Haiteng Jiang

Background: Psychiatric clinical notes in electronic health records (EHRs) provide rich longitudinal information that can support clinical decision-making. Using historical medical data can enable earlier identification of mental illness, better characterization of disease trajectories, and more personalized treatment planning. Natural language processing (NLP) transforms these unstructured notes into analyzable representations for research and care. Objective: This study aims to systematically summarize NLP methodologies for psychiatric clinical notes, compare major modeling paradigms and application areas, and highlight emerging large language model (LLM) trends, key challenges, and future research directions. Methods: Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines, a literature search was conducted for articles on NLP methods based on psychiatric clinical notes published from January 2021 to December 2025 in Ovid MEDLINE, Ovid EMBASE, PubMed, Scopus, Web of Science, the ACM Digital Library, and ScienceDirect. This scoping review analyzed NLP methods applied to psychiatric clinical notes, focusing on major trends, identifying suitable features for traditional machine learning (ML)–based models, applications of pretrained language models (PLMs), and key challenges. Approaches were categorized as rule-based, traditional ML, hybrid, deep learning (DL), and LLM-based methods across information extraction and text classification tasks. Results: In total, 101 studies were eligible for inclusion. Rule-based methods (n=36) and hybrid approaches (n=34) remained the most widely used techniques, largely favored for their interpretability in handling nuanced, subjective clinical notes. These were followed by DL (n=15), traditional ML (n=10), and LLM-based approaches (n=6). Traditional ML studies relied heavily on engineered features, which could be grouped into 5 broad categories: domain knowledge features, lexical and statistical features, vector-based semantic features, emotion-related features, and temporal features. PLMs improved performance mainly through domain adaptation and task-specific fine-tuning, enhancing the handling of psychiatric language, medical terminology, and clinical note structure. LLM-based studies, although still limited in number, indicated a growing shift toward generative and reasoning-based applications. Conclusions: Hybrid NLP approaches remain dominant, combining domain rules with ML for extraction and classification. DL approaches continue to advance, with domain adaptation supporting medical terminology and clinical semantics. LLMs may further automate complex workflows via zero-shot capabilities and reasoning, alongside growing interest in temporal modeling and multimodal integration. Key future needs include improved generalizability across institutions, privacy protection, and careful attention to ethical implications in clinical deployment.

Psychological support hotlines provide immediate help to individuals in crisis, with operators assessing emotional states and suicide risk. However, increasing demand has led to a shortage of trained professionals, emphasizing the need for AI-driven crisis detection models. This study included 1,057 calls from the Hangzhou Hotline (2022–2023) to evaluate the effectiveness of deep learning and pre-trained models in detecting psychological crises using audio (Wave2Vec, Whisper) and transcribed text (RoBERTa, GPT). We adopted two strategies: deep learning classification with pre-trained models and Large Language Models (LLMs)-based prediction via prompt engineering (GPT-4 and DeepSeek series). The deep learning framework, employing GPT embeddings excelled in multi-label predictions compared to auditory model, achieving 80.48% [80.18%, 80.78%] F1 scores for identifying high-risk calls in prospective tests. Fusion experiments revealed that acoustic features offered negligible predictive value compared with text semantics. Notably, GPT-4o and DeepSeek-R1, utilizing few-shot learning, demonstrated performance comparable to the GPT-embedding deep learning model across multiple tasks. This suggests that their advanced Chain-of-Thought reasoning effectively mitigates data dependency gap, enabling LLMs to align with clinical domains using a few examples. Expert evaluation confirmed the clinical applicability of GPT-generated explanations. Taken together, these findings highlight the potential of LLMs in mental health crisis detection and lay the foundation for future research.

Transcranial direct current stimulation (tDCS) and virtual reality (VR) have emerged as promising non-invasive interventions in treating psychiatric disord

Cognitive deficits are a hallmark of Alzheimer’s disease (AD), and effective treatments remain elusive. Transcranial alternating current stimulation (tACS), a non-invasive technique, has shown potential in improving cognitive function across various populations, but further research is needed to investigate its efficacy in AD. In a randomized, double-blind, sham-controlled pilot trial, 36 mild AD patients received active or sham theta-tACS (8 Hz, 1.6 mA, 20-min daily) during n-back task for two weeks, followed by a 10-week follow-up. Cognitive assessments and resting-state EEG were analyzed at baseline, after-treatment, and follow-up. The results showed that the active group demonstrated significant cognitive improvements after treatment (MMSE: t (15) =-3.273, p = 0.005, Cohen’s d = 0.82), particularly in short-term memory (MMSE-recall: Z = -2.11, p = 0.035, r = 0.53), with maintained benefits after 10 weeks. In contrast, the sham group exhibited long-term cognitive decline (MMSE: t (4) = 3.586, p = 0.023, Cohen’s d = -1.60). EEG analysis revealed reduced gamma power (t (23) = 2.689, p = 0.013, Cohen’s d = 1.077) and theta connectivity in active group, particularly in the frontotemporal regions (F4/F7: t (23) = 2.467, p = 0.021, Cohen’s d = 0.988; F4/T3: t (23) = 2.465, p = 0.022, Cohen’s d = 0.987), which was correlated with cognitive improvements (R = –0.57, p = 0.043). In conclusion, tACS combining cognitive training may offer cognitive benefits in mild AD by modulating neural activity, though further studies are needed to clarify its mechanisms.

Human brain achieves dynamic stability-plasticity balance through synaptic homeostasis. Inspired by this biological principle, we propose SPICED: a neuromorphic framework that integrates the synaptic homeostasis mechanism for unsupervised continual EEG decoding, particularly addressing practical scenarios where new individuals with inter-individual variability emerge continually. SPICED comprises a novel synaptic network that enables dynamic expansion during continual adaptation through three bio-inspired neural mechanisms: (1) critical memory reactivation; (2) synaptic consolidation and (3) synaptic renormalization. The interplay within synaptic homeostasis dynamically strengthens task-discriminative memory traces and weakens detrimental memories. By integrating these mechanisms with continual learning system, SPICED preferentially replays task-discriminative memory traces that exhibit strong associations with newly emerging individuals, thereby achieving robust adaptations. Meanwhile, SPICED effectively mitigates catastrophic forgetting by suppressing the replay prioritization of detrimental memories during long-term continual learning. Validated on three EEG datasets, SPICED show its effectiveness.

Objective: Alzheimer's disease, a progressive neurodegenerative disorder, involves neural, genetic, and proteomic factors and impacts multiple cognitive and behavioral domains. Traditional AD prediction largely focuses on univariate disease outcomes, such as disease stages and severity. Multimodal data provide richer disease information than a single modality and may enhance prediction, but are often incomplete due to missing measurements. Recent machine learning approaches show promise in improving prediction accuracy, but their biological relevance remains insufficiently understood. Methods: Integrating missing data analysis, predictive modeling, multimodal data analysis, and explainable AI (XAI), we propose OPTIMUS, a predictive and explainable machine-learning framework, to unveil the many-to-many predictive pathways between multimodal input data and multivariate disease outcomes amidst missing values. Results: OPTIMUS first applies imputation to handle missing data within each modality type while optimizing overall prediction accuracy. It then maps multimodal biomarkers to multivariate outcomes using machine learning and extracts biomarkers that are respectively predictive of each outcome. Finally, outcome prediction targetedly integrating multimodality with unavailable signals (OPTIMUS) incorporates XAI to explain the identified multimodal biomarkers. Conclusion: Using data from 348 cognitively normal subjects, 601 persons with mild cognitive impairment, and 256 AD patients, OPTIMUS identifies neural and transcriptomic signatures that jointly but differentially predict outcomes related to memory, executive function, visuospatial function, and language. Significance: Our work demonstrates the potential of building a predictive and biologically explainable machine-learning framework to uncover multimodal biomarkers that capture disease profiles across varying cognitive landscapes. The results improve our understanding of the complex many-to-many pathways in AD.