Andy Cheuk Nam Hwang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Andy Cheuk Nam Hwang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Elaine Yuen Phin Lee
Department of Diagnostic Radiology, School of Clinical Medicine, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong, China
3 shared publications - Gordon Tin Chun Wong
Department of Anaesthesiology, LKS Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong SAR, China
2 shared publications - Michael Co
Centre for Education and Learning, Department of Surgery University of Hong Kong Hong Kong SAR China
2 shared publications - Pauline Yeung Ng
Adult Intensive Care Unit, Queen Mary Hospital, Hong Kong
2 shared publications - Peng Cao
Department of Diagnostic Radiology The University of Hong Kong Hong Kong China
2 shared publications - Rahul Singh
Department of Diagnostic Radiology, Clinical School of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong SAR, China
2 shared publications
Latest Publications
- Open Access
- 1 July 2026
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 16 (7) , 537
Impact of deep-learning image reconstruction on multiplexed sensitivity encoding diffusion-weighted imaging in the female pelvis
Background: Diffusion-weighted imaging (DWI) in the female pelvis is degraded by artefacts. The study aimed to evaluate the impact of deep-learning image reconstruction (DLRecon) on the image quality of multiplexed sensitivity encoding (MUSE) DWI in the female pelvis. Methods: Female patients scheduled for pelvic magnetic resonance imaging (MRI) with 2-shot MUSE DWI were prospectively recruited. A subset of patients underwent paired 4-shot MUSE DWI. Images were qualitatively reviewed by two radiologists in terms of overall image quality, artefacts, lesions conspicuity and sharpness using a 5-point scale, with and without DLRecon. Intra-observer and inter-observer agreements were evaluated. The signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and apparent diffusion coefficient (ADC) of normal structures and lesions were quantitatively compared between 2-shot MUSE with and without DLRecon. Statistical analyses were performed using Cohen’s kappa, Wilcoxon signed-rank test and paired-sample t-test with Bonferroni correction. Results: Among the 65 female patients evaluated, 27 malignant lesions and 32 benign lesions were detected, while no pathology was found in 6 patients. DLRecon significantly improved the image quality in all aspects evaluated compared to 2-shot MUSE without DLRecon: overall image quality, artefacts, lesions conspicuity and sharpness (N=65, PConclusions: DLRecon improved image quality of MUSE DWI in the female pelvis by reducing impact of artefacts, improving lesion conspicuity and image sharpness, while increasing SNR and CNR but preserving the stability of ADC quantification.
- 1 July 2026
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 16 (7) , 537
Open AccessImpact of deep-learning image reconstruction on multiplexed sensitivity encoding diffusion-weighted imaging in the female pelvis
Background: Diffusion-weighted imaging (DWI) in the female pelvis is degraded by artefacts. The study aimed to evaluate the impact of deep-learning image reconstruction (DLRecon) on the image quality of multiplexed sensitivity encoding (MUSE) DWI in the female pelvis. Methods: Female patients scheduled for pelvic magnetic resonance imaging (MRI) with 2-shot MUSE DWI were prospectively recruited. A subset of patients underwent paired 4-shot MUSE DWI. Images were qualitatively reviewed by two radiologists in terms of overall image quality, artefacts, lesions conspicuity and sharpness using a 5-point scale, with and without DLRecon. Intra-observer and inter-observer agreements were evaluated. The signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and apparent diffusion coefficient (ADC) of normal structures and lesions were quantitatively compared between 2-shot MUSE with and without DLRecon. Statistical analyses were performed using Cohen’s kappa, Wilcoxon signed-rank test and paired-sample t-test with Bonferroni correction. Results: Among the 65 female patients evaluated, 27 malignant lesions and 32 benign lesions were detected, while no pathology was found in 6 patients. DLRecon significantly improved the image quality in all aspects evaluated compared to 2-shot MUSE without DLRecon: overall image quality, artefacts, lesions conspicuity and sharpness (N=65, PConclusions: DLRecon improved image quality of MUSE DWI in the female pelvis by reducing impact of artefacts, improving lesion conspicuity and image sharpness, while increasing SNR and CNR but preserving the stability of ADC quantification.
- Open Access
- 9 March 2026
- Published by JMIR Publications Inc. in JMIR Medical Education
- Vol. 12, e72110
Development of a Deep Learning–Based Feedback Model to Assist Medical Students Learning Renal Ultrasound Acquisition: Mixed Methods Study
Background: Point-of-care ultrasound training is being increasingly integrated into undergraduate medical education, leading to a substantial demand for trained faculty to provide instruction and feedback. Objective: This study aimed to develop an adjunct tool, a deep learning–based feedback model, to facilitate student learning. Methods: Renal ultrasound images (N=2807) were used to train a cascaded deep learning–based feedback model that classified images into three categories: optimal, suboptimal, and incorrect. Suboptimal images were further subcategorized as images with artifact, incorrect gain, and/or incorrect positioning. The model was deployed among year 5 medical students receiving bedside ultrasound training, who were invited to upload renal ultrasound images to an online platform for automated image quality grading and feedback. A mixed methods analysis was used to evaluate students’ learning experience. Focus group interviews were organized to qualitatively analyze the successes and challenges of implementation. Quantitative analysis was based on responses to a 5-point Likert scale questionnaire and performance on the objective structured clinical examination (OSCE). Objective structured clinical examination scores were compared with mean OSCE scores from the 2 years preceding implementation of the deep learning–based feedback model. Results: Focus group interviews identified that the deep learning–based feedback model encouraged self-regulated learning but also recognized that discordant curricular design and hardware limitations impeded its use. The 11-item online questionnaire had a response rate of 42.4% (98/231 students). Among respondents, 32% (31/98) to 48% (47/98) found the model helpful in assisting ultrasound training (Likert score of 4‐5 for items 1-3), while 49% (48/98) to 76% (74/98) were satisfied with its usability and their interaction with the model (Likert score of 4‐5 for items 4-11). The mean OSCE score was 9.73 (SD 0.76) out of 10, compared with mean scores of 9.35 (SD 1.03; P =.06) and 9.45 (SD 0.97; P =.15) out of 10 in the 2 individual years preceding implementation of the model. Conclusions: A cascaded deep learning–based feedback model was developed and deployed among year 5 medical students receiving bedside ultrasound training, with positive learner responses and enhanced self-regulated learning. The innovation was associated with increased student engagement and improved ultrasound skill acquisition among novice learners.
- 9 March 2026
- Published by JMIR Publications Inc. in JMIR Medical Education
- Vol. 12, e72110
Open AccessDevelopment of a Deep Learning–Based Feedback Model to Assist Medical Students Learning Renal Ultrasound Acquisition: Mixed Methods Study
Background: Point-of-care ultrasound training is being increasingly integrated into undergraduate medical education, leading to a substantial demand for trained faculty to provide instruction and feedback. Objective: This study aimed to develop an adjunct tool, a deep learning–based feedback model, to facilitate student learning. Methods: Renal ultrasound images (N=2807) were used to train a cascaded deep learning–based feedback model that classified images into three categories: optimal, suboptimal, and incorrect. Suboptimal images were further subcategorized as images with artifact, incorrect gain, and/or incorrect positioning. The model was deployed among year 5 medical students receiving bedside ultrasound training, who were invited to upload renal ultrasound images to an online platform for automated image quality grading and feedback. A mixed methods analysis was used to evaluate students’ learning experience. Focus group interviews were organized to qualitatively analyze the successes and challenges of implementation. Quantitative analysis was based on responses to a 5-point Likert scale questionnaire and performance on the objective structured clinical examination (OSCE). Objective structured clinical examination scores were compared with mean OSCE scores from the 2 years preceding implementation of the deep learning–based feedback model. Results: Focus group interviews identified that the deep learning–based feedback model encouraged self-regulated learning but also recognized that discordant curricular design and hardware limitations impeded its use. The 11-item online questionnaire had a response rate of 42.4% (98/231 students). Among respondents, 32% (31/98) to 48% (47/98) found the model helpful in assisting ultrasound training (Likert score of 4‐5 for items 1-3), while 49% (48/98) to 76% (74/98) were satisfied with its usability and their interaction with the model (Likert score of 4‐5 for items 4-11). The mean OSCE score was 9.73 (SD 0.76) out of 10, compared with mean scores of 9.35 (SD 1.03; P =.06) and 9.45 (SD 0.97; P =.15) out of 10 in the 2 individual years preceding implementation of the model. Conclusions: A cascaded deep learning–based feedback model was developed and deployed among year 5 medical students receiving bedside ultrasound training, with positive learner responses and enhanced self-regulated learning. The innovation was associated with increased student engagement and improved ultrasound skill acquisition among novice learners.
- Open Access
- 4 February 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Assisting medical students learning in renal ultrasound acquisition: Development of a deep learning-based feedback model (Preprint)
BACKGROUND Point-of-care ultrasound (US) training is increasingly integrated into undergraduate medical education, leading to huge demand for trained faculty members to provide training and feedback. OBJECTIVE This study aimed to develop an adjunct tool, a deep learning-based feedback model to facilitate student learning. METHODS Renal US images (n=2807) were utilized to train a cascaded deep learning-based feedback model, which classified images into 3 labels: optimal, sub-optimal and wrong. Sub-optimal images were further sub-labelled to images with artifact, incorrect gain and/or incorrect position. The model was deployed among year 5 students receiving bedside US training, who were invited to upload renal US images to an online platform for the model to grade the image quality and provide feedback. Mixed-method analysis was used. Students were invited to fill in an online questionnaire to evaluate their learning experience with respect to the effectiveness of the model in enhancing ultrasound training and its usability based on a 5-point scale. Focused group interviews were organized to gain more insights to the successes and challenges of implementation. RESULTS The cohort consisted of 231 year 5 medical students and 98 (42.4%) of them filled in the questionnaire. Among them, 32-48% found the model effective in assisting US training (score 4-5) and 49-75% were satisfied with the usability and their interactions with the model. Focused group interviews identified the model encouraged those who were able to regulate their own learning to be more engaged, but also recognised discordant curricular design and hardware limitations impeded the use of the model. CONCLUSIONS A cascaded deep learning-based feedback model was developed and deployed with positive responses from year 5 medical students receiving bedside US training to support self-regulated learning.
- 4 February 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessAssisting medical students learning in renal ultrasound acquisition: Development of a deep learning-based feedback model (Preprint)
BACKGROUND Point-of-care ultrasound (US) training is increasingly integrated into undergraduate medical education, leading to huge demand for trained faculty members to provide training and feedback. OBJECTIVE This study aimed to develop an adjunct tool, a deep learning-based feedback model to facilitate student learning. METHODS Renal US images (n=2807) were utilized to train a cascaded deep learning-based feedback model, which classified images into 3 labels: optimal, sub-optimal and wrong. Sub-optimal images were further sub-labelled to images with artifact, incorrect gain and/or incorrect position. The model was deployed among year 5 students receiving bedside US training, who were invited to upload renal US images to an online platform for the model to grade the image quality and provide feedback. Mixed-method analysis was used. Students were invited to fill in an online questionnaire to evaluate their learning experience with respect to the effectiveness of the model in enhancing ultrasound training and its usability based on a 5-point scale. Focused group interviews were organized to gain more insights to the successes and challenges of implementation. RESULTS The cohort consisted of 231 year 5 medical students and 98 (42.4%) of them filled in the questionnaire. Among them, 32-48% found the model effective in assisting US training (score 4-5) and 49-75% were satisfied with the usability and their interactions with the model. Focused group interviews identified the model encouraged those who were able to regulate their own learning to be more engaged, but also recognised discordant curricular design and hardware limitations impeded the use of the model. CONCLUSIONS A cascaded deep learning-based feedback model was developed and deployed with positive responses from year 5 medical students receiving bedside US training to support self-regulated learning.