arxiv: 2604.06020 · v1 · submitted 2026-04-07 · ⚛️ physics.med-ph

Recognition: 2 theorem links

· Lean Theorem

Optimizing IMPULSED Acquisition Protocols for Clinical 3T Scanners Through Bayesian Experimental Design

Yan Dai , Xun Jia , Todd Aguilera , Kai Jiang , Arely Perez Rodriguez , Isabelle Vanhaezebrouck , Jie Deng

Authors on Pith no claims yet

Pith reviewed 2026-05-10 18:38 UTC · model grok-4.3

classification ⚛️ physics.med-ph

keywords Bayesian optimizationdiffusion MRIIMPULSED modelprotocol optimization3T clinical scannercellular microstructuretumor imagingexpected information gain

0 comments

The pith

Bayesian optimization designs superior IMPULSED diffusion protocols for clinical 3T scanners

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper aims to show that choosing diffusion MRI measurements for the IMPULSED model through Bayesian experimental design produces protocols that extract more useful information within the time and hardware limits of routine 3T scanners. A sympathetic reader would care because accurate estimates of cell size, density, and other microscopic properties can help characterize tumors and track treatment response without specialized equipment. The work maximizes expected information gain by searching over pulse types, diffusion times, and b-values with Gaussian process surrogates, then tests the resulting design in simulations and living canine tumors.

Core claim

The optimized protocol eliminates OGSEn2 acquisitions, concentrates measurements at high b-values, and uses jointly tuned diffusion timing. Compared with a heuristic baseline, it achieves higher accuracy when distinguishing cell populations and lower errors when estimating IMPULSED parameters across biologically relevant ranges and SNR levels from 5 to 40. These gains hold under varied priors and noise assumptions, and they appear as smoother, higher-quality parameter maps in in-vivo 3T data from canine tumors.

What carries the argument

Bayesian optimization using Gaussian process surrogates to maximize expected information gain over the space of pulse sequence types, diffusion times, and b-values for fitting the IMPULSED model

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same optimization framework could be applied directly to other diffusion MRI models that aim to extract cellular microstructure.
Clinical sites could shorten total scan time while retaining or improving parameter precision by replacing heuristic protocols with this approach.
Incorporating real-time motion or contrast-agent effects into the expected information gain calculation would make the designs even more practical for patient scans.

Load-bearing premise

The prior distributions over IMPULSED parameters and the assumed SNR levels used to compute expected information gain must accurately represent real biological variability and scanner noise.

What would settle it

Acquiring new in-vivo 3T tumor data with both the optimized and baseline protocols and finding equal or worse classification accuracy or parameter estimation error on the optimized version would disprove the claimed improvement.

Figures

Figures reproduced from arXiv: 2604.06020 by Arely Perez Rodriguez, Isabelle Vanhaezebrouck, Jie Deng, Kai Jiang, Todd Aguilera, Xun Jia, Yan Dai.

read the original abstract

To optimize diffusion MRI acquisition protocols for IMPULSED model at clinical 3T scanner using Bayesian experimental design, enabling accurate cellular-scale parameter estimation under realistic scan time and scanner hardware constraints. Expected Information Gain (EIG) was used as the optimization objective to maximize the information content of acquired measurements for IMPULSED model fitting. Bayesian optimization with Gaussian process surrogates efficiently searched the high-dimensional acquisition parameter space, including pulse types (PGSE, OGSEn1, and OGSEn2), diffusion times, and b-values. Optimized protocols were systematically evaluated against a heuristically designed baseline protocol through simulation studies assessing classification accuracy and parameter estimation performance across SNR levels of 5-40. Robustness to optimization assumptions was examined by varying prior distributions and assumed SNR. In-vivo validation was performed using canine tumor data acquired at 3T. The optimized protocol eliminated OGSEn2 acquisitions, concentrated measurements at high b-values, employing concurrently optimized diffusion timing. Compared to the baseline protocol, the optimized design achieved superior classification accuracy for distinguishing cell populations and reduced parameter estimation error across biologically relevant parameter ranges at various SNRs. Performance advantages were consistent across diverse optimization scenarios, demonstrating robustness to prior knowledge and noise assumptions. In-vivo parameter maps showed substantially improved quality and smoothness. Bayesian optimization substantially improves IMPULSED acquisition design for clinical 3T scanners. This principled, algorithm-agnostic framework enables accurate diffusion MRI cytometry under clinical constraints, with potential applications to tumor characterization and treatment monitoring.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Bayesian optimization improves IMPULSED protocols in simulations at 3T but real gains need stronger in-vivo proof.

read the letter

The key point is that Bayesian experimental design produces a protocol for IMPULSED that outperforms the baseline in simulations for clinical 3T scanners, though the in-vivo evidence remains qualitative. They search the acquisition space with expected information gain and Gaussian process surrogates. The optimized protocol drops OGSEn2, concentrates on high b-values, and jointly tunes the diffusion times. Simulations across SNR levels from 5 to 40 show better accuracy in classifying cell populations and lower errors in parameter estimates. Robustness checks by varying priors and SNR keep the gains consistent. The soft spots line up with the stress-test note. The results depend on the chosen priors for cell parameters and the modeled SNR. Real clinical 3T data could have broader variability or different noise floors than tested, which would change which protocol is actually best. The canine in-vivo maps are described as improved in quality and smoothness, but without quantitative metrics or ground-truth comparisons the advantage is hard to measure. No human data is shown. This work is aimed at medical physicists optimizing diffusion MRI for tumor microstructure. Readers looking for a practical example of Bayesian design under scanner constraints will find the simulation framework and the specific protocol useful. The approach is internally consistent and engages the literature on IMPULSED. It deserves a serious referee. Reviewers can press for more quantitative validation while the core idea and simulation results are solid enough to review.

Referee Report

2 major / 2 minor

Summary. The paper claims that using Bayesian experimental design with Expected Information Gain (EIG) as the objective, combined with Bayesian optimization via Gaussian process surrogates, yields an optimized IMPULSED acquisition protocol for clinical 3T scanners. The optimized protocol eliminates OGSEn2 acquisitions, concentrates measurements at high b-values, and jointly optimizes diffusion timings. Simulations across SNR 5-40 show superior cell-population classification accuracy and reduced parameter estimation error versus a heuristic baseline; robustness checks vary priors and SNR; in-vivo canine tumor data at 3T exhibit qualitatively improved parameter-map quality and smoothness.

Significance. If the central claims hold, the work supplies a principled, model-agnostic framework for designing time-efficient protocols that improve cellular-scale parameter recovery under realistic 3T hardware and scan-time limits. The EIG formulation is computed from the forward model and priors independently of any fitted parameters, the simulation studies use held-out data, and robustness checks are performed; these elements strengthen the contribution for clinical translation of IMPULSED-based cytometry.

major comments (2)

[In-vivo validation] In-vivo validation section: superiority is asserted via qualitative statements of 'substantially improved quality and smoothness' on canine tumor maps, but no quantitative head-to-head metrics (e.g., voxel-wise parameter variance, fit residuals, or classification accuracy) against the baseline protocol are reported, weakening the real-data support for the central claim.
[Robustness to optimization assumptions] Robustness analysis: while priors over IMPULSED parameters (cell radius, intracellular fraction, etc.) and SNR (5-40) are varied, the manuscript does not demonstrate that the tested ranges encompass the broader cell-size heterogeneity, partial-volume effects, or scanner-specific noise floors encountered in human tumors at 3T; if the true distributions lie outside these ranges, the selected protocol may not maximize information gain on clinical data.

minor comments (2)

[Methods] Define the precise pulse-sequence parameters for OGSEn1 and OGSEn2 (e.g., number of oscillations, gradient waveforms) in the methods to ensure reproducibility.
[Abstract] The abstract states 'Bayesian optimization substantially improves IMPULSED acquisition design'; this phrasing should be tempered to reflect that the improvement is demonstrated relative to one specific baseline protocol.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive evaluation and constructive comments. We address each major comment below and indicate where revisions have been made to the manuscript.

read point-by-point responses

Referee: [In-vivo validation] In-vivo validation section: superiority is asserted via qualitative statements of 'substantially improved quality and smoothness' on canine tumor maps, but no quantitative head-to-head metrics (e.g., voxel-wise parameter variance, fit residuals, or classification accuracy) against the baseline protocol are reported, weakening the real-data support for the central claim.

Authors: We agree that quantitative head-to-head metrics on the in-vivo data would provide stronger support. The canine tumor scans were performed exclusively with the optimized protocol to demonstrate feasibility under clinical 3T constraints and scan-time limits; a paired acquisition with the baseline protocol on the same animals was not conducted due to ethical and practical considerations. As a result, direct voxel-wise comparisons (variance, residuals, or classification accuracy) against the baseline cannot be computed from the existing dataset. The quantitative superiority is established through the simulation studies with held-out data. In the revised manuscript we have added an explicit discussion of this limitation, clarified the role of the in-vivo results as a qualitative feasibility demonstration, and included additional fit-residual maps from the optimized-protocol data to support the observed improvements in smoothness. revision: partial
Referee: [Robustness to optimization assumptions] Robustness analysis: while priors over IMPULSED parameters (cell radius, intracellular fraction, etc.) and SNR (5-40) are varied, the manuscript does not demonstrate that the tested ranges encompass the broader cell-size heterogeneity, partial-volume effects, or scanner-specific noise floors encountered in human tumors at 3T; if the true distributions lie outside these ranges, the selected protocol may not maximize information gain on clinical data.

Authors: We acknowledge that the original robustness checks, while spanning literature-derived ranges for cell radius and intracellular fraction, do not explicitly cover the full spectrum of human-tumor heterogeneity or partial-volume effects. In the revised manuscript we have expanded the sensitivity analysis to include wider prior distributions informed by human prostate and glioma studies, added a dedicated discussion of partial-volume effects and scanner-specific noise floors at 3T, and noted that the EIG objective is inherently prior-dependent. The framework can be re-optimized with patient-specific priors as new data become available. revision: partial

Circularity Check

0 steps flagged

No load-bearing circularity; EIG optimization and held-out evaluation are independent

full rationale

The paper computes expected information gain directly from the IMPULSED forward model and chosen priors over parameters (cell radius, intracellular fraction, etc.) plus fixed SNR values; this is a forward simulation step with no fitted data or parameters from the target acquisitions entering the objective. Protocol selection via Bayesian optimization with GP surrogates therefore does not reduce to any quantity defined by a fit. Evaluation proceeds on separate simulation draws (held-out from the optimization priors/SNR) and on real in-vivo canine tumor data acquired with both protocols; reported gains in classification accuracy and parameter error are measured against an independently designed baseline protocol. No equation or self-citation chain equates the claimed superiority to the optimization inputs by construction. Minor self-citation risk exists only in the general Bayesian design literature, which is not load-bearing for the specific protocol comparison.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claim rests on the validity of the IMPULSED biophysical forward model, standard assumptions of Gaussian process regression for surrogate modeling, and the choice of priors and noise levels that are tested for sensitivity but still constitute modeling choices.

free parameters (2)

Prior distributions over IMPULSED cellular parameters
Used to compute expected information gain; varied in robustness experiments but selected from biological knowledge.
Assumed SNR range (5-40)
Central to both optimization and performance evaluation; tested across values but fixed for each run.

axioms (2)

domain assumption Gaussian process surrogate provides an accurate and smooth approximation to the expected information gain surface over the acquisition parameter space
Invoked to enable efficient Bayesian optimization in the high-dimensional space of pulse types, timings, and b-values.
domain assumption The IMPULSED model equations correctly capture the relationship between diffusion measurements and underlying cellular parameters in tumor tissue
Required for both the information-gain calculation and the subsequent parameter estimation performance claims.

pith-pipeline@v0.9.0 · 5590 in / 1476 out tokens · 69203 ms · 2026-05-10T18:38:31.969062+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The IMPULSED model is a two-compartment framework... S = Vin ⋅ Sin + (1 − Vin) ⋅ Sex

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

48 extracted references

[1]

Quantitative parameters are obtained by fitting biophysical models to the acquired signals, providing imaging biomarkers about cell microenvironment(3-6)

Introduction Diffusion magnetic resonance imaging (dMRI ) characterizes tissue microstructure by measuring signal attenuation caused by water molecule displacement under diffusion -weighted gradients (1,2). Quantitative parameters are obtained by fitting biophysical models to the acquired signals, providing imaging biomarkers about cell microenvironment(3...
[2]

Conceptually, it connects the observed diffusion signal 𝑆 with contributions from restricted intracellular and extracellular diffusion

Methods 2.1 IMPULSED Model The IMPULSED model is a two -compartment framework assuming isotropic water diffusion both inside and outside sphere -shaped cells with impermeable membranes. Conceptually, it connects the observed diffusion signal 𝑆 with contributions from restricted intracellular and extracellular diffusion. The signal is expressed as a weight...

2000
[3]

The Optimized SNR 20 protocol exhibits several distinctive features compared to the baseline protocol

Results 3.1 Experimental Optimization Results The detailed acquisition parameters of the optimized experiment al design under the uniform prior distributions at SNR of 20 (Optimized SNR 20), together with the baseline protocol (Baseline), are shown in Table 1. The Optimized SNR 20 protocol exhibits several distinctive features compared to the baseline pro...

2000
[4]

Discussion 4.1 Insights into Optimized Acquisition Strategies Across optimization scenarios, the optimized protocols consistently concentrated measurements at high b - values. This pattern indicates that IMPULSED signals —particularly those acquired with OGSEn1 and OGSEn2—are highly sensitive to measurement noise, such that observed signal decay under cli...

2000
[5]

Conclusion In this study, we developed and validated a Bayesian experiment al design optimization framework for IMPULSED diffusion MRI cytometry . We demonstrated that information -theoretic optimization of pulse sequence parameters yields substantially more informative data acquisition, which improves both classification discriminability and parameter es...
[6]

MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders

Le Bihan D, Breton E, Lallemand D, Grenier P , Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161(2):401-407

1986
[7]

Clinical intravoxel incoherent motion and diffusion MR imaging: past, present, and future

Iima M, Le Bihan D. Clinical intravoxel incoherent motion and diffusion MR imaging: past, present, and future. Radiology 2016;278(1):13-32

2016
[8]

The role and potential of using quantitative MRI biomarkers for imaging guidance in brain cancer radiotherapy treatment planning: A systematic review

Aldawsari AM, Al -Qaisieh B, Broadbent DA, Bird D, Murray L, Speight R. The role and potential of using quantitative MRI biomarkers for imaging guidance in brain cancer radiotherapy treatment planning: A systematic review. Physics and imaging in radiation oncology 2023;27:100476

2023
[9]

Time-dependent diffusion MRI for quantitative microstructural mapping of prostate cancer

Wu D, Jiang K, Li H, Zhang Z, Ba R, Zhang Y , Hsu Y -C, Sun Y , Zhang Y-D. Time-dependent diffusion MRI for quantitative microstructural mapping of prostate cancer. Radiology 2022;303(3):578-587

2022
[10]

Iima M, Kataoka M, Honda M, Ohashi A, Kishimoto AO, Ota R, Uozumi R, Urushibata Y , Feiweier T, Toi M. The rate of apparent diffusion coefficient change with diffusion time on breast diffusion - weighted imaging depends on breast tumor types and molecular prognostic biomarker expression. Investigative radiology 2021;56(8):501-508

2021
[11]

Time dependence in diffusion MRI predicts tissue outcome in ischemic stroke patients

Lampinen B, Lätt J, Wasselius J, van Westen D, Nilsson M. Time dependence in diffusion MRI predicts tissue outcome in ischemic stroke patients. Magnet Reson Med 2021;86(2):754-764

2021
[12]

Geometry of nonlinear least squares with applications to sloppy models and optimization

Transtrum MK, Machta BB, Sethna JP . Geometry of nonlinear least squares with applications to sloppy models and optimization. Physical Review E 2011;83(3):036701

2011
[13]

What do we mean by identifiability in mixed effects models? Journal of pharmacokinetics and pharmacodynamics 2016;43:111-122

Lavielle M, Aarons L. What do we mean by identifiability in mixed effects models? Journal of pharmacokinetics and pharmacodynamics 2016;43:111-122

2016
[14]

Deep learning how to fit an intravoxel incoherent motion model to diffusion‐weighted MRI

Barbieri S, Gurney ‐Champion OJ, Klaassen R, Thoeny HC. Deep learning how to fit an intravoxel incoherent motion model to diffusion‐weighted MRI. Magnet Reson Med 2020;83(1):312-321

2020
[15]

Comparison of methods for estimation of the intravoxel incoherent motion (IVIM) diffusion coefficient (D) and perfusion fraction (f)

Jalnefjord O, Andersson M, Montelius M, Starck G, Elf A -K, Johanson V , Svensson J, Ljungberg M. Comparison of methods for estimation of the intravoxel incoherent motion (IVIM) diffusion coefficient (D) and perfusion fraction (f). Magnetic Resonance Materials in Physics, Biology and Medicine 2018;31:715-723

2018
[16]

Otikovs M, Basak A, Frydman L. Spatiotemporal encoding MRI using subspace -constrained sampling and locally -low-rank regularization: Applications to diffusion weighted and diffusion kurtosis imaging of human brain and prostate. Magn Reson Imaging 2022;94:151-160

2022
[17]

qModeL: A plug ‐and‐play model‐based reconstruction for highly accelerated multi‐shot diffusion MRI using learned priors

Mani M, Magnotta VA, Jacob M. qModeL: A plug ‐and‐play model‐based reconstruction for highly accelerated multi‐shot diffusion MRI using learned priors. Magnet Reson Med 2021;86(2):835-851

2021
[18]

Improving microstructural estimation in time ‐dependent diffusion MRI with a Bayesian method

Liu K, Lin Z, Zheng T, Ba R, Zhang Z, Li H, Zhang H, Tal A, Wu D. Improving microstructural estimation in time ‐dependent diffusion MRI with a Bayesian method. Journal of Magnetic Resonance Imaging 2025;61(2):724-734

2025
[19]

IMPULSED model based cytological feature estimation with U ‐Net: Application to human brain tumor at 3T

Wu J, Kang T, Lan X, Chen X, Wu Z, Wang J, Lin L, Cai C, Lin J, Ding X. IMPULSED model based cytological feature estimation with U ‐Net: Application to human brain tumor at 3T. Magnet Reson Med 2023;89(1):411-422

2023
[20]

Time -dependent diffusion MRI helps predict molecular subtypes and treatment response to neoadjuvant chemotherapy in breast cancer

Wang X, Ba R, Huang Y , Cao Y , Chen H, Xu H, Shen H, Liu D, Huang H, Yin T. Time -dependent diffusion MRI helps predict molecular subtypes and treatment response to neoadjuvant chemotherapy in breast cancer. Radiology 2024;313(1):e240288

2024
[21]

Advanced diffusion‐weighted MRI for cancer microstructure assessment in body imaging, and its relationship with histology

Fokkinga E, Hernandez ‐Tamames JA, Ianus A, Nilsson M, Tax CM, Perez ‐Lopez R, Grussu F . Advanced diffusion‐weighted MRI for cancer microstructure assessment in body imaging, and its relationship with histology. Journal of Magnetic Resonance Imaging 2024;60(4):1278-1304

2024
[22]

Magnetic resonance imaging of mean cell size in human breast tumors

Xu J, Jiang X, Li H, Arlinghaus LR, McKinley ET, Devan SP , Hardy BM, Xie J, Kang H, Chakravarthy AB. Magnetic resonance imaging of mean cell size in human breast tumors. Magnet Reson Med 2020;83(6):2002-2014

2020
[23]

Time -dependent diffusion MRI-based microstructural mapping for differentiating high -grade serous ovarian cancer from serous borderline ovarian tumor

Cao Y , Lu Y , Shao W, Zhai W, Song J, Zhang A, Huang S, Zhao X, Cheng W, Wu F . Time -dependent diffusion MRI-based microstructural mapping for differentiating high -grade serous ovarian cancer from serous borderline ovarian tumor. European Journal of Radiology 2024;178:111622

2024
[24]

Values and count optimization of IVIM diffusion weighting parameter

Shafna V , Raju J, John A, Ushadevi AC. Values and count optimization of IVIM diffusion weighting parameter. 2020. IEEE. p 459-464

2020
[25]

An optimized b‐value sampling for the quantification of interstitial fluid using diffusion ‐weighted MRI, a genetic algorithm approach

Drenthen GS, Jansen JF , van der Thiel MM, Voorter PH, Backes WH. An optimized b‐value sampling for the quantification of interstitial fluid using diffusion ‐weighted MRI, a genetic algorithm approach. Magnet Reson Med 2023;90(1):194-201

2023
[26]

An optimized b ‐value distribution for triexponential intravoxel incoherent motion (IVIM) in the liver

Riexinger A, Martin J, Wetscherek A, Kuder TA, Uder M, Hensel B, Laun FB. An optimized b ‐value distribution for triexponential intravoxel incoherent motion (IVIM) in the liver. Magnet Reson Med 2021;85(4):2095-2108

2021
[27]

Optimization of b ‐value distribution for biexponential diffusion ‐weighted MR imaging of normal prostate

Jambor I, Merisaari H, Aronen HJ, Järvinen J, Saunavaara J, Kauko T, Borra R, Pesola M. Optimization of b ‐value distribution for biexponential diffusion ‐weighted MR imaging of normal prostate. Journal of magnetic resonance imaging 2014;39(5):1213-1222

2014
[28]

Optimizing b ‐values schemes for diffusion MRI of the brain with segmented Intravoxel incoherent motion (IVIM) model

Paganelli C, Zampini MA, Morelli L, Buizza G, Fontana G, Anemoni L, Imparato S, Riva G, Iannalfi A, Orlandi E. Optimizing b ‐values schemes for diffusion MRI of the brain with segmented Intravoxel incoherent motion (IVIM) model. Journal of Applied Clinical Medical Physics 2023;24(6):e13986

2023
[29]

A novel approach for b -value optimization in Intravoxel Incoherent Motion Imaging using Metaheuristic algorithm

Raju J, John A. A novel approach for b -value optimization in Intravoxel Incoherent Motion Imaging using Metaheuristic algorithm. Expert Systems with Applications 2021;168:114270

2021
[30]

Toward an optimal distribution of b values for intravoxel incoherent motion imaging

Lemke A, Stieltjes B, Schad LR, Laun FB. Toward an optimal distribution of b values for intravoxel incoherent motion imaging. Magn Reson Imaging 2011;29(6):766-776

2011
[31]

Optimization of b‐value sampling for diffusion ‐weighted imaging of the kidney

Zhang JL, Sigmund EE, Rusinek H, Chandarana H, Storey P , Chen Q, Lee VS. Optimization of b‐value sampling for diffusion ‐weighted imaging of the kidney. Magnet Reson Med 2012;67(1):89 - 97

2012
[32]

D -optimal design of b -values for precise intra -voxel incoherent motion imaging

Sansone M, Fusco R, Petrillo A. D -optimal design of b -values for precise intra -voxel incoherent motion imaging. Biomedical Physics & Engineering Express 2019;5(3):035025

2019
[33]

Optimization of b ‐value schemes for estimation of the diffusion coefficient and the perfusion fraction with segmented intravoxel incoherent motion model fitting

Jalnefjord O, Montelius M, Starck G, Ljungberg M. Optimization of b ‐value schemes for estimation of the diffusion coefficient and the perfusion fraction with segmented intravoxel incoherent motion model fitting. Magnet Reson Med 2019;82(4):1541-1552

2019
[34]

Determination of optimized set of b-values for apparent diffusion coefficient mapping in liver diffusion -weighted MRI

Pena-Nogales O, Hernando D, Aja -Fernandez S, de Luis -Garcia R. Determination of optimized set of b-values for apparent diffusion coefficient mapping in liver diffusion -weighted MRI. Journal of Magnetic Resonance 2020;310:106634

2020
[35]

Comparison of fitting methods and b ‐value sampling strategies for intravoxel incoherent motion in breast cancer

Cho GY , Moy L, Zhang JL, Baete S, Lattanzi R, Moccaldi M, Babb JS, Kim S, Sodickson DK, Sigmund EE. Comparison of fitting methods and b ‐value sampling strategies for intravoxel incoherent motion in breast cancer. Magnet Reson Med 2015;74(4):1077-1085

2015
[36]

Diffusion MRI in peripheral nerves: optimized b values and the role of non -gaussian diffusion

Foesleitner O, Sulaj A, Sturm V , Kronlage M, Godel T, Preisner F , Nawroth PP , Bendszus M, Heiland S, Schwarz D. Diffusion MRI in peripheral nerves: optimized b values and the role of non -gaussian diffusion. Radiology 2022;302(1):153-161

2022
[37]

Optimization of intravoxel incoherent motion (IVIM): variability of parameters measurements using a reduced distribution of b values for breast tumors analysis

Doudou NR, Liu Y , Kampo S, Zhang K, Dai Y , Wang S. Optimization of intravoxel incoherent motion (IVIM): variability of parameters measurements using a reduced distribution of b values for breast tumors analysis. Magnetic Resonance Materials in Physics, Biology and Medicine 2020;33(2):273 - 281

2020
[38]

Effect of combination and number of b values in IVIM analysis with post-processing methodology: simulation and clinical study

Malagi AV , Das CJ, Khare K, Calamante F , Mehndiratta A. Effect of combination and number of b values in IVIM analysis with post-processing methodology: simulation and clinical study. Magnetic Resonance Materials in Physics, Biology and Medicine 2019;32(5):519-527

2019
[39]

Optimized Combination of b‐values for IVIM Perfusion Imaging in Acute Ischemic Stroke Patients

Zhu G, Heit JJ, Martin BW, Marcellus DG, Federau C, Wintermark M. Optimized Combination of b‐values for IVIM Perfusion Imaging in Acute Ischemic Stroke Patients. Clinical neuroradiology 2020;30(3):535-544

2020
[40]

A review of population -based metaheuristics for large -scale black -box global optimization—Part I

Omidvar MN, Li X, Yao X. A review of population -based metaheuristics for large -scale black -box global optimization—Part I. IEEE Transactions on Evolutionary Computation 2021;26(5):802-822

2021
[41]

Quantification of intravoxel incoherent motion with optimized b‐values using deep neural network

Lee W, Kim B, Park H. Quantification of intravoxel incoherent motion with optimized b‐values using deep neural network. Magnet Reson Med 2021;86(1):230-244

2021
[42]

Quantification of cell size using temporal diffusion spectroscopy

Jiang X, Li H, Xie J, Zhao P , Gore JC, Xu J. Quantification of cell size using temporal diffusion spectroscopy. Magnet Reson Med 2016;75(3):1076-1085

2016
[43]

MR cell size imaging with temporal diffusion spectroscopy

Jiang X, Li H, Devan SP , Gore JC, Xu J. MR cell size imaging with temporal diffusion spectroscopy. Magn Reson Imaging 2021;77:109-123

2021
[44]

Characterization of tissue structure at varying length scales using temporal diffusion spectroscopy

Gore JC, Xu J, Colvin DC, Yankeelov TE, Parsons EC, Does MD. Characterization of tissue structure at varying length scales using temporal diffusion spectroscopy. NMR in Biomedicine 2010;23(7):745-756

2010
[45]

In vivo imaging of cancer cell size and cellularity using temporal diffusion spectroscopy

Jiang X, Li H, Xie J, McKinley ET, Zhao P , Gore JC, Xu J. In vivo imaging of cancer cell size and cellularity using temporal diffusion spectroscopy. Magnet Reson Med 2017;78(1):156-164

2017
[46]

Modern Bayesian experimental design

Rainforth T, Foster A, Ivanova DR, Bickford Smith F . Modern Bayesian experimental design. Statistical Science 2024;39(1):100-114

2024
[47]

Maximum entropy sampling and optimal Bayesian experimental design

Sebastiani P , Wynn HP . Maximum entropy sampling and optimal Bayesian experimental design. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2000;62(1):145-157

2000
[48]

On modeling

Novikov DS, Kiselev VG, Jespersen SN. On modeling. Magnet Reson Med 2018;79(6):3172-3193. Tables Protocol Type Sequence 𝛿/Δ⁡(ms) 𝑡diff⁡(ms) Nonzero 𝑏-value (s/mm2) Optimize SNR 20 PGSE 33/59 48 688(1), 800(1), 829(1), 931(1), 2000(4) OGSEn1 41/51 10.25 1220(5) Baseline PGSE 12/74 70 250(1), 500(1), 750(1), 1000(1), 1400(1),1800(1) OGSEn1 41/51 10.25 250(1...

2018