CNClara Lee Ying Ngoh

Clara Lee Ying Ngoh

*This is an algorithmically generated author record based on publicly available information.

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


Clara Lee Ying Ngoh has a total of 31 co-authors. Here are the top co-authors by number of shared publications:

  • Weng Kin Wong

    Department of Medicine, National University Health System, Singapore, Singapore

    2 shared publications
  • Andrew M. T. L. Choong

    SingVaSC, Singapore Vascular Surgical Collaborative, Singapore; Department of Surgery, Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Division of Vascular and Endovascular Surgery, National University Heart Centre, Singapore; Cardiovascular Research Institute, National University of Singapore, Singapore

    1 shared publication
  • Andrew Makmur

    AIO Innovation Office, National University Health System

    1 shared publication
  • Andrew Makmur

    Department of Diagnostic Imaging, National University Hospital, Singapore, Singapore

    1 shared publication
  • Asim Shabbir

    National University Hospital, National University of Singapore, Singapore, Singapore

    1 shared publication
  • Benjamin Paik

    SingVaSC, Singapore Vascular Surgical Collaborative, Singapore

    1 shared publication

Latest Publications

  • Open Access

    The MediVoice implementation journey: ambient artificial intelligence for clinical documentation

    by Jennifer Sumner, Yee Wei Lim, Timotius Tan, Andrew Makmur, Peter Forbes, Jason Phua, Lit Sin Quek, Clara Lee Ying Ngoh

    Healthcare systems are increasingly turning to ambient Artificial Intelligence (AI) scribes to reduce documentation burden and lighten clinicians' cognitive load. In this brief research report, we introduce MediVoice, an ambient AI scribe developed and implemented within the National University Health System, Singapore. MediVoice was piloted across multiple clinical settings and rapidly evaluated through Plan–Do–Study–Act cycles. Doctors, nurses, and allied health professionals assessed its usability, accuracy, workflow fit, and potential time savings. Real-time feedback informed iterative refinement, enabling organisational learning and reinforcing the value of experimentation during early AI adoption. Broader deployment required leadership engagement, an AI community of practice, role-specific training, user champions, recognition of its value, and supporting digital and organisational infrastructure. Looking ahead, routine use will require integration with the electronic medical record, enhanced speech recognition capabilities, and robust AI governance frameworks. MediVoice's trajectory shows that while ambient AI scribes can offer meaningful benefits, success requires more than technical capability. Effective implementation needs continuous adaptation, workflow alignment, cross-professional engagement, governance, and organisational readiness. This case study offers practical lessons for health systems seeking to introduce ambient AI tools within clinical environments.


  • Open Access

    The Use of Large Language Models Tuned with Socratic Methods on the Impact of Medical Students' Learning: A Randomised Controlled Trial (Preprint)

    by Cai Ling Yong, Mohammad Shaheryar Furqan, James Wai Kit Lee, Andrew Makmur, Ragunathan Mariappan, Clara Lee Ying Ngoh, Kee Yuan Ngiam

    BACKGROUND Large Language Models (LLM) are AI models that can generate conversational content based on a trained specified source of information (corpus). OBJECTIVE The aim is to use these corpus-trained LLMs to limit the content offered by LLM, then using prompt engineering to teach using Socratic methods. METHODS Two chatbots were created and deployed, powered by OpenAI’s GPT-3.5 model, with a medical-school textbook corpus. The first chatbot generates a brief summary and open-ended question. The second chatbot generates a case vignette from its pre-trained clinical cases, prompting users for a diagnosis. Both chatbots reply to the user’s response, commenting on the accuracy and asks further questions to encourage critical thinking. A randomised controlled trial was conducted on two groups comprising third year medical students. One group used both chatbots for 10 minutes while the other read the medical textbook. A 15-question test was administered to both groups before and after the intervention. RESULTS Forty students participated in the study. The average of the group before and after reading the textbook (n=20) are 3.9 +/- 1.0 and 7.6 +/- 1.5 respectively (p<0.001). The average of the group before and after using the bot (n=20) are 3.9 +/- 0.9 and 12.8 +/- 1.6 respectively (p<0.001). The respective increase in results was 3.7 and 8.9. CONCLUSIONS Medical students’ learning showed a better performance using a LLM based chatbot compared to self-reading of medical information assessed using a standardised test. More studies are required to determine if LLM-based pedagogical methods are superior to standard education.


  • Clinical characteristics and outcomes of exertional rhabdomyolysis after indoor spinning: a systematic review

    by Yoshio Masuda, Rachel Wam, Benjamin Paik, Clara Ngoh, Andrew Mtl Choong, Jun Jie Ng

    An increasing number of patients are diagnosed with exertional rhabdomyolysis secondary to indoor spinning. We performed a systematic review to characterize the clinical features of this new clinic...


  • Open Access

    Multiple Myeloma-Associated Light Chain Amyloidosis and a Proposed Approach to Monoclonal Immunoglobulin-Associated Renal Disease

    by Zi Yun Chang, Thomas Paulraj Thamboo, Clarice Shi Hui Choong, Weng Kin Wong, Clara Lee Ying Ngoh

    Many challenges remain in diagnosing monoclonal immunoglobulin-associated renal disease, despite widespread application of immunofluorescence (IF) and immunohistochemistry. Here, we report a newly diagnosed case of multiple myeloma with clinical suspicion of renal amyloidosis, which had negative IF staining for kappa and lambda light chains in the glomeruli. Although laser microdissection and mass spectrometry-based proteomic analysis have emerged as important tools for amyloid typing in the literature, such facilities are still not widely available in Asia. We propose that a clinicopathological algorithm for the evaluation of organized monoclonal renal deposits, together with a combined nephrological-haematological approach, will still be adequate to generate an unequivocal diagnosis in the majority of cases.


  • Comparison between lung ultrasonography and current methods for volume assessment in Asian chronic hemodialysis patients

    by Clara L. Y. Ngoh, Hazel L. Teng, Yan T. Chua, Christopher C. H. Leo, Weng K. Wong

    Introduction: Volume assessment in end‐stage kidney disease patients on hemodialysis (HD) remains inadequate by existing methods: clinical examination, bioimpedance spectroscopy, measurement of inferior vena cava diameter by ultrasound (IVCD), and plasma B‐type natriuretic peptide (NT‐pro BNP). This study aims to compare the performance of lung ultrasound against existing methods for volume assessment in a HD cohort.Methods: Two nephrologists independently performed 28‐point lung ultrasound immediately before and after midweek HD in 50 patients. Lung congestion was classified into mild, moderate, and severe categories based on lung ultrasound findings. Clinical examination for crepitations and oedema, change in hydration status (∆HS) measured by bioimpedance spectroscopy, NT‐pro BNP, IVCD during inspiration (IVCDimin), expiration (IVCDimax), and inferior vena cava collapsibility index were also assessed before and after midweek HD.Findings: In all, 61% of patients with normohydration status by bioimpedance spectroscopy had moderate or severe lung congestion on lung ultrasound. There were significant correlations between predialysis lung ultrasound, and NT‐pro BNP (r = 0.432, P = 0.004), ∆HS (r = 0.447, P < 0.001), and IVCD parameters (P < 0.05). Some correlations weakened postdialysis (∆HS [r = 0.322, P = 0.01] and IVCDimax [r = 0.307, P = 0.03]), whereas NT‐pro BNP and ∆HS paradoxically increased in 28% and 30% of the cohort, respectively. On receiver operator curve analysis, most methods of volume assessment had limited discriminatory power to detect mild lung congestion.Discussion: Lung ultrasound demonstrates some comparability with existing volume assessment methods in Asian dialysis patients. However, it appears more effective at detecting subclinical pulmonary congestion, and tracking fluid changes real‐time compared to bioimpedance spectroscopy and NT‐pro BNP.


  • Effect of weight loss after bariatric surgery on kidney function in a multiethnic Asian population

    by Clara Lee Ying Ngoh, Jimmy Bok Yan So, Ho Yee Tiong, Asim Shabbir, Boon Wee Teo