https://orcid.org/0009-0008-9628-3610Links
All Publications
-
Psychological Stress-Related Hematohidrosis: A Case Report
ACTAS DERMO-SIFILIOGRAFICAS
2026; 117 (10): 104702
View details for DOI 10.1016/j.ad.2026.104702
View details for Web of Science ID 001845410200001
View details for PubMedID 42342221
-
Design and rationale of the my heart counts cardiovascular health study: a large-scale, fully digital biobank, and randomized trial of large language model-driven coaching of physical activity.
American journal of preventive cardiology
2026; 28: 101565
Abstract
Cardiovascular disease remains the leading cause of global morbidity and mortality. The original My Heart Counts smartphone application demonstrated the feasibility of large-scale, fully digital recruitment and trial conduct, but was limited by platform exclusivity and the need for human experts to create text-based behavioral interventions.The next-generation My Heart Counts smartphone application is a prospective, observational cohort study with an embedded randomized crossover trial, evaluating personalized text-based coaching prompts, available in both English and Spanish. All study and trial operations will be conducted via the My Heart Counts smartphone application, re-designed using the open-source Stanford Spezi framework to support iOS, with a planned Android release in 2027. The target enrollment is N = 15,000 adults across the United States and United Kingdom. The study establishes a comprehensive digital biobank by synthesizing passive mobile health data (steps, flights climbed, heart rate, sleep, workouts), raw sensor data (e.g., accelerometry), longitudinal clinical surveys, active tasks (6-minute walk test and 12-minute Cooper run test), electrocardiograms (ECG), and electronic health record (EHR) data integrated via HL7 FHIR protocols. The embedded trial evaluates the effect of text-based coaching prompts generated by a large language model (LLM) grounded in the Transtheoretical Model of Change on daily physical activity, as compared to generic prompts.The primary endpoint of the randomized crossover trial is change in daily step count between LLM-driven and generic text-based intervention arms, analyzed using mixed-effects models. Secondary endpoints include change in mean active minutes and calorie burn over each intervention week. Other exploratory analyses include the changes in submaximal (6-minute walk test) and maximal (Cooper 12-minute run test) cardiorespiratory fitness, changes to sensor-derived biomarkers (e.g., sleep quality, resting heart rate, and heart rate variability), and association of sensor-derived biomarkers with EHR-confirmed clinical outcomes.By utilizing autonomous, LLM-driven coaching, modular software design, and cross-platform accessibility, our smartphone application-based study will provide a scalable model for inclusive and decentralized preventive care of patients with cardiovascular disease.Recruitment commenced in March 2026 and is ongoing.
View details for DOI 10.1016/j.ajpc.2026.101565
View details for PubMedID 42395096
View details for PubMedCentralID PMC13325983
-
Artificial Intelligence for Evidence Synthesis of Emerging Biologics to Improve Skeletal Health in Osteogenesis Imperfecta: Systematic Review and Meta-Analysis.
Journal of medical Internet research
2026; 28: e85840
Abstract
Osteogenesis imperfecta (OI) is a rare genetic disorder characterized by bone fragility and recurrent fractures. Emerging biologics demonstrate promise by targeting bone-remodeling pathways, yet evidence for their efficacy and safety remains fragmented and heterogeneous, and no prior systematic review in OI has incorporated artificial intelligence (AI) to synthesize it.This study aims to systematically evaluate the efficacy and safety of novel biologics in patients with OI using an AI-assisted workflow for evidence synthesis.We conducted a systematic review and meta-analysis of interventional trials of denosumab, setrusumab, teriparatide, romosozumab, and fresolimumab. Data were retrieved from PubMed, Web of Science, Embase, ScienceDirect, the Cochrane Library, and ClinicalTrials.gov up to December 1, 2025. Eligible studies enrolled individuals with OI, reported areal bone mineral density (aBMD) and/or fractures, and were randomized, nonrandomized, or single-arm studies; case series were excluded. As a methodological feature, GPT-4o was integrated into the workflow to perform a parallel 2-stage screening (title/abstract and full text) and to assist with risk of bias assessment using an adapted Cochrane RoB 2 tool. The primary outcome, percentage change in aBMD, was synthesized using a random-effects meta-analysis. GPT-4o was benchmarked against human reviewers using sensitivity, specificity, and weighted Cohen κ.Thirteen trials (n=684) were systematically reviewed, of which 10 (n=333) contributed to meta-analyses. In children, denosumab produced the greatest 12-month increase in lumbar spine aBMD (25.49%, 95% CI 17.14%-33.84%). In adults, setrusumab at 12 months yielded the highest improvement (9.38%, 95% CI 6.5%-12.26%). Across trials, no biologic significantly reduced fracture incidence compared to bisphosphonates. Safety profiles varied: denosumab was associated with a high risk of hypercalcemia in children (30.95%), whereas setrusumab had no treatment-related serious adverse events. AI achieved high sensitivity in abstract (97.4%) and full-text (88.9%) screening, and reduced total screening time by over 95%. Although there was substantial agreement with humans in the quality assessment (Cohen κ=0.778, 95% CI 0.710-0.846), the model exhibited optimism and positional biases due to reliance on probabilistic language patterns rather than structured clinical reasoning.This review is the first to synthesize and quantitatively compare skeletal outcomes across multiple biologics in OI with an AI-assisted review workflow. Denosumab and setrusumab demonstrate promising efficacy in improving lumbar spine aBMD across ages, although current evidence does not support superior fracture reduction over bisphosphonates. GPT-4o can substantially accelerate evidence synthesis but should be deployed with explicit human oversight in tasks requiring contextual understanding and clinical reasoning. These findings should be interpreted cautiously given the small and heterogeneous trial base. Taken together, our workflow presented how evidence synthesis may be scaled and operationalized in real-world rare disease research.
View details for DOI 10.2196/85840
View details for PubMedID 42430720
View details for PubMedCentralID PMC13354119
-
Patient Experiences With GLP-1 Receptor Agonists.
JAMA network open
2026; 9 (6): e2616951
Abstract
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are transformational therapies in the treatment of obesity; yet discontinuation rates are high, and patients experience rapid weight regain after stopping treatment. Few scientific publications have reported patients' experiences of taking GLP-1 RAs in a variety of health contexts outside its use as treatment for type 2 diabetes.To provide insight into the lived experiences of patients taking GLP-1 RAs to inform optimal long-term weight management treatment.This qualitative study used inductive thematic analysis of semistructured video interviews conducted between July 22 and September 10, 2025, with participants from 15 US states. Participants included adults who were taking or had previously taken GLP-1 RAs for any treatment indication, recruited through ResearchMatch and snowball sampling.Themes identified through inductive thematic analysis.Of 30 participants, 10 self-identified as men, 1 as nonbinary, and 19 as women; their mean (SD) age was 54.0 (8.8) years. A total of 23 participants were using a GLP-1 RA at time of interview, and 7 participants had stopped GLP-1 RA use. Eight key themes were extracted and categorized into 2 overarching domains. The first domain was patient-reported benefits and trade-offs. It comprised the following themes: (1) reduction in food noise, psychological hunger, or appetite; (2) GLP-1 RAs are not a standalone weight loss solution; (3) wide spectrum of adverse effects experienced; and (4) patients prepared to withstand substantial adverse effects and logistical challenges to achieve weight loss. The second domain was social, clinical, and structural context. It comprised the following themes: (5) perceived stigma associated with taking GLP-1 RAs, influenced by primary indication for prescription; (6) information provision and clinical support are essential but highly variable; (7) access challenges and cost can be prohibitive; and (8) value of shared experiences.In this qualitative study of patient experiences with GLP-1 receptor agonists, participants described GLP-1 RA therapy functioning as a facilitator rather than a replacement for lifestyle change, emphasizing the need for behavioral interventions alongside pharmacotherapy to sustain treatment benefits. Quality of care was highly variable; thus, standardized guidelines for patient education and clinical support could improve expectation management around likely adverse effects and long-term management.
View details for DOI 10.1001/jamanetworkopen.2026.16951
View details for PubMedID 42247231
-
Skin-Brain Axis: Biological Foundations and Clinical Implications.
Journal of the Academy of Consultation-Liaison Psychiatry
2026
View details for DOI 10.1016/j.jaclp.2026.04.006
View details for PubMedID 42031071
-
Purpuric Cutaneous Lesions: A Pathophysiological Framework for Diagnosis
Indian Journal of Dermatology
2026
View details for DOI 10.4103/ijd.ijd_760_24
-
Disseminated coccidioidomycosis with cutaneous, pulmonary and suspected CNS involvement in an elderly patient with Parkinson's disease.
BMJ case reports
2025; 18 (12)
Abstract
An elderly patient with recently diagnosed Parkinson's disease presented with disseminated coccidioidomycosis (DCM) manifesting as ulcerative nasal lesions, pulmonary involvement and acute-on-chronic cognitive decline. Culture-confirmed cutaneous disease, positive serology, and imaging demonstrated pulmonary and central nervous system (CNS) involvement, establishing the diagnosis. The patient achieved marked clinical improvement with high-dose oral fluconazole (800 mg, later tapered to 400 mg daily), with serial imaging documenting resolution of CNS and pulmonary lesions. Lymphocyte subset analysis revealed preserved immune function with subtle findings suggestive of age-related immunosenescence. This case demonstrates that DCM can occur in elderly patients without traditional immunocompromise and emphasises the importance of maintaining high clinical suspicion in endemic areas.
View details for DOI 10.1136/bcr-2025-269449
View details for PubMedID 41436220
-
Progressive Symmetric Erythrokeratoderma in an Adolescent With a Novel GJB3 Variant.
International journal of dermatology
2025
View details for DOI 10.1111/ijd.70200
View details for PubMedID 41366560
-
Artificial Intelligence and Medical Translation: An Editorial on the Ethical Considerations for Emerging Technologies in Dermatology.
Cureus
2025; 17 (10): e95350
Abstract
The growing demand for medical translation services in the U.S. highlights the potential of artificial intelligence (AI), large language models (LLMs) like ChatGPT (OpenAI, San Francisco, CA), to bridge language gaps. However, their use in dermatology raises ethical concerns, including information accuracy, patient privacy, dialectical variations, legal accountability, and algorithm bias for a variety of skin colors. AI models may default to informal language, leading to misunderstandings, and their limited ability to handle less common dialects poses communication challenges. The risk of "hallucination," where incorrect information is generated, and inadequate data oversight further complicate their use. In dermatology, precise translation is crucial due to the field's visual nature and the sensitivity of cosmetic concerns. Linguistic diversity can lead to misinterpretations, affecting patient care. Dermatologists must consider these ethical implications to ensure AI tools address the nuances of dermatological terminology and regional language differences, ultimately improving patient outcomes.
View details for DOI 10.7759/cureus.95350
View details for PubMedID 41293332
-
Fine-tuning LLMs in behavioral psychology for scalable health coaching.
NPJ cardiovascular health
2025; 2 (1): 48
Abstract
Personalized, smartphone-based coaching improves physical activity but relies on static, human-crafted messages. We introduce My Heart Counts (MHC)-Coach, a large language model fine-tuned on the Transtheoretical Model of Change. MHC-Coach generates messages tailored to an individual's psychology (their "stage of change"), providing personalized support to foster long-term physical activity behavior change. To evaluate MHC-Coach's efficacy, 632 participants compared human-expert and MHC-Coach interventions encouraging physical activity. Among messages matched to an individual's stage of change, 68.0% (N = 430) preferred MHC-Coach-generated messages (P < 0.001). Blinded behavioral science experts (N = 2) rated MHC-Coach messages higher than human-expert messages for perceived effectiveness (4.4 vs. 2.8) and Transtheoretical Model alignment (4.1 vs. 3.5) on a 5-point Likert scale. This work demonstrates how language models can operationalize behavioral science frameworks for personalized health coaching, showing the potential for promoting long-term physical activity and reducing cardiovascular disease risk at scale.
View details for DOI 10.1038/s44325-025-00083-5
View details for PubMedID 40994879
View details for PubMedCentralID PMC12454129
-
Post translational Modifications and Protein-Protein Interactome of Endemic Pemphigus in El Bagre, Colombia: A New Variant Analysis.
Dermatology practical & conceptual
2025; 15 (3)
Abstract
A new variant of endemic pemphigus foliaceus, El Bagre-EPF, is an orphan autoimmune disease occurring in El Bagre and neighboring municipalities in Colombia, South America.We aimed to evaluate current state-of-the-art databases and protein-protein interactomes, focusing on post-translational modifications as potential tools to study interactions among El Bagre-EPF antigen proteins.We consulted multiple databases for the known target antigens and their protein-protein interactions. We searched for their network nodes (they represent proteins; each node corresponded to all the proteins produced by a single, protein-coding gene locus). We also searched for any post-translational modifications.We identified similarities in proteins bound by several autoantigens but also found differences in protein linkages. We found that desmoglein, periplakin, desmoplakins, and proteins from subfamilies of the Armadillo repeat proteins and some spectrin domains linked to other cell junctions; these played important roles in membrane-plaque and intermediate filament junctions. A typical drawback in current databases is the lack of information on lipid-protein interactions.Our results show that state-of-the-art protein databases as tools for studies of interactomes fall short in areas including tertiary and quaternary protein interactions and in vivo protein functioning. Enhanced three-dimensional or multi-dimensional functions are required for more accurate interactome analyses.
View details for DOI 10.5826/dpc.1503a5097
View details for PubMedID 40790431
View details for PubMedCentralID PMC12339050
-
Patient-Centered Research Through Artificial Intelligence to Identify Priorities in Cancer Care.
JAMA oncology
2025
Abstract
Patient-centered research is essential for bridging the gap between research and patient care, yet patient perspectives are often inadequately represented in health research.To leverage artificial intelligence (AI) and natural language processing (NLP) to analyze a large dataset of patient messages, defining patient concerns and generating relevant research topics, and to quantify the quality of these AI-generated topics.This case series was conducted using an automated framework involving a 2-staged unsupervised NLP topic model and AI-generated research topic suggestions. The study was based on deidentified patient portal message data from individuals with breast or skin cancer at Stanford Health Care and 22 affiliated centers over July 2013 to April 2024.A widely used large language model (ChatGPT-4o [OpenAI]; April 2024) was used and guided through multiple prompt-engineering strategies to perform multilevel tasks, including knowledge interpretation and summarization (eg, interpreting and summarizing the NLP-defined topics), knowledge generation (eg, generating research ideas corresponding to patients' issues), self-reflection and correction (eg, ensuring and revising the research ideas after searching for scientific articles), and self-reassurance (eg, confirming and finalizing the research ideas).Three breast oncologists (J.L.C., A.W.K., F.R) and 3 dermatologists (K.Y.S, J.Y.T., E.L.) evaluated the meaningfulness and novelty of the AI-generated research topics using a 5-point Likert scale (1 representing exceptional to 5 representing poor). Mean (SD) scores for meaningfulness and novelty were computed for each topic.A total of 614 464 patient messages were analyzed from 25 549 individuals, 10 665 with breast cancer (98.6% female) and 14 884 had skin cancer (49.0% female). The overall mean (SD) scores for meaningfulness and novelty were 3.00 (0.50) and 3.29 (0.74), respectively, for breast cancer topics and 2.67 (0.45) and 3.09 (0.68), respectively, for skin cancer topics. One-third of the AI-suggested research topics were highly meaningful and novel when both scores were lower than the average (5 of 15 for breast cancer and 6 of 15 for skin cancer). Notably, two-thirds of the AI-suggested topics were novel (10 of 15 for breast cancer and 11 of 15 for skin cancer).This case series demonstrates that AI/NLP-driven analysis of large volumes of patient messages can generate quality research topics in cancer care that reflect patient perspectives, providing valuable guidance for future patient-centered health research endeavors.
View details for DOI 10.1001/jamaoncol.2025.0694
View details for PubMedID 40272833
- Major Depressive Disorder in Long COVID and Exposure to Pro‐Inflammatory Cytokines During Infection by SARS‐CoV‐2 Psychiatric Research and Clinical Practice 2025
-
Determining the medical Spanish translation capabilities of three artificial intelligence translation models for Mohs micrographic surgical instructions.
Journal of the American Academy of Dermatology
2024
View details for DOI 10.1016/j.jaad.2024.09.070
View details for PubMedID 39447753
-
Immunofluorescence findings in a reactivating lichenoid photoallergic chronic dermatitis (actinic reticuloid).
Photodermatology, photoimmunology & photomedicine
2024; 40 (5): e12995
Abstract
Chronic photosensitivity dermatitis (CPD) (also named actinic reticuloid) is an unusual disease classically referred often in elderly men. Affected patients have severely itchy, thickened dry skin in areas exposed to the sun throughout the years.A Caucasian female patient who worked most of her life outside who had "chronic dermatitis" in her neck started planting chrysanthemum in her garden on a sunny day. Later, she presented edema, erythema, papules, and a few vesicles in her neck with severe pruritus.A skin biopsy revealed the diagnosis of CPD, along with positive testing for ultraviolet B (UVB), minimal erythema doses (MED) for UVB (MEDB) UVA (MEDA) and PhotoPath.Direct immunofluorescence (DIF) stains using anti-human antibodies against fibrinogen, albumin, IgG, IgM, lambda, kappa, and C3c and C1q were positive at the base membrane area of the dermal epidermal junction, in the papillary dermis, as well as the neurovascular bundles in all the dermis and the extracellular matrix, especially those under the blisters.With this case, we suggest not forgetting the importance of using DIF in reactivated CPD cases in addition to the photo patch testing.
View details for DOI 10.1111/phpp.12995
View details for PubMedID 39145412
- Overall and disease-free survival in women being overweight/obesity at the diagnosis of breast cancer. A singlecenter observational study Oncología (Ecuador) 2023