Tag: nevin manimala

Stat Med. 2026 Sep;45(20-22):e70720. doi: 10.1002/sim.70720.

ABSTRACT

While balancing covariates between groups is central for observational causal inference, selecting which features to balance remains a challenging problem. Kernel balancing is a promising approach that first estimates a kernel that captures similarity across units and then balances a (possibly low-dimensional) summary of that kernel, indirectly learning important features to balance. In this paper, we propose forest kernel balancing, which leverages the underappreciated fact that tree-based machine learning models, namely random forests and Bayesian additive regression trees (BART), implicitly estimate a kernel based on the co-occurrence of observations in the same terminal leaf node. Thus, even though the resulting kernel is solely a function of baseline features, the selected nonlinearities and other interactions are important for predicting the outcome-and therefore are important for addressing confounding. Through simulations and applied illustrations, we show that forest kernel balancing leads to meaningful computational and statistical improvement relative to standard kernel methods, which do not incorporate outcome information when learning features.

PMID:42658030 | DOI:10.1002/sim.70720

Bioinformatics. 2026 Aug 27:btag640. doi: 10.1093/bioinformatics/btag640. Online ahead of print.

ABSTRACT

BACKGROUND: Timeseries clinical transcriptomic datasets offer the opportunity to gain insights into the dynamics of disease mechanisms/treatment responses. However, their utility in uncovering temporal patterns is often limited by high noise levels and small sample sizes. Leveraging foundational gene embeddings and incorporating interaction information can help address these challenges, improve gene network analysis, and enable the detection of subtle changes that drive disease progression or drug response.

RESULTS: We finetuned gene embeddings from foundation models using healthy tissue gene expression data and used them in temporal GNNs to model gene expression of responder and non-responders to treatment in 4 disease datasets – ulcerative colitis, Crohn’s disease, Alopecia Areata, and psoriasis. Application of our method to these datasets confirmed known mechanisms associated with drug action, and also identified key differences between activated and repressed pathways for responders and non-responders, including B-Cell activation and mitochondria-related activity in ulcerative colitis patients.

CONCLUSION: Finetuning gene embeddings from foundation models provides a richer context to model gene expression data compared to using them in their naive state. Even with smaller sample sizes, results from GNN-based temporal models outperform traditional methods by detecting known mechanisms of response and unraveling role of genes and mechanisms not known to be associated with response and non-response.

CODE AVAILABILITY: Code and data are available in a public GitHub repository – https://github.com/Sanofi-Public/GNN-Timeseries. DOI: 10.5281/zenodo.20494035.

PMID:42658019 | DOI:10.1093/bioinformatics/btag640

Transl Vis Sci Technol. 2026 Aug 3;15(8):28. doi: 10.1167/tvst.15.8.28.

ABSTRACT

PURPOSE: To evaluate how reliably single Goldmann applanation tonometry (GAT) measurements reflect the preceding month’s intraocular pressure (IOP) profile in operated open-angle glaucoma.

METHODS: This longitudinal study included 24 patients with a suprachoroidal telemetric IOP sensor analyzing 156 in-hospital GAT measurements, 154 simultaneous telemetric readings, and 13,734 at-home measurements across follow-up visits. Telemetric data from the 30 days prior to each visit were stratified by time of day and matched to corresponding GAT measurements, which, in turn, were ranked relative to the interquartile range (IQR) of the preceding at-home measurements. Peak-and-trough analyses were performed using hourly aggregated data.

RESULTS: The GAT and telemetric IOP measurements showed good agreement (mean ΔIOP = -0.1 ± 3.0 mm Hg). However, 59.7% of the GAT-measured IOP values exceeded the IQR, and 14.3% fell below the IQR of the corresponding telemetric distributions; only 26.0% were within the IQR. Deviations from the telemetric median were 4.2 ± 2.4 mm Hg (above IQR), -2.3 ± 1.5 mm Hg (below IQR), and 0.5 ± 1.2 mm Hg (within IQR). Of the 99 statistically significant extrema (45 peaks and 54 troughs), most peaks occurred within 3 hours before GAT measurements. Troughs were observed both before and after GAT, but with larger nyctohemeral misalignment to GAT.

CONCLUSIONS: Single GAT measurements frequently failed to represent the underlying 30-day IOP profile and often missed clinically relevant peaks and troughs, despite good pointwise agreement with telemetric measurements.

TRANSLATIONAL RELEVANCE: High-frequency telemetric monitoring provides a more comprehensive assessment of IOP dynamics and may improve individualized glaucoma management compared with infrequent GAT snapshot measurements.

PMID:42658016 | DOI:10.1167/tvst.15.8.28

Front Oncol. 2026 Aug 12;16:1828599. doi: 10.3389/fonc.2026.1828599. eCollection 2026.

ABSTRACT

BACKGROUND: Neratinib, an irreversible pan-HER tyrosine kinase inhibitor, is a well-established extended adjuvant therapy for early-stage human epidermal growth factor receptor 2-positive (HER2+) breast cancer. However, its clinical application is often constrained by the high incidence of treatment-related diarrhea. This study sought to assess the dynamic changes in health-related quality of life (HRQoL) and the tolerability of neratinib, with a particular focus on dose escalation strategies and anti-diarrheal prophylaxis, within a real-world Chinese patient cohort.

METHODS: This single-center, prospective observational study included 31 female patients diagnosed with stage I-III HER2-positive breast cancer who initiated neratinib therapy following the completion of trastuzumab-based adjuvant treatment between October 2022 and January 2024. Patients were administered either non-dose-escalation group (240 mg/day) or dose-escalation group (120 mg/160 mg daily during Week 1, 160 mg/200 mg daily during Week 2, and 240 mg daily from Week 3 onward). Diarrhea management was conducted using prophylactic or on-demand loperamide. HRQoL was evaluated longitudinally through the EQ-5D-5L (health utility index and EQ visual analog scale [EQ VAS]) and Functional Assessment of Chronic Illness Therapy-Diarrhea (FACIT-D) questionnaires at predefined time points: baseline (D1), D30, D60, D90, D180, and D360. Diarrhea severity was graded according to Common Terminology Criteria for Adverse Events v5.0 (CTCAE v5.0) criteria. Statistical analyses encompassed descriptive statistics and Wilcoxon signed-rank tests.

RESULTS: The median age of the patients was 48 years (range: 37-65 years), with 58.1% and 41.9% classified as pathological stage II and III, respectively. A total of 64 diarrhea events were recorded (grade 1, 48.4%; grade 2, 50.0%; grade 3, 1.6%), with no grade 4 events or fatalities. Permanent discontinuation of treatment due to diarrhea occurred in 16.1% of patients, 23.5% in the dose-escalation group vs. 7.1% in the non-dose-escalation group. HRQoL temporarily declined during the early treatment phase, as evidenced by a reduction in the EQ-5D-5L health utility index from 0.96 ± 0.07 at baseline to 0.89 ± 0.15 at D30 (p=0.011) and a decrease in the EQ VAS score from 89.94 ± 9.92 to 84.16 ± 12.64 at D30 (p=0.018). However, HRQoL demonstrated progressive recovery from D90 onward, with scores surpassing baseline levels by D180 (EQ-5D-5L utility index: 0.96 ± 0.07; EQ VAS: 89.92 ± 7.74) and D360 (utility index: 0.97 ± 0.06; EQ VAS: 92.84 ± 7.19). The FACIT-D score showed an initial improvement (baseline: 58.81 ± 7.95 vs. D30: 66.90 ± 11.39, p=0.002), followed by stabilization at baseline levels from D180 onward. Notably, no disease recurrence or distant metastasis was observed in patients who completed the treatment.

CONCLUSIONS: Neratinib-associated HRQoL impairment is transient, with initial declines primarily due to diarrhea, which typically resolves within 3-6 months of treatment. With the implementation of appropriate supportive measures, such as loperamide prophylaxis and dose modifications, the majority of patients experience sustained HRQoL recovery, and even improvement, over a 12-month period.

PMID:42656231 | PMC:PMC13506311 | DOI:10.3389/fonc.2026.1828599

Front Psychol. 2026 Aug 12;17:1861330. doi: 10.3389/fpsyg.2026.1861330. eCollection 2026.

ABSTRACT

Physical education assessments that count toward university admission or course grades typically measure only a student’s fitness at a single time point, overlooking incremental progress and thus providing little incentive for those with lower baseline performance. This study evaluated the feasibility of a multidimensional scoring method that captures changes in standing long jump performance, an indicator of lower-limb explosive power and better reflects student learning. This study had two phases. Phase 1 combined a two-round Delphi process, AHP weighting, and standing long jump data from 211 classes (N = 4,109) to develop the multidimensional evaluation model. Phase 2 is a quasi-experimental study where 4 year 1-2 university classes was randomly selected and assigned to the intervention (IG) and control group (CG). Learning motivation was measured using the college student version of the Work Preference Inventory. Difference-in-difference model was used to analyze the association between motivation and standing long jump distance (post-pre) between IG and CG. Both IG and CG showed significant improvements in both standing long jump and motivation, with IG having significantly better outcomes. The time interaction between the groups confirmed the statistical significance. The OLS regression results showed both internal motivation (p < 0.001) and external motivation (p < 0.001) had significant positive relationship. Both boys’ (p < 0.001) and girls’ (p = 0.029) showed significant improvement. These findings suggest that a multidimensional, progress-sensitive scoring system can enhance students’ motivation and foster positive behavioral change in physical education.

PMID:42656224 | PMC:PMC13506395 | DOI:10.3389/fpsyg.2026.1861330

Front Dent Med. 2026 Aug 12;7:1847615. doi: 10.3389/fdmed.2026.1847615. eCollection 2026.

ABSTRACT

INTRODUCTION: Regenerative endodontic therapy aims to restore the pulp-dentin complex through biological approaches; however, current cell-free techniques often fail to achieve true dental pulp regeneration and may result in fibrous tissue formation or canal calcification. This study investigated dynamic hydrogels as advanced scaffolds for dental pulp stem cell (DPSC)-mediated regeneration, leveraging their viscoelastic properties to regulate cellular behavior and immune responses.

METHODS: Dynamic hydrogels were synthesized using host-guest chemistry (Gel-Mal/HA-Ada-CD-SH) to mimic aspects of the native extracellular matrix and were compared with non-dynamic Gel-Mal hydrogels. DPSCs and macrophages were encapsulated within the hydrogels to evaluate cellular viability, morphology, inflammatory marker expression, and gene expression profiles.

RESULTS: Dynamic hydrogels exhibited a distinct porous microarchitecture and significantly increased NF-κB and VEGF expression while reducing IL-1RA expression compared with non-dynamic hydrogels. Gene expression analysis demonstrated significantly increased TGFβ1, SMAD3, and MAPK expression, together with reduced CD80 expression, in macrophages cultured within dynamic hydrogels. Although iNOS expression was lower in the dynamic hydrogel group, the difference was not statistically significant.

DISCUSSION: These findings suggest that dynamic hydrogels do not simply suppress inflammatory signaling but may modulate inflammatory and tissue remodeling pathways while promoting angiogenesis-associated responses. Limitations of the study include the short-term in vitro design and evaluation of a limited panel of inflammatory and fibrosis-associated markers. Within these limitations, dynamic hydrogels demonstrated favorable cellular and immunomodulatory responses and may provide a biologically active microenvironment that supports cellular activities relevant to regenerative endodontic applications. Further in vivo studies are required to validate their translational potential.

PMID:42656218 | PMC:PMC13506308 | DOI:10.3389/fdmed.2026.1847615

Front Public Health. 2026 Aug 12;14:1878385. doi: 10.3389/fpubh.2026.1878385. eCollection 2026.

ABSTRACT

BACKGROUND: About 52 million residents are insured under the single-payer National Health Insurance (NHI) of South Korea. Twenty-six board-certified clinical specialties practice under this system, yet specialty-level matching between physician manpower and NHI-financed workload has not been reported.

METHODS: This study is an ecological cross-sectional analysis using specialty (n = 26) as the unit of observation. For 2023 and 2024, annual NHI reimbursement per specialty was divided by the mean specialist headcount to yield per-specialist reimbursement. This indicator captures facility-level resource use rather than physician effort in isolation. Specialties were placed on a reimbursement × application-rate plane and sorted into four quadrants. Inter-specialty reallocation was simulated at 100, 70, and 50% efficiency.

RESULTS: Per-specialist reimbursement was 52.5 million KRW (approximately USD 40,000) in radiology (n = 4,247) and 2,668.6 million KRW (approximately USD 2.04 million) in radiation oncology (n = 330). The resulting ratio was 50.8-fold. Coefficient of variation exceeded 90% in both years. Five specialties combined above-median reimbursement with sub-100% application rates; seven showed the reverse pattern. Utilization-volume triangulation confirmed that per-specialist reimbursement was strongly correlated with claims per specialist (Spearman ρ = 0.74) and patients per specialist (ρ = 0.66), supporting its interpretation as a workload-related indicator. The rank-order association between reimbursement and application rate was not statistically detectable (ρ = 0.180; 95% confidence interval, -0.24 to 0.54; p = 0.400), and the study was underpowered for this secondary test. A counterfactual calculation, presented as an upper-bound illustration rather than a policy projection, indicated that relocating 100 specialists from lower-reimbursement to higher-reimbursement quadrants at 70% efficiency corresponded to 68.8 billion KRW in annual NHI reimbursement.

CONCLUSION: Across 2 study years, the 50.8-fold gap between these two specialties did not narrow. This dispersion, which reflects NHI-financed resource consumption rather than physician effort alone, is the principal finding. A statistically detectable link between trainee competition and NHI reimbursement was not observed; the secondary correlation test was underpowered, and its null result is not evidence of no association. These findings provide a specialty-level reimbursement baseline for South Korea’s workforce allocation debate.

PMID:42656209 | PMC:PMC13506305 | DOI:10.3389/fpubh.2026.1878385

Front Med (Lausanne). 2026 Aug 12;13:1891922. doi: 10.3389/fmed.2026.1891922. eCollection 2026.

ABSTRACT

Basal cell carcinoma (BCC) arising in sun-protected genital sites is uncommon and lies outside the typical ultraviolet-driven model of disease. We retrospectively reviewed 41 patients with primary genital BCC treated at a tertiary referral center in China between January 2010 and December 2025, assessing clinicopathologic features, sex-stratified outcomes, p16 expression, and high-risk human papillomavirus (HPV) DNA status. The cohort included 29 males and 12 females. Female patients were older at diagnosis than male patients (74.8 ± 10.8 vs. 65.4 ± 12.0 years; p = 0.019). Nodular BCC was the predominant histologic subtype (29/41, 70.7%), and aggressive histologic components were present in 7 cases (17.1%). p16 immunoreactivity was detected in 4 of 37 successfully tested cases (10.8%), and high-risk HPV DNA was detected in 2 of 27 successfully tested cases (7.4%). Clinical presentation descriptors were retrospectively extracted from routine medical records and were incompletely documented. During a median follow-up of 51 months, five patients developed local recurrence and one male patient developed inguinal lymph node metastasis. In this exploratory cohort, no statistically significant sex-associated difference in recurrence-free survival was detected (log-rank p = 0.65), but the small number of events precluded inference of equivalent outcomes or exclusion of clinically meaningful sex-related differences. In this Asian cohort, genital BCC was usually nodular, only infrequently associated with detectable high-risk HPV in available tested specimens, and generally favorable after surgery. Because genital lesions may be concealed and clinically non-specific, biopsy should be considered for persistent or suspicious genital lesions, with ongoing postoperative surveillance.

PMID:42656207 | PMC:PMC13506306 | DOI:10.3389/fmed.2026.1891922

Cureus. 2026 Jul 26;18(7):e113419. doi: 10.7759/cureus.113419. eCollection 2026 Jul.

ABSTRACT

Cardiac telemetry monitoring is an essential tool for assessing cardiac function in hospitalized patients; however, its overuse in low-risk populations can increase healthcare costs, alarm fatigue, and unnecessary resource utilization. This study aimed to assess the use of cardiac telemetry at a tertiary care center in Puerto Rico. Using a retrospective cross-sectional review of medical records, we evaluated whether telemetry was used appropriately according to established clinical criteria and identified patterns of overuse. The findings provide a foundation for developing an institutional telemetry protocol to optimize telemetry utilization, enhance patient safety, and improve cost-effectiveness. A total of 822 hospitalized patients receiving cardiac telemetry were included. Variables such as history of cardiac disease, comorbidities, length of stay, indication for telemetry, and duration of monitoring were analyzed. Our findings revealed that 476 (57.9%) patients received telemetry for non-evidence-based reasons, while 346 (42.1%) had clear clinical indications. Patients with non-evidence-based telemetry had longer monitoring durations, averaging 5.68 ± 5.90 days compared to 2.98 ± 2.94 days for those with evidence-based indications. They also spent slightly less time in the emergency department, with a mean of 7.18 ± 4.73 hours, compared to 7.98 ± 6.44 hours for the evidence-based group. Although not statistically significant (p = 0.137), the mean length of stay was higher among non-evidence-based cases (6.73 ± 8.25 days) than among evidence-based cases (4.46 ± 7.38 days). These differences contributed to an estimated cost of $4,711 per patient for non-evidence-based telemetry compared to $3,122 per patient for evidence-based monitoring. This study suggests that implementing standardized, evidence-based guidelines for telemetry initiation and discontinuation could reduce overuse, lower costs, decrease alarm fatigue, and maintain patient safety while showcasing the potential benefits of a hospital-wide protocol to optimize telemetry practices.

PMID:42656202 | PMC:PMC13506248 | DOI:10.7759/cureus.113419

Philos Trans A Math Phys Eng Sci. 2026 Aug 27;384(2327):20250073. doi: 10.1098/rsta.2025.0073.

ABSTRACT

Conformal prediction (CP) has emerged as a cutting-edge methodology in statistics and machine learning, providing prediction intervals with finite-sample frequentist coverage guarantees. Yet, its interplay with Bayesian statistics-often criticized for lacking frequentist guarantees-remains underexplored. Recent work has suggested that CP can ‘calibrate’ Bayesian prediction regions, thereby imparting frequentist validity and motivating deeper investigation into frequentist-Bayesian hybrids. On the other side, Bayesian procedures have the potential to enhance CP with more informative intervals, towards nearly optimal solutions under a decision-theoretic framework. Thus, the two paradigms can be jointly used for a principled balance between validity and efficiency. This work provides a unified treatment of this emerging interface with open directions. After surveying existing ideas, we consolidate the literature with a Bayesian version of split CP and present a simple analysis of the binomial model, investigating priors’ role, efficiency and computational complexity. This article is part of the theme issue ‘Advancing uncertainty quantification in AI systems’.

PMID:42656165 | DOI:10.1098/rsta.2025.0073