Repeatability of IVIM biomarkers from diffusion‐weighted MRI in head and neck: Bayesian probability versus neural network

Koopman, Thomas; Martens, Roland M.; Gurney‐Champion, Oliver J.; Yaqub, Maqsood; Lavini, Cristina; Graaf, Pim de; Castelijns, Jonas A.; Boellaard, Ronald; Marcus, J. Tim

doi:10.1002/mrm.28671

Cited by 25 publications

(32 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They found that both Bayesian and their neural network approach outperformed NLLS in terms of having lower within-subject coefficient of variations; however, the average IVIM parameter values were not provided. Given the improved repeatability observed by Koopman et al in muscle, 48 and reliability observed in other organs, 86 Bayesian approaches may prove useful for estimation of IVIM parameters in future investigations.…”

Section: Systematic Reviewmentioning

confidence: 92%

“…Various fitting strategies can be employed to obtain the parameters of interest, generally categorized as constrained nonlinear least squares (NLLS), Bayesian, or most recently deep learning approaches. 48 Within the constrained NLLS, all IVIM parameters can be obtained from a single step or a segmented approach can be used. Because D* is generally thought to be at least an order of magnitude greater than D, the intravascular contribution to the bi-exponential decay is negligible at high b-values.…”

Section: Ivim Theorymentioning

confidence: 99%

“…In the studies reviewed here, only one reported IVIM parameters derived from a Bayesian fitting approach to obtain f and D* following NLLS fitting of high b-value data to fit for D. 56 A prior study in the pancreas 86 directly compared results between constrained NLLS approaches and Bayesian where they observed improved reliability of results using a Bayesian approach, but noted that it did not significantly out-perform a NLLS approach that used a fixed value for D*. A recent study by Koopman et al 48 evaluated repeatability of IVIM parameters in the tonsils and pterygoid muscles using a variety of fitting methods including NLLS fitting, Bayesian, and a neural network approach. They found that both Bayesian and their neural network approach outperformed NLLS in terms of having lower within-subject coefficient of variations; however, the average IVIM parameter values were not provided.…”

Section: Systematic Reviewmentioning

confidence: 99%

See 2 more Smart Citations

Intravoxel Incoherent Motion Magnetic Resonance Imaging in Skeletal Muscle: Review and Future Directions

Englund

Reiter

Shahidi

et al. 2021

Magnetic Resonance Imaging

View full text Add to dashboard Cite

Throughout the body, muscle structure and function can be interrogated using a variety of noninvasive magnetic resonance imaging (MRI) methods. Recently, intravoxel incoherent motion (IVIM) MRI has gained momentum as a method to evaluate components of blood flow and tissue diffusion simultaneously. Much of the prior research has focused on highly vascularized organs, including the brain, kidney, and liver. Unique aspects of skeletal muscle, including the relatively low perfusion at rest and large dynamic range of perfusion between resting and maximal hyperemic states, may influence the acquisition, postprocessing, and interpretation of IVIM data. Here, we introduce several of those unique features of skeletal muscle; review existing studies of IVIM in skeletal muscle at rest, in response to exercise, and in disease states; and consider possible confounds that should be addressed for musclespecific evaluations. Most studies used segmented nonlinear least squares fitting with a b-value threshold of 200 sec/mm 2 to obtain IVIM parameters of perfusion fraction (f), pseudo-diffusion coefficient (D*), and diffusion coefficient (D). In healthy individuals, across all muscles, the average AE standard deviation of D was 1.46 AE 0.30 Â 10 À3 mm 2 /sec, D* was 29.7 AE 38.1 Â 10 À3 mm 2 /sec, and f was 11.1 AE 6.7%. Comparisons of reported IVIM parameters in muscles of the back, thigh, and leg of healthy individuals showed no significant difference between anatomic locations. Throughout the body, exercise elicited a positive change of all IVIM parameters. Future directions including advanced postprocessing models and potential sequence modifications are discussed. Level of Evidence: 2 Technical Efficacy: Stage 2

show abstract

Section: Systematic Reviewmentioning

confidence: 92%

Section: Ivim Theorymentioning

confidence: 99%

Section: Systematic Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Intravoxel Incoherent Motion Magnetic Resonance Imaging in Skeletal Muscle: Review and Future Directions

Englund

Reiter

Shahidi

et al. 2021

Magnetic Resonance Imaging

View full text Add to dashboard Cite

show abstract

“…The IVIM feature extraction of the perfusion fraction (f), perfusion coefficient (D*) and diffusion coefficient (D) was performed using MATLAB R2019a software [ 17 ], after motion correction, in order to reduce artifacts.…”

Section: Methodsmentioning

confidence: 99%

Early Response Prediction of Multiparametric Functional MRI and 18F-FDG-PET in Patients with Head and Neck Squamous Cell Carcinoma Treated with (Chemo)Radiation

Martens

Koopman

Lavini

et al. 2022

Cancers

Self Cite

View full text Add to dashboard Cite

Background: Patients with locally-advanced head and neck squamous cell carcinoma (HNSCC) have variable responses to (chemo)radiotherapy. A reliable prediction of outcomes allows for enhancing treatment efficacy and follow-up monitoring. Methods: Fifty-seven histopathologically-proven HNSCC patients with curative (chemo)radiotherapy were prospectively included. All patients had an MRI (DW,-IVIM, DCE-MRI) and 18F-FDG-PET/CT before and 10 days after start-treatment (intratreatment). Primary tumor functional imaging parameters were extracted. Univariate and multivariate analysis were performed to construct prognostic models and risk stratification for 2 year locoregional recurrence-free survival (LRFFS), distant metastasis-free survival (DMFS) and overall survival (OS). Model performance was measured by the cross-validated area under the receiver operating characteristic curve (AUC). Results: The best LRFFS model contained the pretreatment imaging parameters ADC_kurtosis, Kep and SUV_peak, and intratreatment imaging parameters change (Δ) Δ-ADC_skewness, Δ-f, Δ-SUV_peak and Δ-total lesion glycolysis (TLG) (AUC = 0.81). Clinical parameters did not enhance LRFFS prediction. The best DMFS model contained pretreatment ADC_kurtosis and SUV_peak (AUC = 0.88). The best OS model contained gender, HPV-status, N-stage, pretreatment ADC_skewness, D, f, metabolic-active tumor volume (MATV), SUV_mean and SUV_peak (AUC = 0.82). Risk stratification in high/medium/low risk was significantly prognostic for LRFFS (p = 0.002), DMFS (p < 0.001) and OS (p = 0.003). Conclusions: Intratreatment functional imaging parameters capture early tumoral changes that only provide prognostic information regarding LRFFS. The best LRFFS model consisted of pretreatment, intratreatment and Δ functional imaging parameters; the DMFS model consisted of only pretreatment functional imaging parameters, and the OS model consisted ofHPV-status, gender and only pretreatment functional imaging parameters. Accurate clinically applicable risk stratification calculators can enable personalized treatment (adaptation) management, early on during treatment, improve counseling and enhance patient-specific post-therapy monitoring.

show abstract

“…To improve parameter map accuracy, the use of deep neural networks to model the biexponential IVIM fit has been proposed [15][16][17][18][19] . In a previous study, a physics-informed unsupervised approach that could train directly to in-vivo MRI data (IVIM-NET) 15,20 was implemented.…”

Section: Introductionmentioning

confidence: 99%

Self-supervised neural network improves tri-exponential intravoxel incoherent motion model fitting in non-alcoholic fatty liver disease

Troelstra

Dijk

Witjes

et al. 2022

Preprint

View full text Add to dashboard Cite

Neural network generated intravoxel incoherent motion (IVIM) parameter maps are visually more detailed, less noisy and have a higher test-retest repeatability than conventional least-squares (LSQ) generated images. Recent literature suggests that tri-exponential models may fit liver IVIM data more accurately than conventional bi-exponential models, warranting investigation into the use of a neural network (NN) to estimate tri-exponential IVIM parameters. Here, we developed a tri-exponential IVIM unsupervised physics-informed deep neural network (IVIM3-NET), assessed its performance in non-alcoholic fatty liver disease (NAFLD) and compared outcomes with bi-exponential LSQ and NN fits and tri-exponential LSQ fits. Scanning was performed using a 3.0T free-breathing multi-slice diffusion-weighted single-shot echo-planar imaging sequence with 18 b-values. Images were analysed for visual quality, comparing the bi- and tri-exponential IVIM models for LSQ fits and NN fits using parameter-map signal-to-noise ratios (SNR) and adjusted R2. IVIM parameters were compared to histological fibrosis, disease activity and steatosis grades. Parameter map quality improved with bi- and tri-exponential NN approaches, with average parameter-map SNR increasing from 3.38 to 5.59 and 2.45 to 4.01 for bi- and tri-exponential LSQ and NN models respectively. In 33 out of 36 patients, the tri-exponential model exhibited higher adjusted R2 values than the bi-exponential model. Correlating IVIM data to liver histology showed that the bi- and tri-exponential NN outperformed both LSQ models for the majority of IVIM parameters (10 out of 15 significant correlations). Overall, our results support the use of a tri-exponential IVIM model in NAFLD. We show that the IVIM3-NET can be used to improve image quality compared to a tri-exponential LSQ fit and provides promising correlations with histopathology.

show abstract

Repeatability of IVIM biomarkers from diffusion‐weighted MRI in head and neck: Bayesian probability versus neural network

Abstract: This is an open access article under the terms of the Creative Commons Attribution NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

Cited by 25 publications

References 29 publications

Intravoxel Incoherent Motion Magnetic Resonance Imaging in Skeletal Muscle: Review and Future Directions

Intravoxel Incoherent Motion Magnetic Resonance Imaging in Skeletal Muscle: Review and Future Directions

Early Response Prediction of Multiparametric Functional MRI and 18F-FDG-PET in Patients with Head and Neck Squamous Cell Carcinoma Treated with (Chemo)Radiation

Self-supervised neural network improves tri-exponential intravoxel incoherent motion model fitting in non-alcoholic fatty liver disease

Contact Info

Product

Resources

About