Disease Detection (opens in new tab)<\/span><\/a>. ICASSP 2023.<\/li>\n<\/ul>\n\n\n\n
Advancing autism diagnosis: Unsupervised detection of stereotypical behaviors<\/h4>\n\n\n\n
Autism Spectrum Disorder (ASD) typically manifests in early childhood, presenting challenges in social interaction and communication, accompanied by repetitive behaviors. These behaviors, which may include actions like persistent hand-flapping or head-banging, serve as vital indicators for ASD diagnosis. Early detection and intervention are crucial for improving outcomes, yet traditional methods relying on prolonged observation by specialists are not always efficient. Hence, the development of a swift, automated detection system is invaluable.<\/p>\n\n\n\n
![\"Source:](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-4-1024x683.jpg\")
Source: unsplash.com<\/em><\/figcaption><\/figure>\n\n\n\n Traditional approaches have utilized computer vision and supervised learning to analyze video data of individuals with ASD. However, these methods face limitations due to the diverse range of stereotypical behaviors and privacy concerns associated with video data collection.<\/p>\n\n\n\n
Addressing these challenges, Microsoft Research Asia, in collaboration with medical institutions, has innovated an unsupervised approach using video anomaly recognition. The new\u00a0Dual-Stream Stereotypical Behavior Detector (DS-SBD)\u00a0model leverages the temporal dynamics of human posture and repetitive motion patterns. Remarkably, DS-SBD requires only non-anomalous behavior for training and can identify previously unseen stereotypical behaviors during inference, such as identifying circling behaviors in the training data.<\/p>\n\n\n\n
![\"Figure](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-5.gif\")
Figure 2:\u00a0The DS-SBD model\u2019s predictive accuracy spikes when detecting atypical behaviors such as abnormal hand clapping.<\/em><\/figcaption><\/figure>\n\n\n\n Extensive studies validate that DS-SBD\u2019s unsupervised technique has increased the micro-average AUROC from 60.43% to 71.04% and the macro-average AUROC from 56.45% to 73.39%, signifying a substantial improvement in both accuracy and the scope of detectable behaviors. This breakthrough outperforms current state-of-the-art methods and is poised to set a new standard in the field. While DS-SBD marks a significant advancement in recognizing stereotypical behaviors, it represents only one facet of the broader ASD diagnostic process. Comprehensive early diagnosis and intervention strategies will benefit from continued interdisciplinary collaboration and societal engagement.<\/p>\n\n\n\n
Related paper:<\/p>\n\n\n\n
- \n
- Unsupervised Video Anomaly Detection for Stereotypical Behaviors in Autism (opens in new tab)<\/span><\/a>. ICASSP 2023.<\/li>\n<\/ul>\n\n\n\n
Advancing neonatal seizure detection through brainwave analysis<\/h4>\n\n\n\n
Epilepsy in children is a multifaceted, often recurring neurological disorder that predominantly occurs in the formative years (0-18 years). The prompt identification of epilepsy in newborns is vital to safeguard their developmental trajectory.<\/p>\n\n\n\n
![\"Source:](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-6-1024x678.jpg\")
Source: unsplash.com<\/em><\/figcaption><\/figure>\n\n\n\n The genesis of epileptic seizures lies in the abnormal discharges of neurons within the brain, rendering brainwave analysis a pivotal tool for epilepsy diagnosis. Nonetheless, the nascent state of neonatal brain development, coupled with the pronounced noise in brainwave data and the marked variability among infants, renders the detection of neonatal epilepsy a formidable medical challenge.<\/p>\n\n\n\n
Microsoft Research Asia and its collaborators have unveiled a deep learning paradigm, harnessing artificial intelligence and electroencephalogram (EEG) signals – dubbed the Spatial-Temporal EEG Network (STATENet). This model adeptly processes neural signals, nimbly adjusts to neonatal EEG channel variations, and addresses the challenges outlined above. Additionally, the team has introduced a model-level integration technique that synergistically combines outcomes from various spatial-temporal deep models, thereby bolstering the STATENet model\u2019s generalization ability across diverse neonatal subjects.<\/p>\n\n\n\n
Extensive studies utilizing a comprehensive dataset of real-world neonatal EEG data have demonstrated the STATENet model\u2019s substantial enhancement in detection precision. The model\u2019s area under the precision-recall curve (AUPRC) witnessed an improvement exceeding 30% relative to prevailing top-tier methods, equipping physicians with a novel diagnostic instrument for pediatric epilepsy.<\/p>\n\n\n\n
Moreover, Microsoft Research Asia has pioneered the inaugural cross-dataset EEG model capable of deciphering any EEG data, thus achieving a \u2018one-to-many\u2019 brainwave comprehension. This breakthrough underpins the AI Neurologist system, designed to augment brainwave signal analysis in both clinical and research settings, elevating diagnostic accuracy from 75% to 90% in a case study. The associated models are now open source on GitHub, inviting global research participation to extend this technology\u2019s impact across the medical spectrum and catalyze new diagnostic and therapeutic innovations.<\/p>\n\n\n\n
![\"Figure](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/image003.gif\")
Figure 3: The AI Neurologist system<\/em><\/figcaption><\/figure>\n\n\n\n Related papers:<\/p>\n\n\n\n
- \n
- Paper on neonatal seizure detection: Protecting the Future: Neonatal Seizure Detection with Spatial-Temporal Modeling (opens in new tab)<\/span><\/a>. IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2023.<\/li>\n\n\n\n
- Paper on brainwave pre-training model: Learning Topology-Agnostic EEG Representations with Geometry-Aware Modeling. (opens in new tab)<\/span><\/a><\/li>\n<\/ul>\n\n\n\n
Enhancing disease progression prediction and personalized care: The role of AI in precision medicine<\/h3>\n\n\n\n
Precision medicine represents a transformative approach to healthcare, tailoring treatments to individual patient profiles. Despite the promise, the complexity and unique nature of diseases present significant hurdles. AI emerges as a powerful ally, leveraging data analysis, pattern detection, and predictive modeling to forecast disease progression and risks. This capability is especially crucial in chronic disease management, aiding clinicians and patients in mitigating illness severity and preventing complications.<\/p>\n\n\n\n
Graph neural networks: A novel approach to Parkinson\u2019s disease progression<\/h4>\n\n\n\n
Parkinson\u2019s disease, a prevalent neurodegenerative condition in seniors, progresses gradually. Patient conditions may remain stable or even improve over time with proper medication and therapy, maintaining optimal physical functions. Yet, Parkinson\u2019s presents a spectrum of symptoms, from sleep disturbances to motor challenges, making disease progression prediction complex.<\/p>\n\n\n\n
![\"Source:](\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2024\/05\/world-health-day-2024-8-1024x682.jpg\")
Source: pixabay.com<\/em><\/figcaption><\/figure>\n\n\n\n Researchers at Microsoft Research Asia advocate for the analysis of multimodal data to extract similar symptoms, thus enhancing prediction accuracy. Graph neural networks (GNNs) excel in mapping patient interconnections, forming networks where nodes represent patients linked by shared attributes. Selecting these attributes, however, largely demands expert knowledge and experience.<\/p>\n\n\n\n
To overcome this, Microsoft Research Asia collaborated closely with medical institutions. Based on recommendations from professional medical personnel, a new algorithm called AdaMedGraph was proposed. AdaMedGraph autonomously identifies key features to construct patient similarity graphs, harmonizing with existing knowledge and integrating expert-designed graphs into a comprehensive model. Unifying individual and collective data, this innovation simplifies the graph construction process.<\/p>\n\n\n\n
Evaluated on two public datasets, the Parkinson\u2019s Progression Markers Initiative (PPMI) and the Parkinson\u2019s Disease Biomarkers Program (PDBP), AdaMedGraph outperformed benchmarks in predicting Parkinson\u2019s progression over 24 months, setting the stage for personalized treatment strategies.<\/p>\n\n\n\n
Moreover, AdaMedGraph\u2019s robust generalization ability shines in metabolic syndrome prediction, achieving an AUROC of 0.675 on test datasets. This underscores the model\u2019s efficacy in integrating intra- and inter-patient data for individual disease progression forecasting, inspiring new avenues in medical research.<\/p>\n\n\n\n
Related paper:<\/p>\n\n\n\n
- Paper on brainwave pre-training model: Learning Topology-Agnostic EEG Representations with Geometry-Aware Modeling. (opens in new tab)<\/span><\/a><\/li>\n<\/ul>\n\n\n\n
- Paper on neonatal seizure detection: Protecting the Future: Neonatal Seizure Detection with Spatial-Temporal Modeling (opens in new tab)<\/span><\/a>. IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2023.<\/li>\n\n\n\n
- Unsupervised Video Anomaly Detection for Stereotypical Behaviors in Autism (opens in new tab)<\/span><\/a>. ICASSP 2023.<\/li>\n<\/ul>\n\n\n\n
- Paper on speech assessment:\u00a0End-to-End Word-Level Pronunciation Assessment with MASK Pre-training (opens in new tab)<\/span><\/a>. INTERSPEECH 2023.<\/li>\n<\/ul>\n\n\n\n