Tag: statistics

J Appl Clin Med Phys. 2022 Mar 9:e13583. doi: 10.1002/acm2.13583. Online ahead of print.

ABSTRACT

PURPOSE: To develop a 3D-Unet dose prediction model to predict the three-dimensional dose distribution of volumetric modulated arc therapy (VMAT) for cervical cancer and test the dose prediction performance of the model in endometrial cancer to explore the feasibility of model generalization.

METHODS: One hundred and seventeen cases of cervical cancer and 20 cases of endometrial cancer treated with VMAT were used for the model training, validation, and test. The prescribed dose was 50.4 Gy in 28 fractions. Eight independent channels of contoured structures were input to the model, and the dose distribution was used as the output of the model. The 3D-Unet prediction model was trained and validated on the training set (n = 86) and validation set (n = 11), respectively. Then the model was tested on the test set (n = 20) of cervical cancer and endometrial cancer, respectively. The results between clinical dose distribution and predicted dose distribution were compared in the following aspects: (a) the mean absolute error (MAE) within the body, (b) the Dice similarity coefficients (DSCs) under different isodose volumes, (c) the dosimetric indexes including the mean dose (D_mean ), the received dose of 2 cm³ (D_2cc) , the percentage volume of receiving 40 Gy dose of organs-at-risk (V₄₀ ), planning target volume (PTV) D_98% , and homogeneity index (HI), (d) dose-volume histograms (DVHs).

RESULTS: The model can accurately predict the dose distribution of the VMAT plan for cervical cancer and endometrial cancer. The overall average MAE and maximum MAE for cervical cancer were 2.43 ± 3.17% and 3.16 ± 4.01% of the prescribed dose, respectively, and for endometrial cancer were 2.70 ± 3.54% and 3.85 ± 3.11%. The average DSCs under different isodose volumes is above 0.9. The predicted dosimetric indexes and DVHs are equivalent to the clinical dose for both cervical cancer and endometrial cancer, and there is no statistically significant difference.

CONCLUSION: A 3D-Unet dose prediction model was developed for VMAT of cervical cancer, which can predict the dose distribution accurately for cervical cancer. The model can also be generalized for endometrial cancer with good performance.

PMID:35262273 | DOI:10.1002/acm2.13583

J Appl Clin Med Phys. 2022 Mar 9:e13582. doi: 10.1002/acm2.13582. Online ahead of print.

ABSTRACT

This study presents a methodology to develop an X-ray technique chart for portable chest and abdomen imaging which utilizes patient data available in the modality worklist (MWL) to reliably achieve a predetermined exposure index (EI) at the detector for any patient size. The method assumes a correlation between the patients’ tissue equivalent thickness and the square root of the ratio of the patient’s weight to height. To assess variability in detector exposures, the EI statistics for 75 chest examinations and 99 abdominal portable X-ray images acquired with the new technique chart were compared to those from a single portable unit (chest: 3877 images; abdomen: 200 images) using a conventional technique chart with three patient sizes, and to a stationary radiography room utilizing automatic exposure control (AEC) (chest: 360 images; abdomen: 112 images). The results showed that when using the new technique chart on a group of portable units, the variability in EI was significantly reduced (p < 0.01) for both AP chest and AP abdomen images compared to the single portable using a standard technique chart with three patient sizes. The variability in EI for the images acquired with the new chart was comparable to the stationary X-ray room with an AEC system (p > 0.05). This method could be used to streamline the entire imaging chain by automatically selecting an X-ray technique based on patient demographic information contained in the MWL to provide higher quality examinations to clinicians by eliminating outliers. In addition, patient height and weight can be used to estimate the patients’ tissue equivalent thickness.

PMID:35262265 | DOI:10.1002/acm2.13582

J Neuroimaging. 2022 Mar 9. doi: 10.1111/jon.12987. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: To compare quantitative susceptibility mapping (QSM) and high-pass-filtered (HPF) phase imaging for (1) identifying chronic active rim lesions with more myelin damage and (2) distinguishing patients with increased clinical disability in multiple sclerosis.

METHODS: Eighty patients were scanned with QSM for paramagnetic rim detection and Fast Acquisition with Spiral Trajectory and T2prep for myelin water fraction (MWF). Chronic lesions were classified based on the presence/absence of rim on HPF and QSM images. A lesion-level linear mixed-effects model with MWF as the outcome was used to compare myelin damage among the lesion groups. A multiple patient-level linear regression model was fit to establish the association between Expanded Disease Status Scale (EDSS) and the log of the number of rim lesions.

RESULTS: Of 2062 lesions, 188 (9.1%) were HPF rim+/QSM rim+, 203 (9.8%) were HPF rim+/QSM rim-, and the remainder had no rim. In the linear mixed-effects model, HPF rim+/QSM rim+ lesions had significantly lower MWF than both HPF rim+/QSM rim- (p < .001) and HPF rim-/QSM rim- (p < .001) lesions, while the MWF difference between HPF rim+/QSM rim- and HPF rim-/QSM rim- lesions was not statistically significant (p = .130). Holding all other factors constant, the log number of QSM rim+ lesion was associated with EDSS increase (p = .044). The association between the log number of HPF rim+ lesions and EDSS was not statistically significant (p = .206).

CONCLUSIONS: QSM identifies paramagnetic rim lesions that on average have more myelin damage and stronger association with clinical disability than those detected by phase imaging.

PMID:35262241 | DOI:10.1111/jon.12987

Pharm Stat. 2022 Mar 9. doi: 10.1002/pst.2202. Online ahead of print.

ABSTRACT

Databases derived from electronic health records (EHRs) are commonly subject to left truncation, a type of selection bias that occurs when patients need to survive long enough to satisfy certain entry criteria. Standard methods to adjust for left truncation bias rely on an assumption of marginal independence between entry and survival times, which may not always be satisfied in practice. In this work, we examine how a weaker assumption of conditional independence can result in unbiased estimation of common statistical parameters. In particular, we show the estimability of conditional parameters in a truncated dataset, and of marginal parameters that leverage reference data containing non-truncated data on confounders. The latter is complementary to observational causal inference methodology applied to real-world external comparators, which is a common use case for real-world databases. We implement our proposed methods in simulation studies, demonstrating unbiased estimation and valid statistical inference. We also illustrate estimation of a survival distribution under conditionally independent left truncation in a real-world clinico-genomic database.

PMID:35262259 | DOI:10.1002/pst.2202

Mov Disord. 2022 Mar 9. doi: 10.1002/mds.28987. Online ahead of print.

ABSTRACT

BACKGROUND: To date, variants in the GBA gene represent the most frequent large-effect genetic factor associated with Parkinson’s disease (PD). However, the reason why individuals with the same GBA variant may or may not develop neurodegeneration and PD is still unclear.

OBJECTIVES: Therefore, we evaluated the contribution of rare variants in genes responsible for lysosomal storage disorders (LSDs) to GBA-PD risk, comparing the burden of deleterious variants in LSD genes in PD patients versus asymptomatic subjects, all carriers of deleterious variants in GBA.

METHODS: We used a custom next-generation sequencing panel, including 50 LSD genes, to screen 305 patients and 207 controls (discovery cohort). Replication and meta-analysis were performed in two replication cohorts of GBA-variant carriers, of 250 patients and 287 controls, for whom exome or genome data were available.

RESULTS: Statistical analysis in the discovery cohort revealed a significantly increased burden of deleterious variants in LSD genes in patients (P = 0.0029). Moreover, our analyses evidenced that the two strongest modifiers of GBA penetrance are a second variation in GBA (5.6% vs. 1.4%, P = 0.023) and variants in genes causing mucopolysaccharidoses (6.9% vs. 1%, P = 0.0020). These results were confirmed in the meta-analysis, where we observed pooled odds ratios of 1.42 (95% confidence interval [CI] = 1.10-1.83, P = 0.0063), 4.36 (95% CI = 2.02-9.45, P = 0.00019), and 1.83 (95% CI = 1.04-3.22, P = 0.038) for variants in LSD genes, GBA, and mucopolysaccharidosis genes, respectively.

CONCLUSION: The identification of genetic lesions in lysosomal genes increasing PD risk may have important implications in terms of patient stratification for future therapeutic trials. © 2022 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson Movement Disorder Society.

PMID:35262230 | DOI:10.1002/mds.28987

Stat Med. 2022 Mar 9. doi: 10.1002/sim.9370. Online ahead of print.

ABSTRACT

We propose a new approach to test associations between binary trees and covariates. In this approach, binary-tree structured data are treated as sample paths of binary fission Markov branching processes (bMBP). We propose a generalized linear regression model and developed inference procedures for association testing, including variable selection and estimation of covariate effects. Simulation studies show that these procedures are able to accurately identify covariates that are associated with the binary tree structure by impacting the rate parameter of the bMBP. The problem of association testing on binary trees is motivated by modeling hierarchical clustering dendrograms of pixel intensities in biomedical images. By using semi-synthetic data generated from a real brain-tumor image, our simulation studies show that the bMBP model is able to capture the characteristics of dendrogram trees in brain-tumor images. Our final analysis of the glioblastoma multiforme brain-tumor data from The Cancer Imaging Archive identified multiple clinical and genetic variables that are potentially associated with brain-tumor heterogeneity.

PMID:35262202 | DOI:10.1002/sim.9370

Pacing Clin Electrophysiol. 2022 Mar 9. doi: 10.1111/pace.14478. Online ahead of print.

ABSTRACT

BACKGROUND: Cardiac contractility modulation (CCM), being reserved for patients with symptomatic chronic heart failure (HF) and narrow QRS complex under guideline directed medical therapy, can recover initially reduced left ventricular ejection fraction (LVEF); however, the influence of pre-implantation LVEF on long-term outcomes is not fully understood. This study aimed to compare the effects of lower and higher pre-implantation LVEF on long-term outcomes in CCM-therapy.

METHODS: One-hundred seventy-two patients from our single-centre registry were retrospectively included (2002 – 2019). Follow-up data were collected up to five years after implantation. Patients were divided into Group 1 (baseline LVEF≤ 30%) and Group 2 (≥ 31%). Both groups were compared based on differences in survival, echocardiographic- and clinical parameters including LVEF, tricuspid annular plane systolic excursion (TAPSE), NYHA class or Minnesota living with heart failure questionnaire-score (MLWHFQ).

RESULTS: 11 % of the patients did have a LVEF ≥ 31%. Mean LVEF±SD for both groups were 21.98±5.4 vs. 35.2±3.7%, respectively. MLWHFQ (47±21.2 vs. 42±21.4) and mean peak oxygen consumption (VO2, 13.6±4.1 vs. 12.7±2.8 ml/kg/min) were comparable between both groups. LVEF-grouping did not influence survival. Lower baseline LVEF resulted in significantly better recovery of echocardiographic parameters such as LVEF and TAPSE. Irrespective from baseline LVEF, both groups showed nearly comparable improvements for clinical parameters like NYHA-class and MLWHFQ.

CONCLUSION: Long-term biventricular systolic recovery potential in CCM-therapy might be better for pre-implantation LVEF values ≤ 30%, whereas clinical parameters such as NYHA-class can improve irrespective from baseline LVEF. This article is protected by copyright. All rights reserved.

PMID:35262210 | DOI:10.1111/pace.14478

Int J Gynaecol Obstet. 2022 Mar 8. doi: 10.1002/ijgo.14169. Online ahead of print.

ABSTRACT

OBJECTIVE: To explore whether patient characteristics were associated with gender expression, and to further determine impact of gender expression on selection of hysterectomy or uterine-preservation in pelvic organ prolapse (POP) surgery.

METHODS: Within a prospective cohort, a self-reported gender expression tool classified patients as expressing gender polar (i.e., reporting only feminine traits) or non-polar gender scores (i.e., reporting feminine and masculine traits). Multivariate modelling explored associations of gender expression with traditional socio-demographic variables, and with selection of hysterectomy or uterine-preserving surgery. Descriptive statistics of socio-demographic variables were reported by frequency, proportion and mean (SD).

RESULTS: 177 participants completed the gender score questionnaire. Overall the sample had a more feminine gender expression with the majority of respondents classified as gender polar (67.23%, n=119). Participants with non-polar gender scores were 2.53 times (95% 1.05 – 6.09) more likely to choose uterine preservation vs hysterectomy-based surgery. Gender polarity was weakly associated with age, but no other sociodemographic variables.

CONCLUSION: Gender expression is not tightly associated with socio-demographic variables, and is thus a novel measurement in gynecologic research. Gender polarity appears to be associated with choice to undergo hysterectomy. Further research is required to understand this relationship and implications in clinical outcomes.

PMID:35262193 | DOI:10.1002/ijgo.14169

Dalton Trans. 2022 Mar 9. doi: 10.1039/d2dt00295g. Online ahead of print.

ABSTRACT

A Ga-enriched ternary alkali-metal sulfide Na₄Ga₈S₁₄ has been synthesized by a high temperature solid-state reaction. It crystallizes in the centrosymmetric Pbca (no. 61) space group with cell parameters a = 13.5260(4) Å, b = 11.4979(3) Å, c = 29.9592(9) Å, and Z = 8, and exhibits a three-dimensional (3D) network structure constructed from unique [Ga₁₂S₄₂] 12-membered rings, one-dimensional _∞[Ga₄S₁₁] chains, individual [GaS₄] units and Na⁺ ions. The experimental band gap of Na₄Ga₈S₁₄ was measured as ∼3.57 eV. Theoretical calculations indicate that the title compound is a direct band gap compound and the band gap is mainly determined by [GaS₄] units. Meanwhile, statistical analysis shows that the atomic ratio N (N = A^IA^II/Ga, where A^I = alkali-metal, A^II = alkaline earth-metal) can be used to regulate the connection of [GaS₄] units from zero-dimensional (0D) isolated groups, one-dimensional (1D) chains, and two-dimensional (2D) layers to 3D frameworks in Ga-containing alkali- and/or alkaline earth-metal chalcogenides. The results enrich the diversity of alkali-metal sulfides and give an insight into the structural regulation of alkali- and/or alkaline earth-metal chalcogenides.

PMID:35262155 | DOI:10.1039/d2dt00295g

Dev Med Child Neurol. 2022 Mar 9. doi: 10.1111/dmcn.15199. Online ahead of print.

ABSTRACT

AIM: To describe cross-sectional and longitudinal variation in neurorehabilitation content provided to young people after severe paediatric acquired brain injury (pABI) and to relate this to observed functional recovery.

METHOD: This was an observational study in a cohort of admissions to a residential neurorehabilitation centre. Recovery was described using the Pediatric Evaluation of Disability – Computer Adaptive Testing instrument. Rehabilitation content was measured using the recently described Paediatric Rehabilitation Ingredients Measure (PRISM) and examined using multidimensional scaling.

RESULTS: The PRISM reveals wide variation in rehabilitation content between and during admissions primarily reflecting proportions of child active practice, child emotional support, and other management of body structure and function. Rehabilitation content is predicted by pre-admission recovery, suggesting therapist decisions in designing rehabilitation programmes are shaped by their initial expectations of recovery. However, significant correlations persist between plausibly-related aspects of delivered therapy and observed post-admission recovery after adjusting for such effects.

INTERPRETATION: The PRISM approach to the analysis of rehabilitation content shows promise in that it demonstrates significant correlations between plausibly-related aspects of delivered therapy and observed recovery that have been hard to identify with other approaches. However, rigorous, causal analysis will be required to truly understand the contributions of rehabilitation to recovery after pABI.

PMID:35262182 | DOI:10.1111/dmcn.15199