LHLan T. Ho-Pham

Lan T. Ho-Pham

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Lan T. Ho-Pham has a total of 255 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    An integrated deep learning and large language model for automated knee osteoarthritis grading and reporting from the Vietnam Osteoporosis Study

    by Dinh T. Nguyen, Duy K. Hoang, Linh N. Luu, An T. Truong, Hien Q. Kha, Huy T. Tran, Phu D. Ho, Thach S. Tran, Sai Ho Ling, David J. Hunter, et al.


  • Muscle mass, not fat mass, predicts vertebral fracture risk: Vietnam osteoporosis study

    by Hoa T. Nguyen, Nhung H. T. Thai, Tuan V. Nguyen, Lan T. Ho-Pham


  • Open Access

    Artificial intelligence-assisted screening reveals high prevalence of osteoporosis in Vietnamese adults using pelvic and hip radiographs

    by Dat Minh Nguyen, Chih-Hsing Wu, Tuan Van Nguyen, Lan T. Ho-Pham, Kim Thi Hoang Dang, Hy Van Nguyen, Sung-Yen Lin, Chung-Hwan Chen, Ta-Wei Tai


  • Muscle mass, not fat mass, predicts vertebral fracture risk: Vietnam osteoporosis study

    by Hoa T. Nguyen, Bao T. Nguyen, Quyen T. Nguyen, An V. Tran, Nhung H. T. Thai, Quyen X. T. Phan, Linh D. Mai, Thach S. Tran, Tuan V. Nguyen, Lan T. Ho-Pham

    Summary The association between body weight and vertebral fracture is established, but whether muscle mass or fat mass drives this relationship remains unclear. In Vietnamese adults, our study demonstrates that muscle mass, rather than fat mass, predicts vertebral fracture risk. Both low and excessive muscle mass significantly increase this risk, indicating a U-shaped relationship. Purpose The purpose is to examine the relationship between various body composition measures and vertebral fracture in a Vietnamese population. Methods This cross-sectional study included 1,372 Vietnamese adults aged ≥ 50 years from the Vietnam Osteoporosis Study. Body composition was measured using dual-energy X-ray absorptiometry to derive lean body mass index (LBMI), appendicular skeletal muscle mass index (ASMI), body fat mass index (BFMI), and percentage-based indices. Vertebral fractures were identified via thoracolumbar radiographs using Genant’s semiquantitative method. Adjusted logistic regression models and restricted cubic splines were used to assess associations and explore nonlinear relationships. Results Vertebral fractures were diagnosed in 173 participants (12.6%). Compared with those without fractures, affected participants were older (63.4 ± 9.6 vs 58.8 ± 7.3 years; p < 0.001) and more often male (38.7% vs 29.4%; p = 0.017). Both LBMI and ASMI showed U-shaped associations with fracture risk, with higher odds at the extremes; for ASMI, the lowest and highest quartiles had increased odds of fracture (OR 1.72; 95% CI 1.02–2.95 and OR 2.59; 95% CI 1.36–5.02, respectively). In contrast, BFMI was not significantly associated with fracture risk (p = 0.365). Higher trunk lean percentage was consistently related to fracture (ORs 1.82–2.54 across non-reference quartiles), whereas total body and leg lean percentages were associated with increased risk only in the highest quartile (OR 2.18; 95% CI 1.07–4.45 and OR 2.06; 95% CI 1.03–4.12). Conclusions In Vietnamese adults, extremes of muscle mass (low and high), rather than fat mass, are associated with higher vertebral fracture risk. This emphasizes incorporating muscle mass evaluation into fracture risk assessment beyond reliance on BMI alone.


  • Open Access

    Association between knee osteoarthritis and volumetric bone mineral density

    by Duy K. Hoang, An T. Truong, Linh N. Luu, Tan D. Nguyen, Nhan M. Le, Huy G. Nguyen, Lan T. Ho-Pham, David J. Hunter, Tuan V. Nguyen

  • Open Access

    Transition from non-diabetes to diabetes: a population-based longitudinal study in Vietnamese individuals

    by Lan T. Ho-Pham, Linh D. Mai, Thach S. Tran, Tuan V. Nguyen

    Objective: This study sought to determine the incidence of T2DM and identify risk factors among a Vietnamese urban cohort without previous T2DM diagnoses to support early intervention. Methods: This population-based prospective study included 1971 individuals (1368 women) free of T2DM at baseline. Participants were re-assessed after a median follow-up of 2.0 years. Glycosylated hemoglobin (HbA1c) levels were measured using high-pressure liquid chromatography (HPLC) analyzers (ADAMS A1c HA-8160, Arkray, Kyoto, Japan). T2DM was defined as an HbA1c value ≥ 6.5%, while prediabetes was defined as an HbA1c value between 5.7% and 6.4%. Incident T2DM was identified as newly developed diabetes from a non-diabetic status. The association between potential risk factors and T2DM was analyzed using logistic regression. Results: At baseline, 1137 participants (58%) were classified as ‘Normal’ and 834 (42%) as ‘Pre-diabetes’. Over the follow-up, 16% of initially normal individuals progressed to prediabetes, and 10% of those with prediabetes progressed to diabetes. Overall, 4.3% transitioned to diabetes. Advancing age (OR per year: 1.03; 95% CI, 1.01 to 1.06) and higher BMI (OR per unit: 1.20; 95% CI, 1.12 to 1.27) were significantly associated with an increased risk of developing T2DM. A predictive model incorporating age, body mass index (BMI), and baseline HbA1c resulted in an area under the ROC curve of 0.90. Conclusions: The study revealed a notable 4.3% incidence of T2DM over two years among Vietnamese adults. Significant predictors included age and BMI, underscoring the need for targeted preventive measures focusing on weight management and monitoring in high-risk groups.

  • Open Access

    Minimizing and quantifying uncertainty in AI-informed decisions: Applications in medicine

    by Samuel D. Curtis, Sambit Panda, Adam Li, Haoyin Xu, Yuxin Bai, Itsuki Ogihara, Eliza O’REilly, Yuxuan Wang, Lisa Dobbyn, Maria Popoli, et al.

    AI is now a cornerstone of modern dataset analysis. In many real world applications, practitioners are concerned with controlling specific kinds of errors, rather than minimizing the overall number of errors. For example, biomedical screening assays may primarily be concerned with mitigating the number of false positives rather than false negatives. Quantifying uncertainty in AI-based predictions, and in particular those controlling specific kinds of errors, remains theoretically and practically challenging. We develop a strategy called multidimensional informed generalized hypothesis testing (MIGHT) which we prove accurately quantifies uncertainty and confidence given sufficient data, and concomitantly controls for particular error types. Our key insight was that it is possible to integrate canonical cross-validation and parametric calibration procedures within a nonparametric ensemble method. Simulations demonstrate that while typical AI based-approaches cannot be trusted to obtain the truth, MIGHT can be. We apply MIGHT to answer an open question in liquid biopsies using circulating cell-free DNA (ccfDNA) in individuals with or without cancer: Which biomarkers, or combinations thereof, can we trust? Performance estimates produced by MIGHT on ccfDNA data have coefficients of variation that are often orders of magnitude lower than other state of the art algorithms such as support vector machines, random forests, and Transformers, while often also achieving higher sensitivity. We find that combinations of variable sets often decrease rather than increase sensitivity over the optimal single variable set because some variable sets add more noise than signal. This work demonstrates the importance of quantifying uncertainty and confidence—with theoretical guarantees—for the interpretation of real-world data.


  • Open Access

    Fragmentation signatures in cancer patients resemble those of patients with vascular or autoimmune diseases

    by Samuel D. Curtis, Tingshan Liu, Yuxin Bai, Yuxuan Wang, Sambit Panda, Adam Li, Haoyin Xu, Eliza O’REilly, Lisa Dobbyn, Maria Popoli, et al.

    Multiple case-controlled studies have shown that analyzing fragmentation patterns in plasma cell-free DNA (cfDNA) can distinguish individuals with cancer from healthy controls. However, there have been few studies that investigate various types of cfDNA fragmentomics patterns in individuals with other diseases. We therefore developed a comprehensive statistic, called fragmentation signatures, that integrates the distributions of fragment positioning, fragment length, and fragment end-motifs in cfDNA. We found that individuals with venous thromboembolism, systemic lupus erythematosus, dermatomyositis, or scleroderma have cfDNA fragmentation signatures that closely resemble those found in individuals with advanced cancers. Furthermore, these signatures were highly correlated with increases in inflammatory markers in the blood. We demonstrate that these similarities in fragmentation signatures lead to high rates of false positives in individuals with autoimmune or vascular disease when evaluated using conventional binary classification approaches for multicancer earlier detection (MCED). To address this issue, we introduced a multiclass approach for MCED that integrates fragmentation signatures with protein biomarkers and achieves improved specificity in individuals with autoimmune or vascular disease while maintaining high sensitivity. Though these data put substantial limitations on the specificity of fragmentomics-based tests for cancer diagnostics, they also offer ways to improve the interpretability of such tests. Moreover, we expect these results will lead to a better understanding of the process—most likely inflammatory—from which abnormal fragmentation signatures are derived.