Yunshu Zhu

Yunshu Zhu

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 February 2025 in Journal of Healthcare Informatics Research
Verified
Journal Article
Open access
PDF file
DVDinithi Vithanage
Dinithi Vithanage
CDChao Deng
Chao Deng
LWLei Wang
Lei Wang
MYMengyang Yin
Mengyang Yin
MAMohammad Alkhalaf
Mohammad Alkhalaf
Ping YuPing Yu
Ping Yu
Information extraction (IE) of unstructured electronic health records is challenging due to the semantic complexity of textual data. Generative large language models (LLMs) offer promising solutions to address this challenge. However, identifying the best training methods to adapt LLMs for IE in residential aged care settings remains underexplored. This research addresses this challenge by evaluating the effects of zero-shot and few-shot learning, both with and without parameter-efficient fine-tuning (PEFT) and retrieval-augmented generation (RAG) using Llama 3.1-8B. The study performed named entity recognition (NER) to nursing notes from Australian aged care facilities (RACFs), focusing on agitation in dementia and malnutrition risk factors. Performance evaluation includes accuracy, macro-averaged precision, recall, and F1 score. We used non-parametric statistical methods to compare if the differences were statistically significant. Results show that zero-shot and few-shot learning, whether combined with PEFT or RAG, achieve comparable performance across the clinical domains when the same prompting template is used. Few-shot learning significantly outperforms zero-shot learning when neither PEFT nor RAG is applied. Notably, PEFT significantly improves model performance in both zero-shot and few-shot learning; however, RAG significantly improves performance only in few-shot learning. After PEFT, the performance of zero-shot learning reaches a comparable level with few-shot learning. However, few-shot learning with RAG significantly outperforms zero-shot learning with RAG. We also found a similar level of performance between few-shot learning with RAG and zero-shot learning with PEFT. These findings provide valuable insights for researchers, practitioners, and stakeholders to optimize the use of generative LLMs in clinical IE.

1 April 2022 in Journal of Gerontological Nursing
Verified
Journal Article
Closed access
YZYunshu Zhu
Yunshu Zhu
TSTing Song
Ting Song
ZZZhenyu Zhang
Zhenyu Zhang
CDChao Deng
Chao Deng
MAMohammad Alkhalaf
Mohammad Alkhalaf
PYPing Yu
Ping Yu
Using a suite of artificial intelligence technologies, the current study sought to determine the prevalence of agitated behaviors in people with dementia in residential aged care facilities (RACFs) in Australia. Computerized natural language processing allowed extraction of agitation instances from the free-text nursing progress notes, a component of electronic health records in RACFs. In total, 59 observable agitated behaviors were found. No difference was found in dementia prevalence between female and male clients (44.1%), across metropolitan and regional facilities (42.1% [ SD = 17.9%]), or for agitation prevalence in dementia (76.5% [ SD = 18.4%]). The top 10 behaviors were resisting, wandering, speaking in excessively loud voice, pacing, restlessness, pushing, shouting, complaining, frustration, and using profane language. Four to 17 agitated behaviors coexisted in 53% of people with dementia agitation, indicating high caregiver burden in these RACFs. Improving workforce training and redesigning care models are urgent for sustainability of dementia care in RACFs. [ Journal of Gerontological Nursing, 48 (4), 57–64.]

25 January 2024
Verified
Book Chapter
Open access
PDF file
YZYunshu Zhu
Yunshu Zhu
TSTing Song
Ting Song
ZZZhenyu Zhang
Zhenyu Zhang
MYMengyang Yin
Mengyang Yin
Ping YuPing Yu
Ping Yu
Natural Language Processing (NLP) is a powerful technique for extracting valuable information from unstructured electronic health records (EHRs). However, a prerequisite for NLP is the availability of high-quality annotated datasets. To date, there is a lack of effective methods to guide the research effort of manually annotating unstructured datasets, which can hinder NLP performance. Therefore, this study develops a five-step workflow for manually annotating unstructured datasets, including (1) annotator training and familiarising with the text corpus, (2) vocabulary identification, (3) annotation schema development, (4) annotation execution, and (5) result validation. This framework was then applied to annotate agitation symptoms from the unstructured EHRs of 40 Australian residential aged care facilities. The annotated corpus achieved an accuracy rate of 96%. This suggests that our proposed annotation workflow can be used in manual data processing to develop annotated training corpus for developing NLP algorithms.

25 January 2024
Verified
Book Chapter
Open access
PDF file
DVDinithi Vithanage
Dinithi Vithanage
YZYunshu Zhu
Yunshu Zhu
ZZZhenyu Zhang
Zhenyu Zhang
CDChao Deng
Chao Deng
MYMengyang Yin
Mengyang Yin
Ping YuPing Yu
Ping Yu
Nursing staff record observations about older people under their care in free-text nursing notes. These notes contain older people’s care needs, disease symptoms, frequency of symptom occurrence, nursing actions, etc. Therefore, it is vital to develop a technique to uncover important data from these notes. This study developed and evaluated a deep learning and transfer learning-based named entity recognition (NER) model for extracting symptoms of agitation in dementia from the nursing notes. We employed a Clinical BioBERT model for word embedding. Then we applied bidirectional long-short-term memory (BiLSTM) and conditional random field (CRF) models for NER on nursing notes from Australian residential aged care facilities. The proposed NER model achieves satisfactory performance in extracting symptoms of agitation in dementia with a 75% F1 score and 78% accuracy. We will further develop machine learning models to recommend the optimal nursing actions to manage agitation.

30 September 2024 in Psychiatry International
Verified
Journal Article
Open access
PDF file
YZYunshu Zhu
Yunshu Zhu
Ping YuPing Yu
Ping Yu
WLWanqing Li
Wanqing Li
TSTing Song
Ting Song
ZZZhenyu Zhang
Zhenyu Zhang
CDChao Deng
Chao Deng
Agitation is one of the most common and persistent behavioral and psychological symptoms among persons with dementia (PWD) in residential aged care facilities (RACFs). While most studies have explored the general impact of COVID-19 on the mental health and well-being of aged care residents, there has been limited research on the pandemic’s impact on agitation in dementia within RACFs. This knowledge is crucial to ensuring that RACFs are better equipped to handle future public health emergencies. Therefore, this retrospective cohort study investigates the impact of the pandemic on agitation incidents within 40 Australian RACFs. Using Python, we extracted agitation symptoms from nursing notes and compared the frequency and percentage of symptom occurrence pre-pandemic versus during-pandemic. Chi-square tests examined any significant change in the prevalence of dementia and agitation in dementia between the comparative groups and periods. Dementia prevalence dropped significantly during the second year of the pandemic, with a concurrent increase in early-onset dementia cases. Overall, agitation symptoms decreased by 20.1%, but six symptoms significantly increased: resisting (28.98%), wandering (11.78%), restlessness (3.19%), complaining (10.1%), arguing (2.36%), and outbursts (1.74%). Conversely, pacing decreased by 15.88% and speaking loudly decreased by 10.9%. Over half of the care recipients with dementia experienced agitation symptoms 2–5 times each year, increasing from 50.56% in 2019 to 58.28% in 2021. Despite the co-occurrence of symptoms such as falls, confusion, and weakness, there was no evidence to suggest that these comorbidities were specific to COVID-19; rather, they appeared to be prevalent even before the pandemic. Persons with agitation in dementia had a significantly higher death rate during the COVID-19 pandemic than otherwise.

27 June 2024 in medRxiv
Verified
Preprint
Open access
PDF file
DVDinithi Vithanage
Dinithi Vithanage
CDChao Deng
Chao Deng
LWLei Wang
Lei Wang
MYMengyang Yin
Mengyang Yin
MAMohammad Alkhalaf
Mohammad Alkhalaf
Ping YuPing Yu
Ping Yu
Multi-label classification of unstructured electronic health records (EHR) is challenging due to the semantic complexity of textual data. Identifying the most effective machine learning method for EHR classification is useful in real-world clinical settings. Advances in natural language processing (NLP) using large language models (LLMs) offer promising solutions. Therefore, this experimental research aims to test the effects of zero-shot and few-shot learning prompting, with and without parameter-efficient fine-tuning (PEFT) and retrieval-augmented generation (RAG) of LLMs, on the multi-label classification of unstructured EHR data from residential aged care facilities (RACFs) in Australia. The four clinical tasks examined are agitation in dementia, depression in dementia, frailty index, and malnutrition risk factors, using the Llama 3.1-8B. Performance evaluation includes accuracy, macro-averaged precision, recall, and F1 score, supported by non-parametric statistical analyses. Results indicate that both zero-shot and few-shot learning, regardless of the use of PEFT and RAG, demonstrate equivalent performance across the clinical tasks when using the same prompting template. Few-shot learning consistently outperforms zero-shot learning when neither PEFT nor RAG is applied. Notably, PEFT significantly enhances model performance in both zero-shot and few-shot learning; however, RAG improves performance only in few-shot learning. After PEFT, the performance of zero-shot learning is equal to that of few-shot learning across clinical tasks. Additionally, few-shot learning with RAG surpasses zero-shot learning with RAG, while no significant difference exists between few-shot learning with RAG and zero-shot learning with PEFT. These findings offer crucial insights into LLMs for researchers, practitioners, and stakeholders utilizing LLMs in clinical document analysis.