Jiali Yao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jiali Yao has a total of 73 co-authors. Here are the top co-authors by number of shared publications:
- Falk Müller-Riemenschneider
Saw Swee Hock School of Public Health, National University of Singapore, Singapore, Singapore
5 shared publications - Andre Matthias Müller
Saw Swee Hock School of Public Health, National University of Singapore, Singapore, Singapore.
3 shared publications - Chuen Seng Tan
Saw Swee Hock School of Public Health, National University of Singapore and National University Health System, Singapore, Singapore
3 shared publications - Jeremy Tan
Policy, Research and Surveillance Division Health Promotion Board Singapore
3 shared publications - Nicole Lim
Policy, Research and Surveillance Division Health Promotion Board Singapore
3 shared publications - Agnes Tan
Health Promotion Board Singapore, Singapore, Singapore
2 shared publications
Latest Publications
- 7 October 2026
- Published by Elsevier in The American Journal of Clinical Nutrition
- Vol. 124 (6) , 101574
Lifestyle determinants of glycemic responses to snacking in free-living adults without diabetes: an intensive longitudinal observational study
- 7 October 2026
- Published by Elsevier in The American Journal of Clinical Nutrition
- Vol. 124 (6) , 101574
Lifestyle determinants of glycemic responses to snacking in free-living adults without diabetes: an intensive longitudinal observational study
- Open Access
- 1 November 2025
- Published by American Society of Clinical Oncology (ASCO) in JCO Clinical Cancer Informatics
- Vol. 9 (9) , e2400243
Augmenting Large Language Models With National Comprehensive Cancer Network Guidelines for Improved and Standardized Adjuvant Therapy Recommendations in Postoperative Breast Cancer Cases
PURPOSE: Multidisciplinary breast tumor boards (MTBs) are essential for optimizing breast cancer treatment but face challenges related to logistics, variability in expertise, and lack of standardization. Large language models may support clinical decision making. This study evaluates the accuracy of adjuvant therapy recommendations generated by an artificial intelligence (AI)–driven tool, TheSerenityBot (TSB), in comparison with Claude-2 and GPT-4, using expert MTB consensus as the reference. METHODS: Postoperative breast cancer cases reviewed at the National University Hospital, Singapore, between June and November 2023 were retrospectively analyzed. Eligible patients were women with invasive or preinvasive breast cancer who underwent surgery. Metastatic cases were excluded. TSB, a Claude-2–based model augmented with 2023 National Comprehensive Cancer Network guidelines, generated adjuvant therapy recommendations across seven treatment modalities. Outputs from TSB, Claude-2, and GPT-4 were evaluated for concordance with MTB recommendations. Model performance was assessed using generalized estimating equations. RESULTS: Fifty patients were included (mean age, 59.8 years); 75.5% had hormone receptor–positive tumors, and 60.0% underwent breast-conserving surgery. TSB demonstrated the highest overall accuracy (0.89), followed by Claude-2 (0.86) and GPT-4 (0.78). GPT-4 showed significantly lower accuracy in genetic testing recommendations (odds ratio [OR], 0.05 [95% CI, 0.015 to 0.149]; P < .001), whereas Claude-2 was less accurate in radiotherapy recommendations (OR, 0.41 [95% CI, 0.17 to 0.98]; P = .040). CONCLUSION: A guideline-augmented AI tool such as TSB shows promise in supporting adjuvant therapy decisions in breast cancer. To improve clinical relevance, future iterations will incorporate individualized patient factors, broader guideline frameworks, and electronic health record integration. Prospective trials are ongoing to assess the real-world impact.
- 1 November 2025
- Published by American Society of Clinical Oncology (ASCO) in JCO Clinical Cancer Informatics
- Vol. 9 (9) , e2400243
Open AccessAugmenting Large Language Models With National Comprehensive Cancer Network Guidelines for Improved and Standardized Adjuvant Therapy Recommendations in Postoperative Breast Cancer Cases
PURPOSE: Multidisciplinary breast tumor boards (MTBs) are essential for optimizing breast cancer treatment but face challenges related to logistics, variability in expertise, and lack of standardization. Large language models may support clinical decision making. This study evaluates the accuracy of adjuvant therapy recommendations generated by an artificial intelligence (AI)–driven tool, TheSerenityBot (TSB), in comparison with Claude-2 and GPT-4, using expert MTB consensus as the reference. METHODS: Postoperative breast cancer cases reviewed at the National University Hospital, Singapore, between June and November 2023 were retrospectively analyzed. Eligible patients were women with invasive or preinvasive breast cancer who underwent surgery. Metastatic cases were excluded. TSB, a Claude-2–based model augmented with 2023 National Comprehensive Cancer Network guidelines, generated adjuvant therapy recommendations across seven treatment modalities. Outputs from TSB, Claude-2, and GPT-4 were evaluated for concordance with MTB recommendations. Model performance was assessed using generalized estimating equations. RESULTS: Fifty patients were included (mean age, 59.8 years); 75.5% had hormone receptor–positive tumors, and 60.0% underwent breast-conserving surgery. TSB demonstrated the highest overall accuracy (0.89), followed by Claude-2 (0.86) and GPT-4 (0.78). GPT-4 showed significantly lower accuracy in genetic testing recommendations (odds ratio [OR], 0.05 [95% CI, 0.015 to 0.149]; P < .001), whereas Claude-2 was less accurate in radiotherapy recommendations (OR, 0.41 [95% CI, 0.17 to 0.98]; P = .040). CONCLUSION: A guideline-augmented AI tool such as TSB shows promise in supporting adjuvant therapy decisions in breast cancer. To improve clinical relevance, future iterations will incorporate individualized patient factors, broader guideline frameworks, and electronic health record integration. Prospective trials are ongoing to assess the real-world impact.
- Open Access
Scaling Precision Medicine: Singapore’s PRECISE-SG100K Genomic Engine
The whole-of-nation National Precision Medicine (NPM) programme in Singapore aims to integrate a genomics backbone into standard healthcare. Phase II recently completed the sequencing of 102,202 genomes from the diverse Asian population of Singapore, generating the PRECISE-SG100K dataset. The scaling from the pilot SG10K_Health phase to national production scale demanded a central shift: from earlier generation workflows to an integrated, industrial grade “ Genomic Engine ”. The initiative was co-developed under a public-private partnership model. Our Genomic Engine was designed with an automation-first approach, as a modular platform linking a purpose-built, centralized biospecimen processing and DNA extraction facility (EXTRACT Lab), a high-throughput sequencing operation (Illumina Data Generation Infrastructure, IDGI), with the cloud-native analytics (GRIDS) running on a fully integrated PRECISE Sample Tracking System (STS). Here we describe the key design features of our Genomic Engine. Highlighting the Stress Test phase where the automated DNA extraction yields failed to meet expectations. We share the operational adjustments which were implemented to stabilize production and resulted in a three-fold reduction in sample triage due to improved DNA recovery. Without establishing the operational guardrails during the Stress Test, the Engine would not have reached a steady-state production evidenced by >3,000 genomes per month, median turnaround time 11 days, 99.6% operational success, within in six months. This framework offers a practical blueprint for other population-scale initiatives seeking to bridge the gap between scientific piloting and industrial delivery.
Open AccessScaling Precision Medicine: Singapore’s PRECISE-SG100K Genomic Engine
The whole-of-nation National Precision Medicine (NPM) programme in Singapore aims to integrate a genomics backbone into standard healthcare. Phase II recently completed the sequencing of 102,202 genomes from the diverse Asian population of Singapore, generating the PRECISE-SG100K dataset. The scaling from the pilot SG10K_Health phase to national production scale demanded a central shift: from earlier generation workflows to an integrated, industrial grade “ Genomic Engine ”. The initiative was co-developed under a public-private partnership model. Our Genomic Engine was designed with an automation-first approach, as a modular platform linking a purpose-built, centralized biospecimen processing and DNA extraction facility (EXTRACT Lab), a high-throughput sequencing operation (Illumina Data Generation Infrastructure, IDGI), with the cloud-native analytics (GRIDS) running on a fully integrated PRECISE Sample Tracking System (STS). Here we describe the key design features of our Genomic Engine. Highlighting the Stress Test phase where the automated DNA extraction yields failed to meet expectations. We share the operational adjustments which were implemented to stabilize production and resulted in a three-fold reduction in sample triage due to improved DNA recovery. Without establishing the operational guardrails during the Stress Test, the Engine would not have reached a steady-state production evidenced by >3,000 genomes per month, median turnaround time 11 days, 99.6% operational success, within in six months. This framework offers a practical blueprint for other population-scale initiatives seeking to bridge the gap between scientific piloting and industrial delivery.
- Open Access
- 6 February 2023
- Published by Royal Society of Chemistry (RSC) in Food & Function
- Vol. 14 (4) , 2260-2269
Associations between oral processing, saliva, and bolus properties on daily glucose excursions amongst people at risk of type-2 diabetes
Greater bolus saliva uptake was associated with a higher time spent above the normal range for blood glucose. Salivary amylase and saliva flow rate were not significantly associated with the time spent above the normal range.
- 6 February 2023
- Published by Royal Society of Chemistry (RSC) in Food & Function
- Vol. 14 (4) , 2260-2269
Open AccessAssociations between oral processing, saliva, and bolus properties on daily glucose excursions amongst people at risk of type-2 diabetes
Greater bolus saliva uptake was associated with a higher time spent above the normal range for blood glucose. Salivary amylase and saliva flow rate were not significantly associated with the time spent above the normal range.
- Open Access
- 21 June 2022
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 11 (12) , e022508
Evaluation of a Population‐Wide Mobile Health Physical Activity Program in 696 907 Adults in Singapore
Background: Evidence of scaled‐up physical activity interventions is scarce. This study evaluates the uptake, engagement, and effectiveness of one such intervention program. Methods and Results: The program was open to individuals aged ≥17 years in Singapore. The main intervention components comprised device‐based daily physical activity recording paired with step count goals and financial rewards. According to the different reward opportunities, we divided the evaluation period (August 2017 to June 2018) into the baseline monitoring phase, the main challenge phase, and the maintenance phase. Uptake was assessed by the number of individuals registered, and engagement by the step recording duration after registration. The effectiveness was defined as changes in mean daily step count from baseline to the main challenge phase and the maintenance phase. A total of 696 907 participants registered, including more Singapore citizens (versus noncitizens), women, and younger (aged 17–39 years) individuals. The evaluation of engagement and effectiveness included 421 388 (60.5%) participants who provided plausible characteristic information and step count data. The median duration of engagement was 74 (IQR, 14–149) days. Compared with the baseline of 7509 (SD, 3467) steps, mean daily step count increased by 1579 (95% CI, 1564–1594) steps during the main challenge phase and 934 (95% CI, 916–952) steps during the maintenance phase. Greater engagement and activity increase were found in participants who are citizens, women, aged ≥40 years, non‐obese, and using separate wearables (versus smartphones). Conclusions: Mobile health physical activity interventions can successfully reach a large population and be effective in increasing physical activity, despite declining program engagement over time.
- 21 June 2022
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 11 (12) , e022508
Open AccessEvaluation of a Population‐Wide Mobile Health Physical Activity Program in 696 907 Adults in Singapore
Background: Evidence of scaled‐up physical activity interventions is scarce. This study evaluates the uptake, engagement, and effectiveness of one such intervention program. Methods and Results: The program was open to individuals aged ≥17 years in Singapore. The main intervention components comprised device‐based daily physical activity recording paired with step count goals and financial rewards. According to the different reward opportunities, we divided the evaluation period (August 2017 to June 2018) into the baseline monitoring phase, the main challenge phase, and the maintenance phase. Uptake was assessed by the number of individuals registered, and engagement by the step recording duration after registration. The effectiveness was defined as changes in mean daily step count from baseline to the main challenge phase and the maintenance phase. A total of 696 907 participants registered, including more Singapore citizens (versus noncitizens), women, and younger (aged 17–39 years) individuals. The evaluation of engagement and effectiveness included 421 388 (60.5%) participants who provided plausible characteristic information and step count data. The median duration of engagement was 74 (IQR, 14–149) days. Compared with the baseline of 7509 (SD, 3467) steps, mean daily step count increased by 1579 (95% CI, 1564–1594) steps during the main challenge phase and 934 (95% CI, 916–952) steps during the maintenance phase. Greater engagement and activity increase were found in participants who are citizens, women, aged ≥40 years, non‐obese, and using separate wearables (versus smartphones). Conclusions: Mobile health physical activity interventions can successfully reach a large population and be effective in increasing physical activity, despite declining program engagement over time.
- 19 December 2019
- Published by BMJ in British Journal of Sports Medicine
- Vol. 54 (17) , 1047-1048
Bright spots, physical activity investments that work: National Steps Challenge, Singapore: a nationwide mHealth physical activity programme
- 19 December 2019
- Published by BMJ in British Journal of Sports Medicine
- Vol. 54 (17) , 1047-1048
Bright spots, physical activity investments that work: National Steps Challenge, Singapore: a nationwide mHealth physical activity programme
- Open Access
- 2 October 2019
- Published by JMIR Publications Inc. in JMIR mHealth and uHealth
- Vol. 7 (10) , e14120
Heart Rate Measures From Wrist-Worn Activity Trackers in a Laboratory and Free-Living Setting: Validation Study
Background: Wrist-worn activity trackers are popular, and an increasing number of these devices are equipped with heart rate (HR) measurement capabilities. However, the validity of HR data obtained from such trackers has not been thoroughly assessed outside the laboratory setting. Objective: This study aimed to investigate the validity of HR measures of a high-cost consumer-based tracker (Polar A370) and a low-cost tracker (Tempo HR) in the laboratory and free-living settings. Methods: Participants underwent a laboratory-based cycling protocol while wearing the two trackers and the chest-strapped Polar H10, which acted as criterion. Participants also wore the devices throughout the waking hours of the following day during which they were required to conduct at least one 10-min bout of moderate-to-vigorous physical activity (MVPA) to ensure variability in the HR signal. We extracted 10-second values from all devices and time-matched HR data from the trackers with those from the Polar H10. We calculated intraclass correlation coefficients (ICCs), mean absolute errors, and mean absolute percentage errors (MAPEs) between the criterion and the trackers. We constructed decile plots that compared HR data from Tempo HR and Polar A370 with criterion measures across intensity deciles. We investigated how many HR data points within the MVPA zone (≥64% of maximum HR) were detected by the trackers. Results: Of the 57 people screened, 55 joined the study (mean age 30.5 [SD 9.8] years). Tempo HR showed moderate agreement and large errors (laboratory: ICC 0.51 and MAPE 13.00%; free-living: ICC 0.71 and MAPE 10.20%). Polar A370 showed moderate-to-strong agreement and small errors (laboratory: ICC 0.73 and MAPE 6.40%; free-living: ICC 0.83 and MAPE 7.10%). Decile plots indicated increasing differences between Tempo HR and the criterion as HRs increased. Such trend was less pronounced when considering the Polar A370 HR data. Tempo HR identified 62.13% (1872/3013) and 54.27% (5717/10,535) of all MVPA time points in the laboratory phase and free-living phase, respectively. Polar A370 detected 81.09% (2273/2803) and 83.55% (9323/11,158) of all MVPA time points in the laboratory phase and free-living phase, respectively. Conclusions: HR data from the examined wrist-worn trackers were reasonably accurate in both the settings, with the Polar A370 showing stronger agreement with the Polar H10 and smaller errors. Inaccuracies increased with increasing HRs; this was pronounced for Tempo HR.
- 2 October 2019
- Published by JMIR Publications Inc. in JMIR mHealth and uHealth
- Vol. 7 (10) , e14120
Open AccessHeart Rate Measures From Wrist-Worn Activity Trackers in a Laboratory and Free-Living Setting: Validation Study
Background: Wrist-worn activity trackers are popular, and an increasing number of these devices are equipped with heart rate (HR) measurement capabilities. However, the validity of HR data obtained from such trackers has not been thoroughly assessed outside the laboratory setting. Objective: This study aimed to investigate the validity of HR measures of a high-cost consumer-based tracker (Polar A370) and a low-cost tracker (Tempo HR) in the laboratory and free-living settings. Methods: Participants underwent a laboratory-based cycling protocol while wearing the two trackers and the chest-strapped Polar H10, which acted as criterion. Participants also wore the devices throughout the waking hours of the following day during which they were required to conduct at least one 10-min bout of moderate-to-vigorous physical activity (MVPA) to ensure variability in the HR signal. We extracted 10-second values from all devices and time-matched HR data from the trackers with those from the Polar H10. We calculated intraclass correlation coefficients (ICCs), mean absolute errors, and mean absolute percentage errors (MAPEs) between the criterion and the trackers. We constructed decile plots that compared HR data from Tempo HR and Polar A370 with criterion measures across intensity deciles. We investigated how many HR data points within the MVPA zone (≥64% of maximum HR) were detected by the trackers. Results: Of the 57 people screened, 55 joined the study (mean age 30.5 [SD 9.8] years). Tempo HR showed moderate agreement and large errors (laboratory: ICC 0.51 and MAPE 13.00%; free-living: ICC 0.71 and MAPE 10.20%). Polar A370 showed moderate-to-strong agreement and small errors (laboratory: ICC 0.73 and MAPE 6.40%; free-living: ICC 0.83 and MAPE 7.10%). Decile plots indicated increasing differences between Tempo HR and the criterion as HRs increased. Such trend was less pronounced when considering the Polar A370 HR data. Tempo HR identified 62.13% (1872/3013) and 54.27% (5717/10,535) of all MVPA time points in the laboratory phase and free-living phase, respectively. Polar A370 detected 81.09% (2273/2803) and 83.55% (9323/11,158) of all MVPA time points in the laboratory phase and free-living phase, respectively. Conclusions: HR data from the examined wrist-worn trackers were reasonably accurate in both the settings, with the Polar A370 showing stronger agreement with the Polar H10 and smaller errors. Inaccuracies increased with increasing HRs; this was pronounced for Tempo HR.
- Open Access
- 24 March 2019
- Published by JMIR Publications Inc. in JMIR Preprints
Heart Rate Measures From Wrist-Worn Activity Trackers in a Laboratory and Free-Living Setting: Validation Study (Preprint)
BACKGROUND Wrist-worn activity trackers are popular, and an increasing number of these devices are equipped with heart rate (HR) measurement capabilities. However, the validity of HR data obtained from such trackers has not been thoroughly assessed outside the laboratory setting. OBJECTIVE This study aimed to investigate the validity of HR measures of a high-cost consumer-based tracker (Polar A370) and a low-cost tracker (Tempo HR) in the laboratory and free-living settings. METHODS Participants underwent a laboratory-based cycling protocol while wearing the two trackers and the chest-strapped Polar H10, which acted as criterion. Participants also wore the devices throughout the waking hours of the following day during which they were required to conduct at least one 10-min bout of moderate-to-vigorous physical activity (MVPA) to ensure variability in the HR signal. We extracted 10-second values from all devices and time-matched HR data from the trackers with those from the Polar H10. We calculated intraclass correlation coefficients (ICCs), mean absolute errors, and mean absolute percentage errors (MAPEs) between the criterion and the trackers. We constructed decile plots that compared HR data from Tempo HR and Polar A370 with criterion measures across intensity deciles. We investigated how many HR data points within the MVPA zone (≥64% of maximum HR) were detected by the trackers. RESULTS Of the 57 people screened, 55 joined the study (mean age 30.5 [SD 9.8] years). Tempo HR showed moderate agreement and large errors (laboratory: ICC 0.51 and MAPE 13.00%; free-living: ICC 0.71 and MAPE 10.20%). Polar A370 showed moderate-to-strong agreement and small errors (laboratory: ICC 0.73 and MAPE 6.40%; free-living: ICC 0.83 and MAPE 7.10%). Decile plots indicated increasing differences between Tempo HR and the criterion as HRs increased. Such trend was less pronounced when considering the Polar A370 HR data. Tempo HR identified 62.13% (1872/3013) and 54.27% (5717/10,535) of all MVPA time points in the laboratory phase and free-living phase, respectively. Polar A370 detected 81.09% (2273/2803) and 83.55% (9323/11,158) of all MVPA time points in the laboratory phase and free-living phase, respectively. CONCLUSIONS HR data from the examined wrist-worn trackers were reasonably accurate in both the settings, with the Polar A370 showing stronger agreement with the Polar H10 and smaller errors. Inaccuracies increased with increasing HRs; this was pronounced for Tempo HR.
- 24 March 2019
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessHeart Rate Measures From Wrist-Worn Activity Trackers in a Laboratory and Free-Living Setting: Validation Study (Preprint)
BACKGROUND Wrist-worn activity trackers are popular, and an increasing number of these devices are equipped with heart rate (HR) measurement capabilities. However, the validity of HR data obtained from such trackers has not been thoroughly assessed outside the laboratory setting. OBJECTIVE This study aimed to investigate the validity of HR measures of a high-cost consumer-based tracker (Polar A370) and a low-cost tracker (Tempo HR) in the laboratory and free-living settings. METHODS Participants underwent a laboratory-based cycling protocol while wearing the two trackers and the chest-strapped Polar H10, which acted as criterion. Participants also wore the devices throughout the waking hours of the following day during which they were required to conduct at least one 10-min bout of moderate-to-vigorous physical activity (MVPA) to ensure variability in the HR signal. We extracted 10-second values from all devices and time-matched HR data from the trackers with those from the Polar H10. We calculated intraclass correlation coefficients (ICCs), mean absolute errors, and mean absolute percentage errors (MAPEs) between the criterion and the trackers. We constructed decile plots that compared HR data from Tempo HR and Polar A370 with criterion measures across intensity deciles. We investigated how many HR data points within the MVPA zone (≥64% of maximum HR) were detected by the trackers. RESULTS Of the 57 people screened, 55 joined the study (mean age 30.5 [SD 9.8] years). Tempo HR showed moderate agreement and large errors (laboratory: ICC 0.51 and MAPE 13.00%; free-living: ICC 0.71 and MAPE 10.20%). Polar A370 showed moderate-to-strong agreement and small errors (laboratory: ICC 0.73 and MAPE 6.40%; free-living: ICC 0.83 and MAPE 7.10%). Decile plots indicated increasing differences between Tempo HR and the criterion as HRs increased. Such trend was less pronounced when considering the Polar A370 HR data. Tempo HR identified 62.13% (1872/3013) and 54.27% (5717/10,535) of all MVPA time points in the laboratory phase and free-living phase, respectively. Polar A370 detected 81.09% (2273/2803) and 83.55% (9323/11,158) of all MVPA time points in the laboratory phase and free-living phase, respectively. CONCLUSIONS HR data from the examined wrist-worn trackers were reasonably accurate in both the settings, with the Polar A370 showing stronger agreement with the Polar H10 and smaller errors. Inaccuracies increased with increasing HRs; this was pronounced for Tempo HR.