Xianyan Chen’s research while affiliated with University of Georgia and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (34)


Exploring hyperelastic material model discovery for human brain cortex: multivariate analysis vs. artificial neural network approaches
  • Article

February 2025

·

18 Reads

Journal of the Mechanical Behavior of Biomedical Materials

·

·

Xianyan Chen

·

[...]

·





P-1019. Incidence And Risk Factors for Invasive Fungal Infections in Pre-Transplant and Transplant-Ineligible Multiple Myeloma in the U.S.: A Claims Analysis (2017-2021)

January 2025

Open Forum Infectious Diseases

Background Current research on risk factors for invasive fungal infections (IFIs) in multiple myeloma (MM) has limitations due to heterogeneous patient populations, including post-transplant, or reliance on subgroup analyses. This study aimed to address this gap by evaluating the incidence of IFIs and identifying risk factors in patients with MM receiving treatment before or ineligible for transplant. Methods We analyzed data from the Merative MarketScan Database (2017-2021) to identify adults (≥ 18 years) diagnosed with and treated for MM with proteasome inhibitors, lymphodepleting agents, thalidomide or derivatives, anti-SLAMF7 monoclonal antibodies, or exportin 1 inhibitors, with or without dexamethasone. We evaluated the incidence and risk factors for IFIs following anti-MM therapy initiation. All patients were followed for at least one year. Patients without an IFI were censored at the end of the study period. Results Among 3054 individuals with MM, 6% (n=195) were diagnosed with an IFI. Candidiasis was most common (87%), followed by pneumocystis (6.2%) and aspergillosis (3.6%). Patients with an IFI were younger with a higher burden of comorbidities compared to those without an IFI (table 1). Notably, neutropenia, thrombocytopenia, chronic heart failure, chronic liver failure, hypercalcemia, and hyperglycemia were significantly more common in the IFI group. Anti-MM therapies were similar between groups, with a high prevalence of both thalidomide or derivatives and dexamethasone (table 2). Antifungal prophylaxis was uncommon while nearly half of each group received antibacterial prophylaxis. Patients with an IFI were more likely to have received multiple lines of anti-MM therapy. Multivariate analysis identified recent dexamethasone use (HR 5.85, 95% CI: 4.08-8.40), neutropenia (HR 2.77, 95% CI: 1.87-4.11), and a greater number of anti-MM therapies within the preceding year (HR 2.15, 95% CI: 1.71-2.69) as significant risk factors for IFI. Conclusion Candidiasis was the most common IFI in patients with MM. Younger age, higher comorbidity burden, and neutropenia were associated with IFIs. Additionally, recent dexamethasone use and a higher number of prior anti-MM therapies significantly increased the risk of IFIs. Disclosures All Authors: No reported disclosures


P-1929. Impact of the COVID-19 Pandemic on the Risk Factors and Outcomes Associated with Candidemia

January 2025

Open Forum Infectious Diseases

Background Cases of candidemia surged during the COVID-19 pandemic, possibly due to the intensive treatments used in critically ill patients. This study aims to explore the impact of the pandemic on risk factors and clinical outcomes associated with candidemia. Methods We retrospectively analyzed patients ≥ 18 years diagnosed with candidemia at a single facility in Albany, Georgia, between January 2017 and May 2023. Patients were categorized into three groups based on their COVID-19 status and admission date (January 21, 2020, marking the first US case): 1) pre-pandemic (before January 21, 2020), 2) pandemic with COVID-19 (after January 21, 2020), and 3) pandemic without COVID-19 (after January 21, 2020). We compared risk factors for candidemia and in-hospital mortality across these groups. Results A total of 89 patients were included, with 44% pre-pandemic and 56% during the pandemic (half with COVID-19) (Table 1). Baseline characteristics were similar, except for a higher median number of comorbidities in the pre-pandemic group (p=0.008). Notably, the COVID-19 group had a significantly higher proportion of patients with no comorbidities (12%) compared to the pre-pandemic (0%) and pandemic without COVID-19 (4%) groups (p=0.01). Patients with COVID-19 were predominantly diagnosed with candidemia in the ICU (92%) compared to pre-pandemic (49%) and pandemic without COVID-19 (44%) groups (p< 0.001). Additionally, mechanical ventilation and vascular catheters were more frequent in the COVID-19 group (88% and 96%, respectively) compared to the pre-pandemic (31% and 62%) and pandemic without COVID-19 (40% and 68%) groups (p< 0.001 for both) (Table 2). While immunosuppressive medications were uncommon, glucocorticoid use was significantly higher in the COVID-19 group (76%) compared to the pre-pandemic (18%) and pandemic without COVID-19 (24%) groups (p=0.008). In-hospital mortality was highest among patients with COVID-19 (76%) compared to patients before the pandemic (44%) and those without COVID-19 during the pandemic (32%) (p< 0.005). Conclusion Patients with COVID-19 displayed a higher prevalence of candidemia risk factors and significantly worse clinical outcomes, including higher in-hospital mortality, compared to the other groups. Disclosures All Authors: No reported disclosures


Role of Data-driven Regional Growth Model in Shaping Brain Folding Patterns
  • Article
  • Full-text available

January 2025

·

27 Reads

Soft Matter

The surface morphology of the developing mammalian brain is crucial for understanding brain function and dysfunction. Computational modeling offers valuable insights into the underlying mechanisms for early brain folding. Recent...

Download

Unsupervised machine learning analysis to identify patterns of ICU medication use for fluid overload prediction

January 2025

·

9 Reads

Pharmacotherapy

Background Fluid overload (FO) in the intensive care unit (ICU) is common, serious, and may be preventable. Intravenous medications (including administered volume) are a primary cause for FO but are challenging to evaluate as a FO predictor given the high frequency and time‐dependency of their use and other factors affecting FO. We sought to employ unsupervised machine learning methods to uncover medication administration patterns correlating with FO. Methods This retrospective cohort study included 927 adults admitted to an ICU for ≥72 h. FO was defined as a positive fluid balance ≥7% of admission body weight. After reviewing medication administration record data in 3‐h periods, medication exposure was categorized into clusters using principal component analysis (PCA) and Restricted Boltzmann Machine (RBM). Medication regimens of patients with and without FO were compared within clusters to assess their temporal association with FO. Results FO occurred in 127 (13.7%) of 927 included patients. Patients received a median (interquartile range) of 31(13–65) discrete intravenous medication administrations over the 72‐h period. Across all 47,803 intravenous medication administrations, 10 unique medication clusters, containing 121 to 130 medications per cluster, were identified. The mean number of Cluster 7 medications administered was significantly greater in the FO cohort compared with patients without FO (25.6 vs.10.9, p < 0.0001). A total of 51 (40.2%) of 127 unique Cluster 7 medications were administered in more than five different 3‐h periods during the 72‐h study window. The most common Cluster 7 medications included continuous infusions, antibiotics, and sedatives/analgesics. Addition of Cluster 7 medications to an FO prediction model including the Acute Physiologic and Chronic Health Evaluation (APACHE) II score and receipt of diuretics improved model predictiveness from an Area Under the Receiver Operation Characteristic (AUROC) curve of 0.719 to 0.741 ( p = 0.027). Conclusions Using machine learning approaches, a unique medication cluster was strongly associated with FO. Incorporation of this cluster improved the ability to predict FO compared to traditional prediction models. Integration of this approach into real‐time clinical applications may improve early detection of FO to facilitate timely intervention.


Figure 1. Directed acyclic graph for the causal pathway relating comprehensive medication management to medications that patients receive and patient outcomes.
Results of propensity-matched analysis Pharmacist interventions >3 during the ICU stay (n = 4,029)
Effect of comprehensive medication management on mortality in critically ill patients

October 2024

·

33 Reads

Background: Medication management in the intensive care unit (ICU) is causally linked to both treatment success and potential adverse drug events (ADEs), often associated with deleterious consequences. Patients with higher severity of illness tend to require more management. The purpose of this evaluation was to explore the effect of comprehensive medication management (CMM) on mortality in critically ill patients. Methods: In this retrospective cohort study of adult ICU patients, CMM was measured by critical care pharmacist (CCP) medication interventions. Propensity score matching was performed to generate a balanced 1:1 matched cohort, and logistic regression was applied for estimating propensity scores. The primary outcome was the odds of hospital mortality. Hospital and ICU length of stay were also assessed. Results: In a cohort of 10,441 ICU patients, the unadjusted mortality rate was 11% with a mean APACHE II score of 9.54 and Medication Regimen Complexity-Intensive Care Unit (MRC-ICU) score of 5.78. Compared with CCP interventions less than 3, more CCP interventions was associated with a significantly reduced risk of mortality (estimate -0.04, 95% confidence interval -0.06 - -0.03, p < 0.01) and shorter length of ICU stay (estimate -2.77, 95% CI -2.98 - - 2.56, p < 0.01). Conclusions: The degree by which CCPs deliver CMM in the ICU is directly associated with reduced hospital mortality independent of patient characteristics and medication regimen complexity.



Citations (9)


... Application to rubber-like materials and capturing rate-independent inelasticity using two feed-forward neural networks (FFN) for the inelastic evolution and the free energy has been discussed by Ghaderi et al. (2020), Masi et al. (2021) and Masi and Stefanou (2022). Automated model discovery to link stresses and strains has been applied by Martonová et al. (2024) and Peng et al. (2021) and on human brain cortex by Hou et al. (2024) and have been embedded into finite element analysis software . ...

Reference:

Accounting for plasticity: An extension of inelastic Constitutive Artificial Neural Networks
Automated Data-Driven Discovery of Material Models Based on Symbolic Regression: A Case Study on the Human Brain Cortex
  • Citing Article
  • September 2024

Acta Biomaterialia

... 500 For education purposes, multiple authors have raised the discussion about how LLMs can be used to support educators' and instructors' daily work. [523][524][525][526][527] Finally, in this direction, Mouriño et al. 510 developed i-Digest, an agent whose perception module can understand audio tracks and video recordings. These audio recordings are transcribed to text using the Whisper 138 model, and therefore, i-Digest is a digital tutor that generates questions to help students test their knowledge about the course material. ...

Assessing Large Language Models in Mechanical Engineering Education: A Study on Mechanics-Focused Conceptual Understanding

... Also, many researchers have used the machine learning method to study the behavior of structures. At present, significant attention is directed towards evaluating fatigue properties [117][118][119], forecasting the mechanical properties of metamaterials [120][121][122], inversely designing metamaterials [123][124][125][126], as well as designing auxetic tubes [83]. By utilizing appropriate machine learning models, many researchers have successfully made accurate predictions of mechanical behavior [127,128] and optimized the design of structures [129][130][131][132][133][134]. ...

Machine Learning-Based Morphological and Mechanical Prediction of Kirigami-Inspired Active Composites
  • Citing Article
  • December 2023

International Journal of Mechanical Sciences

... In recent years, there has been growing interest in integrating dynamic variables and Machine Learning (ML) approaches to improve mortality prediction in the ICU [15]. The ML models offer advantages over classical statistical models in handling complex, high-dimensional data typical of ICU environments, providing more accurate and dynamic predictions by capturing nonlinear relationships and interactions that traditional models may miss [16]. ...

Machine learning vs. traditional regression analysis for fluid overload prediction in the ICU

... [9,10] Additionally, a pilot study of six machine learning methods also showed that incorporation of medication data and the medication regimen complexity-intensive care unit (MRC-ICU) score improved mortality prediction, and adding MRC-ICU to severity of illness improved traditional regression as well. [11] These examples offer credence to the concept that incorporating information on medication regimens is useful in predicting both shot-term and long-term outcomes for ICU patients. ...

Evaluation of medication regimen complexity as a predictor for mortality

... [4][5][6][7] Given the complexity and prolific nature of mediation use in the ICU, data driven strategies are increasingly being employed to parse meaningful patterns for fluid overload prediction. [8][9][10] While research is ongoing regarding identification of predictors for fluid overload, minimal research has evaluated the impact of medications as potential contributors. 11,12 These studies have shown that medication regimen complexity, as measured by the medication regimen complexity-ICU (MRC-ICU), was related to fluid overload risk, using both traditional regression and supervised machine learning approaches. ...

Machine learning vs. traditional regression analysis for fluid overload prediction in the ICU
  • Citing Preprint
  • June 2023

... From here, the derived medication list comprised of 30,550 discrete medication order entries for 991 ICU patients. [24] When a filter for the generic drug name, dose, and administration route was applied, a total of 1,868 unique medication products were identified. When only those medications incorporated in the MRC-ICU Scoring Tool were considered, a total of 889 discrete medication products remained for review and coding by the panel. ...

Cluster analysis driven by unsupervised latent feature learning of intensive care unit medications to identify novel pharmaco-phenotypes of critically ill patients

... Formula can be a good carrier of complex nutrients so as to ensure the brain and neurocognitive development of young children from 0 to 2 years old [21]. In preadolescents, a 9-month randomized, parallel-group, doubleblinded trial demonstrated that consuming foods containing added milk powder instead of snacks or meals enhanced inhibitory control and selective attention [22]. Therefore, it is crucial to study the nutritional requirements of infants and young children and to propose new early-life nutritional strategies. ...

The effect of 9-mo of formulated whole egg or milk powder food products as meal/snack replacements on executive function in preadolescents: A randomized placebo controlled trial
  • Citing Article
  • September 2022

American Journal of Clinical Nutrition

... Challapalli and Li [145] in their work proposed 20 different shaped symmetric optimal RVEs for obtaining higher buckling loads. With the help of a convolution neural network, Tian et al. [146] predicted the Poisson's ratio of different RVE-based auxetics. Using a feed-forward neural network, Fernández et al. [147] carried out the optimisation of beam lattice microstructures for anisotropic and hyperplastic materials. ...

Machine Learning-based Prediction and Inverse Design of 2D Metamaterial Structures with Tunable Deformation-Dependent Poisson’s Ratio
  • Citing Article
  • August 2022

Nanoscale