Citations

Background: Heart failure with preserved ejection fraction (HFpEF) accounts for approximately half of all heart failure hospitalizations; yet, tools for individualized 1-year mortality risk stratification remain limited. We conducted a retrospective study to develop and validate a machine learning model using genetic programming and automated model search to predict 1-year all-cause mortality in HFpEF patients admitted for decompensated HF. Methods: Electronic medical records from Mayo Clinic (2010-2020) were queried to identify 7,840 adult patients hospitalized with HF exacerbation and a left ventricular ejection fraction>50% on echocardiograms performed between six months pre-admission and seven days post-discharge. Demographic, clinical, laboratory, medication, comorbidity, and imaging data available at the time of index hospitalization were extracted. The cohort was split into 80% training (n=6,272) and 20% held-out test (n=1,568) sets. After standard preprocessing, we performed an evolutionary AutoML search. The final pipeline applied a RobustScaler, two sequential feature-union blocks, and an XGBoost classifier. Model performance on the test set was assessed by area under the receiver operating characteristic curve (AUC), Brier score, accuracy, precision, recall, and F1 score, each with bootstrap-derived 95% confidence intervals (CIs). For interpretability, SHAP (SHapley Additive exPlanations) values were computed using the full scaled training set as background. Results: On the held-out test set, the XGBoost-based model achieved an AUC of 0.7595 (95% CI: 0.7307-0.7870) and a Brier score of 0.1764 (95% CI: 0.1652-0.1890). Accuracy was 0.7406 (95% CI: 0.7143-0.7645), precision 0.6349 (95% CI: 0.5759-0.6942), recall 0.4135 (95% CI: 0.3636-0.4609), and F1 score 0.5008 (95% CI: 0.4529-0.5444). The global SHAP summary plot (Figure 1) ranked age, serum albumin, blood urea nitrogen (BUN), presence of renal failure, and NT-proBNP as the top 5 predictors. Conclusions: In this large HFpEF cohort, an AutoML-derived XGBoost pipeline achieved robust discrimination (AUC≈0.76) and calibration (Brier≈0.18) for 1-year mortality prediction. SHAP-based explanations highlighted the importance of age, nutritional status (albumin), renal biomarkers (BUN, renal failure), and natriuretic peptides (NT-proBNP) as principal drivers of risk. These interpretable machine learning findings may guide personalized risk stratification and identify therapeutic targets in HFpEF.

Introduction: Cardiovascular risks following SARS-CoV-2 infection are incompletely defined. We evaluated evolving cardiovascular outcomes across COVID-19 variant eras using a large community-based cohort to inform risk stratification and clinical care. Methods: We conducted a retrospective cohort study of 162,471 adults with COVID-19 between March 2020 and December 2023 using the Rochester Epidemiology Project. Patients were stratified by variant era (Pre-Delta, Delta, and Omicron). Cardiovascular outcomes included major adverse cardiovascular events (MACE: myocardial infarction, stroke, death), ischemic and inflammatory heart disease, thrombotic events, and dysrhythmias. Diagnoses were identified using ICD-10 codes and validated in a subset. Cumulative incidence was estimated using Kaplan-Meier and competing risk models. Cox proportional hazards regression was used to assess differences by era and age. Results: MACE occurred in 4,922 patients, primarily in those ≥80 (2-year incidence: 17%). Event rates peaked within 30 days post-infection, with marked variation by variant. In patients ≥80, 30-day MACE risk was highest in the Pre-Delta (HR 3.85; 95% CI, 3.14–4.71) and Delta (HR 2.72; 95% CI, 2.13–3.47) eras versus Omicron. After 30 days, event rates fell and era differences narrowed. Thrombotic events (n=2,090; 1.4%) decreased across eras. Ischemic and inflammatory heart disease rates were stable. Myocarditis/pericarditis were rare (0.1%) without age trends. Dysrhythmias affected 6,765 patients (2-year incidence: 4.9%) and rose with age (2.0% in <40 vs. 19.9% in ≥80). Within 30 days, risk was higher in the Pre-Delta and Delta eras for age ≥40, especially ≥80 (Delta HR 3.08; 95% CI, 2.08–4.56) and 60–79 (HR 2.23; 95% CI, 1.75–2.85). Among those 40–59, Delta posed the highest risk (HR 2.74; 95% CI, 1.94–3.86 vs. Pre-Delta HR 1.77; 95% CI, 1.26–2.49). No differences were observed in <40. Beyond 30 days, risk converged across variants. Conclusion: Cardiovascular risk after SARS-CoV-2 infection was concentrated in the first 30 days, with attenuation thereafter. While MACE and thrombotic event rates declined over time, early dysrhythmia risk varied by variant and age, especially in older adults. These findings define a time-sensitive, variant-specific risk window and support short-term, risk-stratified post-COVID monitoring. This work informs targeted follow-up protocols by infection era and age, with implications for quality improvement, resource use, and equitable care.

Abstract The all‐inside anterior cruciate ligament reconstruction technique has gained popularity because of its ability to preserve bone stock, accommodate shorter graft lengths, and reduce postoperative pain. Traditionally, this technique uses metallic cortical suspensory buttons for graft fixation, which are reliable but may cause soft‐tissue irritation and hardware prominence. A recent modification to traditional all‐inside suspensory fixation uses all‐suture cortical buttons, eliminating the need for metallic or screw‐based implants while maintaining secure fixation. This Technical Note describes an all‐inside anterior cruciate ligament reconstruction technique that uses an all–soft tissue quadriceps tendon autograft, secured with only suture‐based buttons (no metal, no plastic) on both the femoral and tibial sides. Technique Video video See video under supplementary data.

Importance Adjuvant radiotherapy can improve locoregional control and survival in patients with olfactory neuroblastoma (ONB), particularly with advanced-stage and histologic-grade disease. Standard radiotherapy treatment is with intensity-modulated radiotherapy (IMRT). Proton beam radiotherapy (PBRT) provides theoretical advantages in greater sparing of dose to uninvolved organs at risk. Objective To investigate if there are differences in the effectiveness and radiation treatment–related adverse events (RTAEs) between adjuvant IMRT and PBRT for patients with ONB. Design, Setting, and Participants This propensity score–matched cohort study included patients with ONB treated between February 2005 and April 2021 with either IMRT or PBRT at 9 academic tertiary care centers in North America. Patients were matched 1:2 based on age, modified Kadish stage, and Hyams grade. Data were analyzed from July 2024 to January 2025. Exposure Adjuvant IMRT or adjuvant PBRT. Main Outcomes and Measures Local recurrence-free survival (RFS), any RFS, and overall survival (OS). RTAEs, ie, grade 2 events or higher based on Common Terminology Criteria for Adverse Events, were recorded for both modalities. Results Of 54 included patients, 27 (50%) were female, and the mean (SD) age was 46.2 (15.4) years. A total of 18 were treated with PBRT and 36 were treated with IMRT. Most patients had modified Kadish stage C disease (33 of 54 [61%]), and 24 patients (44%) had Hyams grade III or IV disease. The RTAE rate was 20% (8 of 40); IMRT had a rate of 21% (6 of 29), and PBRT had a rate of 18% (2 of 11). The difference in the point estimates for 10-year RFS showed a potential clinical benefit favoring IMRT, although the wide confidence interval indicates uncertainty (10-year RFS: IMRT, 63.3%; 95% CI, 44.6-89.8; PBRT, 37.8%; 95% CI, 14.2-100; difference, 25.5 percentage points; 95% CI, −17.6 to 68.6). There were no clinically meaningful differences in 10-year local RFS (IMRT, 75.6%; 95% CI, 59.8-95.4; PBRT, 72.7%; 95% CI, 45.2-100; difference, 2.9 percentage points; 95% CI, −35.9 to 41.7) or 10-year OS (IMRT, 61.8%; 95% CI, 42.8-89.1; PBRT, 57.1%; 95% CI, 24.3-100; difference, 4.7 percentage points; 95% CI, −49.2 to 58.6), although wide confidence intervals indicate considerable uncertainty. Conclusions and Relevance Due to the imprecision of estimates, no definitive conclusions can be made regarding the comparative effectiveness of IMRT vs PBRT for patients with ONB. These preliminary data may inform the design of appropriately powered prospective studies evaluating the efficacy of PBRT vs IMRT in this population.

Abstract INTRODUCTION M0 images were missing in Siemens ASL data in Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, prohibiting cerebral blood flow (CBF) quantification. METHODS A conditional latent diffusion model was trained and evaluated on in‐house datasets, then applied to the Siemens data in ADNI‐3. Regional CBF differences by Alzheimer's disease (AD) stages, their accuracy for AD classification, and CBF trajectory slopes were compared between generated data (Siemens) and acquired data (General Electric). RESULTS The diffusion model generated M0 images with high fidelity (SSIM = 0.918 ± 0.023, PSNR = 31.361 ± 2.537) and minimal CBF bias (mean difference is 0.21 ± 1.58 mL/100 g/min). Both generated and acquired CBF showed similar spatial patterns and decreasing trends with AD progression in specific AD‐related regions. Generated CBF also improved accuracy in classifying AD stages compared to qualitative perfusion images. CONCLUSION This study shows the potential of diffusion models for imputing missing modalities in large‐scale studies exploring the use of ASL as a biomarker of AD. Highlights Using latent diffusion model, we can generate M0 image from control image in arterial spin labeling (ASL) with high fidelity. The generated M0 can be used for cerebral blood flow (CBF) quantification in Alzheimer's Disease Neuroimaging Initiative dataset. The performance of classification between Alzheimer's disease (AD) patients and cognitive normal people is better when using generated CBF maps than using non‐quantitative perfusion images. ASL CBF decreases with AD progression in key AD‐related brain regions.

OBJECTIVES The Pediatric Emergency Care Applied Research Network (PECARN) cervical spine injury (CSI) clinical prediction rule (CPR) was developed to reduce unnecessary imaging in pediatric trauma patients and ensure accurate diagnosis. This study aimed to identify barriers and facilitators to future implementation of this rule in diverse emergency department (ED) settings. METHODS We conducted 40 interviews with 48 clinicians, including ED physicians, surgeons, nurses, and advanced practice providers at 21 health care facilities across the United States. Semistructured interviews explored determinants impacting implementation, guided by the Tailored Implementation for Chronic Diseases (TICD) framework. We conducted a deductive thematic analysis using the TICD domains as a priori codes. Three researchers independently coded the transcripts and identified key themes aligned with the TICD framework. RESULTS Barriers included limited guideline accessibility, workflow disruptions, and inconsistent integration with electronic health records. Clinicians’ concerns regarding the applicability and adherence to guidelines were also significant. Facilitators identified included institutional support, stakeholder engagement, and targeted education that addressed knowledge gaps. Implementation strategies varied by context, with suggestions for electronic health record-based decision tools in resource-equipped facilities and external aids, such as pocket cards, in lower-resource settings. Addressing legal concerns, clinician resistance, and workflow alignment emerged as critical priorities. CONCLUSION Implementing the PECARN CSI CPRs requires addressing contextual barriers such as workflow challenges, clinician adherence, and technology integration. Context-sensitive strategies are essential for successful adoption and improved decision-making in pediatric trauma care. Future research should evaluate the impact of these strategies on clinician behavior and patient outcomes.

Abstract INTRODUCTION Preclinical Alzheimer's disease (AD) can be described relative to biomarker positivity onset time. METHODS We estimated time from amyloid positivity (A+) using sampled iterative local approximation (SILA) in a longitudinal autosomal dominant AD (ADAD) sample ( N = 379) with amyloid positron emission tomography. We compared (1) predicted age at A+ to imputed age, (2) estimated age at A+ to estimated age at symptom onset, and (3) variance in cognitive performance explained. RESULTS Mean error between imputed and SILA‐estimated age at A+ ( N = 26) was 1.15 years. Age at A+ explained 39% of estimated years to symptom onset (EYO) variance. Time from A+ explained 19% of cognitive composite variance and 14% of Clinical Dementia Rating Sum of Boxes CDR‐SB variance; EYO explained 43% and 57%, respectively. DISCUSSION SILA estimates A+ age in ADAD with reasonably good accuracy. SILA‐estimated time from A+ describes the start of pathology, but the time from A+ onset to symptoms is variable in ADAD and better described by EYO. Highlights Amyloid chronicity predicts a 14‐year preclinical AD phase in ADAD. SILA accurately estimates age at A+ (MAE < 2 years). EYO outperforms chronicity in predicting symptom onset. APP mutation carriers show atypical amyloid accumulation. Chronicity models help reveal AD heterogeneity in preclinical stages.

Background The literature on dyslipidemia and cognitive trajectories among cognitively unimpaired (CU) persons remains inconclusive. Objective To investigate the association between baseline dyslipidemia and change in global and domain specific (i.e., memory, language, attention/executive function, and visuospatial skills) cognition in a population-based setting and whether the association differs by sex, age, or APOE ɛ4 carrier status. Methods We conducted a longitudinal study derived from the Mayo Clinic Study of Aging, involving 4236 CU persons aged ≥ 50 years. We ran linear mixed-effect models to examine baseline dyslipidemia in predicting longitudinal cognitive global and domain-specific z-scores adjusted for age, sex, education, medical comorbidity, repeated cognitive testing, and APOE ɛ4. We examined interactions among dyslipidemia, years since baseline, and separately by sex, age, and APOE ɛ4 carrier status. Results Over a median follow-up period of 6.4 years, CU individuals with dyslipidemia showed faster decline in z-scores of all domains, i.e., memory, language, attention/executive function, visuospatial and global cognition relative to CU individuals without dyslipidemia. Three-way interactions showed that dyslipidemia's effect on decline in z-scores of global cognition, attention and visuospatial skills was less pronounced in males than females. Higher age increased dyslipidemia's effect on decline in z-scores of attention but not global cognition or any other domain. APOE ɛ4 carrier status did not modify the effect of dyslipidemia on cognitive z-scores. Conclusions Dyslipidemia was associated with faster global and domain-specific cognitive decline over time in older community-dwelling individuals who were cognitively unimpaired at baseline. The effect of dyslipidemia on cognitive trajectories may be sex-influenced.

Abstract Background Although recent guidance recommends early surgical debridement for native joint septic arthritis (NJSA), supporting data, particularly on long-term outcomes, remains scarce. Methods We conducted a retrospective multicenter cohort study of adults (≥18 years) with NJSA who underwent surgery across Mayo Clinic campuses between 2012 and 2021. Clinical outcomes at 1 year were assessed using a 9-level Desirability of Outcome Ranking (DOOR) scale, incorporating survival, treatment failure (relapse, reinfection, readmission, or significant surgical events), and joint recovery. Time to surgery from hospital admission was analyzed both as a continuous variable and as categories: &lt; 1, 1–2, or ≥3 days. Results Among 268 patients, 30% underwent surgery &lt;1 day from admission, 47% in 1–2 days, and 24% in ≥3 days. At 1 year, 57% achieved full recovery without unfavorable events (DOOR score 1), while treatment failure occurred in 34%. In unadjusted analyses, longer surgical delay was significantly associated with higher (worse) DOOR scores (per IQR increase [from 0 to 2 days], OR: 1.5; 95% CI: 1.0–2.1; P = .026), increased 1-year mortality (HR: 1.7; 95% CI: 1.1–2.6; P = .019) and treatment failure (HR: 1.5; 95% CI: 1.1–2.0; P = .007). Even after adjusting for age and Charlson comorbidity index, the association between surgical delay and treatment failure remained significant (HR: 1.5; 95% CI, 1.1–2.0; P = .016). Conclusions The finding that delayed surgical intervention is associated with an increased risk of treatment failure reinforces current expert recommendations for timely surgical management in NJSA.

Abstract Increased demand to make medical linear accelerator acceptance testing procedures (ATP) more efficient exists while the technology being evaluated for clinical use has become increasingly complex. The most recent American Association of Physicists in Medicine (AAPM) task group giving guidelines to medical physicists on ATP was Task Group report 45, published in 1994. Several new technologies, enhanced machine performance, and possible future changes to ATP since that time require updated guidance from the AAPM. The goal of this task group is to empower the medical physicist to be meaningfully involved with the design and execution of comprehensive and effective linear accelerator ATP beginning with the purchasing contract negotiation to the contractual handover of the linear accelerator to the department for clinical use. Additionally, we discussed aspects related to beam matching, ATP at the time of replacing major linear accelerator's components as well as what we perceive as the future of the ATP process given the trends of standardizing linear accelerators.