Search NASA⌕ Search

SEARCH · Search NASA

Results for “Randomized methods”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 955 records · Page 53
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

Urchin-like artificial gallium oxide nanowires grown by a novel MOCVD/CVD-based route for random laser application

A novel procedure based on a two-step method was developed to obtain β-Ga{sub 2}O{sub 3} nanowires by the chemical vapor deposition (CVD) method. The first step consists in the gallium micro-spheres growth inside a metal-organic chemical vapor deposition environment, using an organometallic precursor. Nanoscale spheres covering the microspheres were obtained. The second step involves the CVD oxidization of the gallium micro-spheres, which allow the formation of β-Ga{sub 2}O{sub 3} nanowires on the micro-sphere surface, with the final result being a nanostructure mimicking nature's sea urchin morphology. The grown nanomaterial is characterized by several techniques, including X-ray diffraction, scanning electron microscopy, energy-dispersive X-ray, transmission electron microscopy, and photoluminescence. A discussion about the growth mechanism and the optical properties of the β-Ga{sub 2}O{sub 3} material is presented considering its unknown true bandgap value (extending from 4.4 to 5.68 eV). As an application, the scattering properties of the nanomaterial are exploited to demonstrate random laser emission (around 570 nm) when it is permeated with a laser dye liquid solution.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Generation of random geological models using multi-randomization for machine learning

Generating high-fidelity geological models is essential for advancing machine learning (ML) methods in automated seismic interpretation. For instance, seismic images paired with corresponding fault labels are foundational for ML-based fault detection from seismic migration sections. While several open-access datasets of random geological models exist, open-source tools specifically designed to produce large volumes of such models for ML applications remain scarce. To address this gap, we present RGM (Random Geological Model), an open-source software package for efficiently generating 2D and 3D synthetic geological models tailored for ML workflows. RGM supports the creation of diverse model components, including medium property distributions (P-/S-wave velocities and density), seismic reflectivity images (i.e., synthetic migration sections), relative geological time, and discrete fault attributes such as probability, dip, strike, rake, and displacement. It also accommodates the creation of complex geological features such as salt bodies and unconformities. The model generation algorithm employs a multi-randomization strategy, yielding an effectively infinite-dimensional model space that encompasses a wide range of geological scenarios and associated seismic features. Furthermore, RGM incorporates a method to generate synthetic elastic migration images using analytical elastic reflection coefficients combined with frequency-dependent scaling. This functionality enables the creation of training datasets for ML models that leverage elastic seismic images. RGM is implemented in modern object-oriented Fortran, allowing users to flexibly control statistical parameters governing model variability. We demonstrate the capability, performance, and geological realism of the package through comprehensive 2D and 3D examples.

58 GEOSCIENCES↗
DOE OSTI2015

Prospective Randomized Comparison of the Effectiveness of Radiation Therapy and Local Steroid Injection for the Treatment of Plantar Fasciitis

Purpose: The purpose of this study was to conduct a randomized trial of radiation therapy for plantar fasciitis and to compare radiation therapy with local steroid injections. Methods and Materials: Between March 2013 and April 2014, 128 patients with plantar fasciitis were randomized to receive radiation therapy (total dose of 6.0 Gy applied in 6 fractions of 1.0 Gy three times a week) or local corticosteroid injections a 1 ml injection of 40 mg methylprednisolone and 0.5 ml 1% lidocaine under the guidance of palpation. The results were measured using a visual analog scale, a modified von Pannewitz scale, and a 5-level function score. The fundamental phase of the study was 3 months, with a follow-up period of up to 6 months. Results: The median follow-up period for all patients was 12.5 months (range, 6.5-18.6 months). For the radiation therapy patients, the median follow-up period was 13 months (range, 6.5-18.5 months), whereas in the palpation-guided (PG) steroid injection arm, it was 12.1 months (range, 6.5-18.6 months). After 3 months, results in the radiation therapy arm were significantly superior to those in the PG steroid injection arm (visual analog scale, P<.001; modified von Pannewitz scale, P<.001; 5-level function score, P<.001). Requirements for a second treatment did not significantly differ between the 2 groups, but the time interval for the second treatment was significantly shorter in the PG steroid injection group (P=.045). Conclusion: This study confirms the superior analgesic effect of radiation therapy compared to mean PG steroid injection on plantar fasciitis for at least 6 months after treatment.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Note: Non-invasive optical method for rapid determination of alignment degree of oriented nanofibrous layers

This paper presents a rapid non-destructive method that provides information on the anisotropic internal structure of nanofibrous layers. A laser beam of a wavelength of 632.8 nm is directed at and passes through a nanofibrous layer prepared by electrostatic spinning. Information about the structural arrangement of nanofibers in the layer is directly visible in the form of a diffraction image formed on a projection screen or obtained from measured intensities of the laser beam passing through the sample which are determined by the dependency of the angle of the main direction of polarization of the laser beam on the axis of alignment of nanofibers in the sample. Both optical methods were verified on Polyvinyl alcohol (PVA) nanofibrous layers (fiber diameter of 470 nm) with random, single-axis aligned and crossed structures. The obtained results match the results of commonly used methods which apply the analysis of electron microscope images. The presented simple method not only allows samples to be analysed much more rapidly and without damaging them but it also makes possible the analysis of much larger areas, up to several square millimetres, at the same time.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2011

Enhanced Ablation of High Intensity Focused Ultrasound with Microbubbles: An Experimental Study on Rabbit Hepatic VX2 Tumors

Purpose: This study was designed to assess the enhanced effect of high intensity focused ultrasound (HIFU) ablation with microbubbles on rabbit hepatic VX2 tumors and to compare the detection sensitivity of CEUS and CECT to determine the residual viable tissue after ablation of HIFU. Methods: Forty rabbits with hepatic VX2 tumors were randomly separated into two groups (20 animals per group) before HIFU ablation. A bolus of 0.2 mL of saline or a microbubble-based ultrasound (US) contrast agent was injected intravenously to group I rabbits and group II rabbits, respectively. The HIFU ablation procedure was started 15 s after the injection. Tumors were examined with grayscale contrast-enhanced ultrasound (CEUS) and contrast-enhanced computed tomography (CECT) immediately before and after HIFU ablation. Histopathologic assessment was performed immediately after treatment imaging. Results: Before ablation, intense contrast enhancement during arterial phase was observed at the whole tumors or the periphery of the tumors by CEUS and CECT. Lower HIFU energy was used in group II than in group I (P < 0.001). Histopathologic assessment revealed local residual viable tumor tissues due to incomplete ablation in 47.4% (9/19) of tumors in group I and 10% (2/20) of tumors in group II (P < 0.05). The concordance rate of CEUS (90.9%) with histopathology on residual tumor detection was higher than that of CECT (27.3%, P < 0.05). Conclusions: Introduction of the microbubble agent enhances HIFU therapeutic efficacy. CEUS proves to have high sensitivity in assessment of residual viable rabbit VX2 tumor after HIFU.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2024

Machine Learning Analysis of Impact of Western US Fires on Central US Hailstorms

Fires, including wildfires, harm air quality and essential public services like transportation, communication, and utilities. These fires can also influence atmospheric conditions, including temperature and aerosols, potentially affecting severe convective storms. Here, we investigate the remote impacts of fires in the western United States (WUS) on the occurrence of large hail (size: $\geqslant$ 2.54 cm) in the central US (CUS) over the 20-year period of 2001–20 using the machine learning (ML), Random Forest (RF), and Extreme Gradient Boosting (XGB) methods. The developed RF and XGB models demonstrate high accuracy (> 90%) and F1 scores of up to 0.78 in predicting large hail occurrences when WUS fires and CUS hailstorms coincide, particularly in four states (Wyoming, South Dakota, Nebraska, and Kansas). The key contributing variables identified from both ML models include the meteorological variables in the fire region (temperature and moisture), the westerly wind over the plume transport path, and the fire features (i.e., the maximum fire power and burned area). Importantly, the results confirm a linkage between WUS fires and severe weather in the CUS, corroborating the findings of our previous modeling study conducted on case simulations with a detailed physics model.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2019

Elastic Modulus Estimation for Copper Syntactic Foams Reinforced with Iron Hollow Spheres of Different Wall Thicknesses

The present study estimates the effect on Young’s modulus of the percentage and thickness of iron hollow spheres reinforcing copper syntactic foams. The resulting materials were modeled by inserting hollow spheres in a random arrangement, using a combination of Discrete Element Method (DEM) for generating the sphere location coordinates and Finite Element Analysis (FEA). Estimations for syntactic foams were compared to those obtained for conventional copper foams with the same porosity. Results showed that the elastic modulus of both syntactic and conventional foams drops significantly as the percentage of spheres or porosity increases. Furthermore, the increase in the wall thickness of the iron hollow spheres leads to significant rises in the Young’s modulus. Depending on the desired mechanical properties, the outcomes presented in this work could be used for selecting the kind of foam (syntactic or conventional) and predesign some characteristics such as porosity or thickness of the spheres or pores.

36 MATERIALS SCIENCE↗
DOE OSTI2018

On an asymptotic viscosity solution property of solutions of discrete Hamilton–Jacobi–Bellman equations

In this paper, we show that a proper limit of solutions of discrete Hamilton–Jacobi–Bellman (dHJB) equations in a random walk model becomes a viscosity solution of a Hamilton–Jacobi–Bellman (HJB) variational inequality in a continuous-time geometric Brownian model. HJB variational inequalities are used to analyze singular stochastic control problems in mathematical finance. By our result, with the help of dHJB equations, we can obtain viscosity solutions of HJB variational inequalities which are usually identified with the value functions of the singular stochastic control problems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

The role of surface chemistry in the synthesis of supported CuPd bimetallic/intermetallic catalysts for selective hydrogenation reactions

Understanding and controlling the structure of supported bimetallic/intermetallic catalysts are crucial for many heterogeneous catalytic reactions. However, commonly used synthesis techniques such as coimpregnation and coadsorption often yield nonuniform alloying and phase segregation, largely owing to the lack of interactions between different metal precursors. Here we show that by adopting sequential adsorption of complex metal cations and anions with the assistance of a proper ligand, the interaction between different metal precursors is enhanced, thus resulting in uniformly alloyed intermetallic catalysts. Importantly, the supported CuPd intermetallic catalyst exhibits greatly improved selectivity towards monoolefins in the semihydrogenation of acetylene and butadiene compared to those random CuPd alloys synthesized via coimpregnation and coadsorption methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2011

Evaluation of Biologic Effective Dose and Schedule of Fractionation for Preoperative Radiotherapy for Rectal Cancer: Meta-Analyses and Meta-Regression;Rectal cancer; Preoperative radiotherapy; Biologic effective dose; Meta-analysis

Purpose: To evaluate whether the risk of local recurrence depends on the biologic effective dose (BED) or fractionation dose in patients with resectable rectal cancer undergoing preoperative radiotherapy (RT) compared with surgery alone. Methods and Materials: A meta-analysis of randomized controlled trials (RCTs) was performed. The MEDLINE, Embase, CancerLit, and Cochrane Library databases were systematically searched for evidence. To evaluate the dose-response relationship, we conducted a meta-regression analysis. Four subgroups were created: Group 1, RCTs with a BED >30 Gy{sub 10} and a short RT schedule; Group 2, RCTs with BED >30 Gy{sub 10} and a long RT schedule; Group 3, RCTs with BED {<=}30 Gy{sub 10} and a short RT schedule; and Group 4, RCTs with BED {<=}30 Gy{sub 10} and a long RT schedule. Results: Our review identified 21 RCTs, yielding 9,097 patients. The pooled results from these 21 randomized trials of preoperative RT showed a significant reduction in mortality for groups 1 (p = .004) and 2 (p = .03). For local recurrence, the results were also significant in groups 1 (p = .00001) and 2 (p = .00001).The only subgroup that showed a greater sphincter preservation (SP) rate than surgery was group 2 (p = .03). The dose-response curve was linear (p = .006), and RT decreased the risk of local recurrence by about 1.7% for each Gy{sub 10} of BED. Conclusion: Our data have shown that RT with a BED of >30 Gy{sub 10} is more efficient in reducing local recurrence and mortality rates than a BED of {<=}30 Gy{sub 10}, independent of the schedule of fractionation used. A long RT schedule with a BED of >30 Gy{sub 10} should be recommended for sphincter preservation.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2012

Feasibility and Acute Toxicity of Hypofractionated Radiation in Large-breasted Patients

Purpose: To determine the feasibility of and acute toxicity associated with hypofractionated whole breast radiation (HypoRT) after breast-conserving surgery in patients excluded from or underrepresented in randomized trials comparing HypoRT with conventional fractionation schedules. Methods and Materials: A review was conducted of all patients consecutively treated with HypoRT at University of Chicago. All patients were treated to 42.56 Gy in 2.66 Gy daily fractions in either the prone or supine position. Planning was performed in most cases using wedges and large segments or a 'field-in-field' technique. Breast volume was estimated using volumetric measurements of the planning target volume (PTV). Dosimetric parameters of heterogeneity (V105, V107, V110, and maximum dose) were recorded for each treatment plan. Acute toxicity was scored for each treated breast. Results: Between 2006 and 2010, 78 patients were treated to 80 breasts using HypoRT. Most women were overweight or obese (78.7%), with a median body mass index of 29.2 kg/m{sup 2}. Median breast volume was 1,351 mL. Of the 80 treated breasts, the maximum acute skin toxicity was mild erythema or hyperpigmentation in 70.0% (56/80), dry desquamation in 21.25% (17/80), and focal moist desquamation in 8.75% (7/80). Maximum acute toxicity occurred after the completion of radiation in 31.9% of patients. Separation >25 cm was not associated with increased toxicity. Breast volume was the only patient factor significantly associated with moist desquamation on multivariable analysis (p = 0.01). Patients with breast volume >2,500 mL experienced focal moist desquamation in 27.2% of cases compared with 6.34% in patients with breast volume <2,500 mL (p = 0.03). Conclusions: HypoRT is feasible and safe in patients with separation >25 cm and in patients with large breast volume when employing modern planning and positioning techniques. We recommend counseling regarding expected increases in skin toxicity in women with a PTV volume >2,500 mL.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2012

Skin-sparing Helical Tomotherapy vs 3D-conformal Radiotherapy for Adjuvant Breast Radiotherapy: In Vivo Skin Dosimetry Study

Purpose: We investigated whether treatment-planning system (TPS)-calculated dose accurately reflects skin dose received for patients receiving adjuvant breast radiotherapy (RT) with standard three-dimensional conformal RT (3D-CRT) or skin-sparing helical tomotherapy (HT). Methods and Materials: Fifty patients enrolled in a randomized controlled trial investigating acute skin toxicity from adjuvant breast RT with 3D-CRT compared to skin-sparing HT, where a 5-mm strip of ipsilateral breast skin was spared. Thermoluminescent dosimetry or optically stimulated luminescence measurements were made in multiple locations and were compared to TPS-calculated doses. Skin dosimetric parameters and acute skin toxicity were recorded in these patients. Results: With HT there was a significant correlation between calculated and measured dose in the medial and lateral ipsilateral breast (r = 0.67, P<.001; r = 0.44, P=.03, respectively) and the medial and central contralateral breast (r = 0.73, P<.001; r = 0.88, P<.001, respectively). With 3D-CRT there was a significant correlation in the medial and lateral ipsilateral breast (r = 0.45, P=.03; r = 0.68, P<.001, respectively); the medial and central contralateral breast (r = 0.62, P=.001; r = 0.86, P<.001, respectively); and the mid neck (r = 0.42, P=.04, respectively). On average, HT-calculated dose overestimated the measured dose by 14%; 3D-CRT underestimated the dose by 0.4%. There was a borderline association between highest measured skin dose and moist desquamation (P=.05). Skin-sparing HT had greater skin homogeneity (homogeneity index of 1.39 vs 1.65, respectively; P=.005) than 3D-CRT plans. HT plans had a lower skin{sub V50} (1.4% vs 5.9%, respectively; P=.001) but higher skin{sub V40} and skin{sub V30} (71.7% vs 64.0%, P=.02; and 99.0% vs 93.8%, P=.001, respectively) than 3D-CRT plans. Conclusion: The 3D-CRT TPS more accurately reflected skin dose than the HT TPS, which tended to overestimate dose received by 14% in patients receiving adjuvant breast RT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2013

Quality Research in Radiation Oncology Analysis of Clinical Performance Measures in the Management of Gastric Cancer

Background: The specific aim was to determine national patterns of radiation therapy (RT) practice in patients treated for stage IB-IV (nonmetastatic) gastric cancer (GC). Methods and Materials: A national process survey of randomly selected US RT facilities was conducted which retrospectively assessed demographics, staging, geographic region, practice setting, and treatment by using on-site record review of eligible GC cases treated from 2005 to 2007. Three clinical performance measures (CPMs), (1) use of computed tomography (CT)-based treatment planning; (2) use of dose volume histograms (DVHs) to evaluate RT dose to the kidneys and liver; and (3) completion of RT within the prescribed time frame; and emerging quality indicators, (i) use of intensity modulated RT (IMRT); (ii) use of image-guided tools (IGRT) other than CT for RT target delineation; and (iii) use of preoperative RT, were assessed. Results: CPMs were computed for 250 eligible patients at 45 institutions (median age, 62 years; 66% male; 60% Caucasian). Using 2000 American Joint Committee on Cancer criteria, 13% of patients were stage I, 29% were stage II, 32% were stage IIIA, 10% were stage IIIB, and 12% were stage IV. Most patients (43%) were treated at academic centers, 32% were treated at large nonacademic centers, and 25% were treated at small to medium sized facilities. Almost all patients (99.5%) underwent CT-based planning, and 75% had DVHs to evaluate normal tissue doses to the kidneys and liver. Seventy percent of patients completed RT within the prescribed time frame. IMRT and IGRT were used in 22% and 17% of patients, respectively. IGRT techniques included positron emission tomography (n=20), magnetic resonance imaging (n=1), respiratory gating and 4-dimensional CT (n=22), and on-board imaging (n=10). Nineteen percent of patients received preoperative RT. Conclusions: This analysis of radiation practice patterns for treating nonmetastatic GC indicates widespread adoption of CT-based planning with use of DVH to evaluate normal tissue doses. Most patients completed adjuvant RT in the prescribed time frame. IMRT and IGRT were not routinely incorporated into clinical practice during the 2005-2007 period. These data will be a benchmark for future Quality Research in Radiation Oncology GC surveys.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2013

Chromosome Damage and Cell Proliferation Rates in In Vitro Irradiated Whole Blood as Markers of Late Radiation Toxicity After Radiation Therapy to the Prostate

Purpose: In vitro irradiated blood samples from prostate cancer patients showing late normal tissue damage were examined for lymphocyte response by measuring chromosomal aberrations and proliferation rate. Methods and Materials: Patients were selected from a randomized trial evaluating the optimal timing of dose-escalated radiation and short-course androgen deprivation therapy. Of 438 patients, 3% experienced grade 3 late radiation proctitis and were considered to be radiosensitive. Blood samples were taken from 10 of these patients along with 20 matched samples from patients with grade 0 proctitis. The samples were irradiated at 6 Gy and, along with control samples, were analyzed for dicentric chromosomes and excess fragments per cell. Cells in first and second metaphase were also enumerated to determine the lymphocyte proliferation rate. Results: At 6 Gy, there were statistically significant differences between the radiosensitive and control cohorts for 3 endpoints: the mean number of dicentric chromosomes per cell (3.26 ± 0.31, 2.91 ± 0.32; P=.0258), the mean number of excess fragments per cell (2.27 ± 0.23, 1.43 ± 0.37; P<.0001), and the proportion of cells in second metaphase (0.27 ± 0.10, 0.46 ± 0.09; P=.0007). Conclusions: These results may be a valuable indicator for identifying radiosensitive patients and for tailoring radiation therapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Nomograms for Prediction of Outcome With or Without Adjuvant Radiation Therapy for Patients With Endometrial Cancer: A Pooled Analysis of PORTEC-1 and PORTEC-2 Trials

Background: Postoperative radiation therapy for stage I endometrial cancer improves locoregional control but is without survival benefit. To facilitate treatment decision support for individual patients, accurate statistical models to predict locoregional relapse (LRR), distant relapse (DR), overall survival (OS), and disease-free survival (DFS) are required. Methods and Materials: Clinical trial data from the randomized Post Operative Radiation Therapy for Endometrial Cancer (PORTEC-1; N=714 patients) and PORTEC-2 (N=427 patients) trials and registered group (grade 3 and deep invasion, n=99) were pooled for analysis (N=1240). For most patients (86%) pathology review data were available; otherwise original pathology data were used. Trial variables which were clinically relevant and eligible according to data constraints were age, stage, given treatment (pelvic external beam radiation therapy (EBRT), vaginal brachytherapy (VBT), or no adjuvant treatment, FIGO histological grade, depth of invasion, and lymph-vascular invasion (LVSI). Multivariate analyses were based on Cox proportional hazards regression model. Predictors were selected based on a backward elimination scheme. Model results were expressed by the c-index (0.5-1.0; random to perfect prediction). Two validation sets (n=244 and 291 patients) were used. Results: Accuracy of the developed models was good, with training accuracies between 0.71 and 0.78. The nomograms validated well for DR (0.73), DFS (0.69), and OS (0.70), but validation was only fair for LRR (0.59). Ranking of variables as to their predictive power showed that age, tumor grade, and LVSI were highly predictive for all outcomes, and given treatment for LRR and DFS. The nomograms were able to significantly distinguish low- from high-probability patients for these outcomes. Conclusions: The nomograms are internally validated and able to accurately predict long-term outcome for endometrial cancer patients with observation, pelvic EBRT, or VBT after surgery. These models facilitate decision support in daily clinical practice and can be used for patient counseling and shared decision making, selecting patients who benefit most from adjuvant treatment, and generating new hypotheses.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2023

Machine learning models for estimating contamination across different curbside collection strategies

Contaminated recyclables, which are frequently discarded as waste, pose a significant challenge to the implementation of a circular economy. These contaminated recyclables impede the circulation of resources, resulting in higher processing costs at material recovery facilities (MRFs). Over the past few decades, machine learning (ML) models such as linear regression (LR), support vector machine (SVM), and random forest (RF) have evolved to provide new methods for predicting inbound contamination rates in addition to traditional statistical models. In this study, we applied ML models to predict inbound contamination rates using demographic features from 15 counties in the U.S. with different curbside collection strategies. In general, we found that ML models outperformed linear mixed models. Specifically, SVM models had the highest performance (R 2 = 0.75; mean absolute error (MAE) = 0.06), which may be due to their ability to model nonlinear relationships between features and inbound contamination rates. Further, the key predictor was population, with poverty rate being positively correlated and median age negatively correlated with inbound contamination rates. To improve the management of contamination and enhance the implementation of a circular economy, better models are needed to understand and estimate inbound contamination rates as well as identify critical factors in the present and future.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Fast-RF-Shimming: Accelerate RF shimming in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) offers an elevated signal-to-noise ratio (SNR), enabling exceptionally high spatial resolution that benefits both clinical diagnostics and advanced research. However, the jump to higher fields introduces complications, particularly transmit radiofrequency (RF) field ($B^{+}_{1}$) inhomogeneities, manifesting as uneven flip angles and image intensity irregularities. These artifacts can degrade image quality and impede broader clinical adoption. Traditional RF shimming methods, such as Magnitude Least Squares (MLS) optimization, effectively mitigate $B^{+}_{1}$ inhomogeneity, but remain time-consuming. Recent machine learning approaches, including RF Shim Prediction by Iteratively Projected Ridge Regression and other deep learning architectures, suggest alternative pathways. Although these approaches show promise, challenges such as extensive training periods, limited network complexity, and practical data requirements persist. In this paper, we introduce a holistic learning-based framework called Fast-RF-Shimming, which achieves a 5000 ​× ​speed-up compared to the traditional MLS method. In the initial phase, we employ random-initialized Adaptive Moment Estimation (Adam) to derive the desired reference shimming weights from multi-channel $B^{+}_{1}$ fields. Next, we train a Residual Network (ResNet) to map $B^{+}_{1}$ fields directly to the ultimate RF shimming outputs, incorporating the confidence parameter into its loss function. Finally, we design Non-uniformity Field Detector (NFD), an optional post-processing step, to ensure the extreme non-uniform outcomes are identified. Comparative evaluations with standard MLS optimization underscore notable gains in both processing speed and predictive accuracy, which indicates that our technique shows a promising solution for addressing persistent inhomogeneity challenges.

Deep learning↗