Kay-Yaw Tay
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Kay-Yaw Tay has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Shermyn Neo
Department of Neurology, National Neuroscience Institute, Singapore, Singapore
7 shared publications - Wing Lok Au
Department of Neurology, National Neuroscience Institute
7 shared publications - Louis Chew Seng Tan
Department of Neurology, National Neuroscience Institute, Singapore, Singapore
6 shared publications - Adeline Su Lyn Ng
Department of Neurology, National Neuroscience Institute, Singapore, Singapore
5 shared publications - Zheyu Xu
Department of Neurology, National Neuroscience Institute, Singapore, Singapore
4 shared publications - Eng-King Tan
National Neuroscience Institute, Singapore 308433, Singapore
3 shared publications
Latest Publications
- 28 March 2026
- Published by Elsevier in General Hospital Psychiatry
- Vol. 100, 156-162
Neuropsychiatric outcomes after deep brain stimulation for Parkinson's disease – Real-world data from a national movement disorders unit
- 28 March 2026
- Published by Elsevier in General Hospital Psychiatry
- Vol. 100, 156-162
Neuropsychiatric outcomes after deep brain stimulation for Parkinson's disease – Real-world data from a national movement disorders unit
- Open Access
- 3 December 2024
- Published by MDPI AG in Biomedicines
- Vol. 12 (12) , 2758
Machine Learning for Early Detection of Cognitive Decline in Parkinson’s Disease Using Multimodal Biomarker and Clinical Data
Background: Parkinson’s disease (PD) is the second most common neurodegenerative disease, primarily affecting the middle-aged to elderly population. Among its nonmotor symptoms, cognitive decline (CD) is a precursor to dementia and represents a critical target for early risk assessment and diagnosis. Accurate CD prediction is crucial for timely intervention and tailored management of at-risk patients. This study used machine learning (ML) techniques to predict the CD risk over five-year in early-stage PD. Methods: Data from the Early Parkinson’s Disease Longitudinal Singapore (2014 to 2018) was used to predict CD defined as a one-unit annual decrease or a one-unit decline in Montreal Cognitive Assessment over two consecutive years. Four ML methods—AutoScore, Random Forest, K-Nearest Neighbors and Neural Network—were applied using baseline demographics, clinical assessments and blood biomarkers. Results: Variable selection identified key predictors of CD, including education year, diastolic lying blood pressure, diastolic standing blood pressure, systolic lying blood pressure, Hoehn and Yahr scale, body mass index, phosphorylated tau at threonine 181, total tau, Neurofilament light chain and suppression of tumorigenicity 2. Random Forest was the most effective, achieving an AUC of 0.93 (95% CI: 0.89, 0.97), using 10-fold cross-validation. Conclusions: Here, we demonstrate that ML-based models can identify early-stage PD patients at high risk for CD, supporting targeted interventions and improved PD management.
- 3 December 2024
- Published by MDPI AG in Biomedicines
- Vol. 12 (12) , 2758
Open AccessMachine Learning for Early Detection of Cognitive Decline in Parkinson’s Disease Using Multimodal Biomarker and Clinical Data
Background: Parkinson’s disease (PD) is the second most common neurodegenerative disease, primarily affecting the middle-aged to elderly population. Among its nonmotor symptoms, cognitive decline (CD) is a precursor to dementia and represents a critical target for early risk assessment and diagnosis. Accurate CD prediction is crucial for timely intervention and tailored management of at-risk patients. This study used machine learning (ML) techniques to predict the CD risk over five-year in early-stage PD. Methods: Data from the Early Parkinson’s Disease Longitudinal Singapore (2014 to 2018) was used to predict CD defined as a one-unit annual decrease or a one-unit decline in Montreal Cognitive Assessment over two consecutive years. Four ML methods—AutoScore, Random Forest, K-Nearest Neighbors and Neural Network—were applied using baseline demographics, clinical assessments and blood biomarkers. Results: Variable selection identified key predictors of CD, including education year, diastolic lying blood pressure, diastolic standing blood pressure, systolic lying blood pressure, Hoehn and Yahr scale, body mass index, phosphorylated tau at threonine 181, total tau, Neurofilament light chain and suppression of tumorigenicity 2. Random Forest was the most effective, achieving an AUC of 0.93 (95% CI: 0.89, 0.97), using 10-fold cross-validation. Conclusions: Here, we demonstrate that ML-based models can identify early-stage PD patients at high risk for CD, supporting targeted interventions and improved PD management.
- Open Access
- 5 November 2024
- Published by MDPI AG in Preprints.org
Developing a Machine Learning Model to Predict the Risk of Cognitive Decline in Early Parkinson's Disease
Background: Cognitive decline (CD) is a significant concern in Parkinson's disease (PD), highlighting the need for reliable risk prediction models for early intervention. This study used machine learning (ML) techniques to predict the CD risk over five-year in early-stage PD. Methods: Data from the Early Parkinson's Disease Longitudinal Singapore (2014 to 2018) was used to predict CD defined as a one-unit annual decrease or a one-unit decline in Montreal Cognitive Assessment (MoCA) over two consecutive years. Four ML methods—AutoScore, Random Forest, K-Nearest Neighbors and Neural Network—were applied using baseline demographics, clinical assessments and blood biomarkers. Model performance was evaluated using area under the curve (AUC), sensitivity and specificity. Results: Variable selection identified key predictors of CD, including education year, diastolic lying blood pressure, diastolic standing blood pressure, systolic lying blood pressure, Hoehn and Yahr scale, body mass index, phosphorylated tau at threonine 181, total tau, Neurofilament light chain and suppression of tumorigenicity 2. Random Forest was the most effective, achieving an AUC of 0.93 (95% CI: 0.89, 0.97), using 10-fold cross-validation. Conclusion: ML-based models offer potential for early identification of patients at high risk of CD, facilitating targeted interventions and improving patient outcomes in PD management.
- 5 November 2024
- Published by MDPI AG in Preprints.org
Open AccessDeveloping a Machine Learning Model to Predict the Risk of Cognitive Decline in Early Parkinson's Disease
Background: Cognitive decline (CD) is a significant concern in Parkinson's disease (PD), highlighting the need for reliable risk prediction models for early intervention. This study used machine learning (ML) techniques to predict the CD risk over five-year in early-stage PD. Methods: Data from the Early Parkinson's Disease Longitudinal Singapore (2014 to 2018) was used to predict CD defined as a one-unit annual decrease or a one-unit decline in Montreal Cognitive Assessment (MoCA) over two consecutive years. Four ML methods—AutoScore, Random Forest, K-Nearest Neighbors and Neural Network—were applied using baseline demographics, clinical assessments and blood biomarkers. Model performance was evaluated using area under the curve (AUC), sensitivity and specificity. Results: Variable selection identified key predictors of CD, including education year, diastolic lying blood pressure, diastolic standing blood pressure, systolic lying blood pressure, Hoehn and Yahr scale, body mass index, phosphorylated tau at threonine 181, total tau, Neurofilament light chain and suppression of tumorigenicity 2. Random Forest was the most effective, achieving an AUC of 0.93 (95% CI: 0.89, 0.97), using 10-fold cross-validation. Conclusion: ML-based models offer potential for early identification of patients at high risk of CD, facilitating targeted interventions and improving patient outcomes in PD management.
- Open Access
- 5 June 2020
- Published by Springer Nature in Molecular Neurodegeneration
- Vol. 15 (1) , 1-8
Utility of plasma Neurofilament light as a diagnostic and prognostic biomarker of the postural instability gait disorder motor subtype in early Parkinson’s disease
Background: The main motor subtypes of Parkinson’s disease (PD) include tremor-dominant (TD) and postural instability gait disorder (PIGD), with varying disease course that warrant the development of biomarkers capable of predicting progression according to motor subtype. The PIGD subtype is associated with a poorer prognosis, hence identification of a biomarker associated with PIGD is clinically relevant. Neurofilament light (NfL) chain is a potential biomarker of disease severity in neurological disorders including PD. However, no study has investigated NfL and PD motor subtypes. Here, we aimed to investigate the diagnostic and prognostic utility of plasma NfL for PD motor subtypes in early Parkinson’s disease. Given the higher risk for cognitive and motor decline in PIGD, we hypothesized that plasma NfL is a potential biomarker for PIGD. Methods: Plasma NfL was measured in 199 participants (149 PD and 50 healthy controls, HC) using an ultrasensitive single molecule array. Patients were classified into TD or PIGD based on MDS-UPDRS components. After 2 years, 115 patients were reassessed. Association between NfL and clinical measures in PIGD and TD at baseline and at 2-year follow-up were analysed. Results: At baseline, plasma NfL levels were higher in PD than HC (8.8 ± 3.4 vs 16.2 ± 7.6 pg/ml, p < 0.0001), and differentiated PD from HC with a good diagnostic accuracy (AUC = 0.833, p < 0.001). At 2 years, NfL was higher in PIGD than TD (18.4 ± 14.5 vs 12.6 ± 4.4 pg/ml, p = 0.039). Within the PIGD group, higher NfL associated significantly with worse global cognition and UPDRS motor scores at baseline, and was able to predict motor and cognitive decline at a mean follow-up duration of 1.9 years, controlled for age, sex and disease duration. Conclusions: In this longitudinal study, we demonstrated for the first time the potential utility of plasma NfL as a diagnostic and prognostic biomarker in PIGD even at early stages of PD. These important novel findings will require further confirmation in larger, longitudinal PD cohorts.
- 5 June 2020
- Published by Springer Nature in Molecular Neurodegeneration
- Vol. 15 (1) , 1-8
Open AccessUtility of plasma Neurofilament light as a diagnostic and prognostic biomarker of the postural instability gait disorder motor subtype in early Parkinson’s disease
Background: The main motor subtypes of Parkinson’s disease (PD) include tremor-dominant (TD) and postural instability gait disorder (PIGD), with varying disease course that warrant the development of biomarkers capable of predicting progression according to motor subtype. The PIGD subtype is associated with a poorer prognosis, hence identification of a biomarker associated with PIGD is clinically relevant. Neurofilament light (NfL) chain is a potential biomarker of disease severity in neurological disorders including PD. However, no study has investigated NfL and PD motor subtypes. Here, we aimed to investigate the diagnostic and prognostic utility of plasma NfL for PD motor subtypes in early Parkinson’s disease. Given the higher risk for cognitive and motor decline in PIGD, we hypothesized that plasma NfL is a potential biomarker for PIGD. Methods: Plasma NfL was measured in 199 participants (149 PD and 50 healthy controls, HC) using an ultrasensitive single molecule array. Patients were classified into TD or PIGD based on MDS-UPDRS components. After 2 years, 115 patients were reassessed. Association between NfL and clinical measures in PIGD and TD at baseline and at 2-year follow-up were analysed. Results: At baseline, plasma NfL levels were higher in PD than HC (8.8 ± 3.4 vs 16.2 ± 7.6 pg/ml, p < 0.0001), and differentiated PD from HC with a good diagnostic accuracy (AUC = 0.833, p < 0.001). At 2 years, NfL was higher in PIGD than TD (18.4 ± 14.5 vs 12.6 ± 4.4 pg/ml, p = 0.039). Within the PIGD group, higher NfL associated significantly with worse global cognition and UPDRS motor scores at baseline, and was able to predict motor and cognitive decline at a mean follow-up duration of 1.9 years, controlled for age, sex and disease duration. Conclusions: In this longitudinal study, we demonstrated for the first time the potential utility of plasma NfL as a diagnostic and prognostic biomarker in PIGD even at early stages of PD. These important novel findings will require further confirmation in larger, longitudinal PD cohorts.
- 30 March 2020
- Published by Hindawi Limited in Parkinson's Disease
Evolution of Initial Pharmacologic Treatment of Newly Diagnosed Parkinson's Disease Patients over a Decade in Singapore
Objective. The aim of this study is to compare Parkinson's disease (PD) treatment practices by movement disorder (MD) specialists across a decade, and to determine the factors that influence drug choice for the motor symptoms of PD in newly diagnosed drug-naive patients. Methods. This prospective temporal analysis included patients seen at the National Neuroscience Institute in Singapore and diagnosed with PD by MD specialists in the years 2007 and 2017. Primary outcomes were use of specific PD drugs and changes in drug-prescribing patterns. Descriptive analyses and multivariable logistic regression models determined the extent to which patient characteristics were associated with type of PD treatment. Results. Of 230 patients with PD (mean (SD) age, 66.7 (10.3) years), 131 (57.0%) were male. From 2007 to 2017, the use of ergot dopamine agonists and anticholinergics decreased from 19.3% to 2.0% (P<0.001) and from 12.0% to 2.7% (P=0.004), respectively. The use of monoamine oxidase B inhibitors (MAOBI) increased from 13.3% to 25.2% (P=0.033). The use of levodopa (LD)-sparing strategies decreased nonsignificantly from 33.7% to 24.5% (P=0.133). Overall, 196 (85.2%) patients were initiated on symptomatic monotherapy, with LD being the most commonly prescribed. MAOBI was the most common drug used in combination therapy. Age <= 70 (adjusted OR, 11.9; 95% CI, 4.5-31.5) and Hoehn and Yahr (HY) stage <2 (adjusted OR, 3.4; 95% CI, 1.5-7.7) were independent factors for LD-sparing strategies. Non-LD prescriptions (13 of 92; 14.1%) were more likely to be discontinued compared to LD ones (6 of 149; 4.0%) (P=0.005). Conclusions. Drug-prescribing patterns in PD have changed significantly through the last decade, influenced by emerging evidence and reports of adverse drug effects. Choosing drugs based on the patient's age and disease severity remain sound guiding principles across the years. It is important that international and national guidelines for pharmacotherapy in PD be updated consistently throughout different socioeconomic settings to optimize care.
- 30 March 2020
- Published by Hindawi Limited in Parkinson's Disease
Evolution of Initial Pharmacologic Treatment of Newly Diagnosed Parkinson's Disease Patients over a Decade in Singapore
Objective. The aim of this study is to compare Parkinson's disease (PD) treatment practices by movement disorder (MD) specialists across a decade, and to determine the factors that influence drug choice for the motor symptoms of PD in newly diagnosed drug-naive patients. Methods. This prospective temporal analysis included patients seen at the National Neuroscience Institute in Singapore and diagnosed with PD by MD specialists in the years 2007 and 2017. Primary outcomes were use of specific PD drugs and changes in drug-prescribing patterns. Descriptive analyses and multivariable logistic regression models determined the extent to which patient characteristics were associated with type of PD treatment. Results. Of 230 patients with PD (mean (SD) age, 66.7 (10.3) years), 131 (57.0%) were male. From 2007 to 2017, the use of ergot dopamine agonists and anticholinergics decreased from 19.3% to 2.0% (P<0.001) and from 12.0% to 2.7% (P=0.004), respectively. The use of monoamine oxidase B inhibitors (MAOBI) increased from 13.3% to 25.2% (P=0.033). The use of levodopa (LD)-sparing strategies decreased nonsignificantly from 33.7% to 24.5% (P=0.133). Overall, 196 (85.2%) patients were initiated on symptomatic monotherapy, with LD being the most commonly prescribed. MAOBI was the most common drug used in combination therapy. Age <= 70 (adjusted OR, 11.9; 95% CI, 4.5-31.5) and Hoehn and Yahr (HY) stage <2 (adjusted OR, 3.4; 95% CI, 1.5-7.7) were independent factors for LD-sparing strategies. Non-LD prescriptions (13 of 92; 14.1%) were more likely to be discontinued compared to LD ones (6 of 149; 4.0%) (P=0.005). Conclusions. Drug-prescribing patterns in PD have changed significantly through the last decade, influenced by emerging evidence and reports of adverse drug effects. Choosing drugs based on the patient's age and disease severity remain sound guiding principles across the years. It is important that international and national guidelines for pharmacotherapy in PD be updated consistently throughout different socioeconomic settings to optimize care.
- Open Access
- 19 February 2020
- Published by Hindawi Limited in Parkinson's Disease
- Vol. 2020, 1-6
Attitude of Asian Parkinson Patients towards Clinical Research and Tissue Donation
Objective. The success of clinical research and tissue donation programs are highly dependent on recruitment of willing volunteers. A comprehensive survey of patient preferences and attitudes can help identify and address barriers hindering the recruitment for research. Method. This is a cross-sectional study on 105 Parkinson’s disease patients who completed an interviewer-administered questionnaire. Results. Out of 105 respondents, 48% of patients had either already participated in clinical research or were keen to participate. About 80% believed clinical research to be safe for their health and privacy. More than 70% of participants were willing to donate blood, urine, or stool, while 16% were agreeable for cerebrospinal fluid sample donation. Motivating factors for clinical research included altruism (64%) and contribution to advance medical knowledge (64%). Common reasons for unwillingness towards clinical research included the risks involved (43%), time constraints (33%), and mobility challenges (24%). Conclusion. The attitude of Singaporean Parkinson patients toward clinical research and tissue donation is encouraging with about half of the participants willing to support clinical research. Three-quarters of patients would support tissue donations. Participation in research may be further increased with greater patient and public education to overcome misconceptions and also by limiting the demands of studies.
- 19 February 2020
- Published by Hindawi Limited in Parkinson's Disease
- Vol. 2020, 1-6
Open AccessAttitude of Asian Parkinson Patients towards Clinical Research and Tissue Donation
Objective. The success of clinical research and tissue donation programs are highly dependent on recruitment of willing volunteers. A comprehensive survey of patient preferences and attitudes can help identify and address barriers hindering the recruitment for research. Method. This is a cross-sectional study on 105 Parkinson’s disease patients who completed an interviewer-administered questionnaire. Results. Out of 105 respondents, 48% of patients had either already participated in clinical research or were keen to participate. About 80% believed clinical research to be safe for their health and privacy. More than 70% of participants were willing to donate blood, urine, or stool, while 16% were agreeable for cerebrospinal fluid sample donation. Motivating factors for clinical research included altruism (64%) and contribution to advance medical knowledge (64%). Common reasons for unwillingness towards clinical research included the risks involved (43%), time constraints (33%), and mobility challenges (24%). Conclusion. The attitude of Singaporean Parkinson patients toward clinical research and tissue donation is encouraging with about half of the participants willing to support clinical research. Three-quarters of patients would support tissue donations. Participation in research may be further increased with greater patient and public education to overcome misconceptions and also by limiting the demands of studies.
- 3 October 2019
- Published by Springer Nature in Cell and Tissue Banking
- Vol. 20 (4) , 585-590
Attitudes of Asian Parkinson patients towards brain donation
Histopathological examination of brain tissue is required for better understanding of neurodegenerative conditions such as Parkinson’s disease and related disorders. However, patient willingness remains the greatest hurdle hampering participation in brain donation for research. While there is extensive research being conducted on the subject in West, to the best of our knowledge, there are no studies done in this regard in Asia. This cross-sectional survey was conducted on 105 Parkinson’s disease patients to assess their knowledge, beliefs and attitude towards brain donation in an Asian population. The majority of the participants (78%) acknowledged the importance of donation of brain for research, and 70% believed that their donated brain samples would be handled professionally. Fifty percent participants were willing to consider donating their brain for research. Motivating factors for brain donation included altruism (87%) and contribution to advance medical knowledge (80%). Common reasons for unwillingness towards brain donation were stress for family (30%), disfigurement of body (25%), and having a conservative mindset (23%). About one-third of the participants preferred to be approached for brain donation after their first clinic visit. Most patients preferred either their treating neurologists (66%) or research staff (18%) to discuss brain donation with. Participation for brain donation may be increased further with greater patient and public education to overcome misconceptions and change mindsets.
- 3 October 2019
- Published by Springer Nature in Cell and Tissue Banking
- Vol. 20 (4) , 585-590
Attitudes of Asian Parkinson patients towards brain donation
Histopathological examination of brain tissue is required for better understanding of neurodegenerative conditions such as Parkinson’s disease and related disorders. However, patient willingness remains the greatest hurdle hampering participation in brain donation for research. While there is extensive research being conducted on the subject in West, to the best of our knowledge, there are no studies done in this regard in Asia. This cross-sectional survey was conducted on 105 Parkinson’s disease patients to assess their knowledge, beliefs and attitude towards brain donation in an Asian population. The majority of the participants (78%) acknowledged the importance of donation of brain for research, and 70% believed that their donated brain samples would be handled professionally. Fifty percent participants were willing to consider donating their brain for research. Motivating factors for brain donation included altruism (87%) and contribution to advance medical knowledge (80%). Common reasons for unwillingness towards brain donation were stress for family (30%), disfigurement of body (25%), and having a conservative mindset (23%). About one-third of the participants preferred to be approached for brain donation after their first clinic visit. Most patients preferred either their treating neurologists (66%) or research staff (18%) to discuss brain donation with. Participation for brain donation may be increased further with greater patient and public education to overcome misconceptions and change mindsets.