Jonathan M Scott1, Arvin Arani2, Armando Manduca3, Kiaran P McGee2, Joshua D Trzasko2, John Huston2, Richard L Ehman2, Matthew C Murphy4. 1. Mayo Clinic Medical Scientist Training Program, 200 First Street SW, Rochester 55905, MN, USA. 2. Department of Radiology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester 55905, MN, USA. 3. Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, 200 First Street SW, Rochester 55905, MN, USA. 4. Department of Radiology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester 55905, MN, USA. Electronic address: murphy.matthew@mayo.edu.
Abstract
PURPOSE: To test the hypothesis that removing the assumption of material homogeneity will improve the spatial accuracy of stiffness estimates made by Magnetic Resonance Elastography (MRE). METHODS: An artificial neural network was trained using synthetic wave data computed using a coupled harmonic oscillator model. Material properties were allowed to vary in a piecewise smooth pattern. This neural network inversion (Inhomogeneous Learned Inversion (ILI)) was compared against a previous homogeneous neural network inversion (Homogeneous Learned Inversion (HLI)) and conventional direct inversion (DI) in simulation, phantom, and in-vivo experiments. RESULTS: In simulation experiments, ILI was more accurate than HLI and DI in predicting the stiffness of an inclusion in noise-free, low-noise, and high-noise data. In the phantom experiment, ILI delineated inclusions ≤ 2.25 cm in diameter more clearly than HLI and DI, and provided a higher contrast-to-noise ratio for all inclusions. In a series of stiff brain tumors, ILI shows sharper stiffness transitions at the edges of tumors than the other inversions evaluated. CONCLUSION: ILI is an artificial neural network based framework for MRE inversion that does not assume homogeneity in material stiffness. Preliminary results suggest that it provides more accurate stiffness estimates and better contrast in small inclusions and at large stiffness gradients than existing algorithms that assume local homogeneity. These results support the need for continued exploration of learning-based approaches to MRE inversion, particularly for applications where high resolution is required.
PURPOSE: To test the hypothesis that removing the assumption of material homogeneity will improve the spatial accuracy of stiffness estimates made by Magnetic Resonance Elastography (MRE). METHODS: An artificial neural network was trained using synthetic wave data computed using a coupled harmonic oscillator model. Material properties were allowed to vary in a piecewise smooth pattern. This neural network inversion (Inhomogeneous Learned Inversion (ILI)) was compared against a previous homogeneous neural network inversion (Homogeneous Learned Inversion (HLI)) and conventional direct inversion (DI) in simulation, phantom, and in-vivo experiments. RESULTS: In simulation experiments, ILI was more accurate than HLI and DI in predicting the stiffness of an inclusion in noise-free, low-noise, and high-noise data. In the phantom experiment, ILI delineated inclusions ≤ 2.25 cm in diameter more clearly than HLI and DI, and provided a higher contrast-to-noise ratio for all inclusions. In a series of stiff brain tumors, ILI shows sharper stiffness transitions at the edges of tumors than the other inversions evaluated. CONCLUSION: ILI is an artificial neural network based framework for MRE inversion that does not assume homogeneity in material stiffness. Preliminary results suggest that it provides more accurate stiffness estimates and better contrast in small inclusions and at large stiffness gradients than existing algorithms that assume local homogeneity. These results support the need for continued exploration of learning-based approaches to MRE inversion, particularly for applications where high resolution is required.
Authors: Marvin M Doyley; John B Weaver; Elijah E W Van Houten; Francis E Kennedy; Keith D Paulsen Journal: Med Phys Date: 2003-04 Impact factor: 4.071
Authors: M D J McGarry; E E W Van Houten; C L Johnson; J G Georgiadis; B P Sutton; J B Weaver; K D Paulsen Journal: Med Phys Date: 2012-10 Impact factor: 4.071
Authors: Meng Yin; Kevin J Glaser; Jayant A Talwalkar; Jun Chen; Armando Manduca; Richard L Ehman Journal: Radiology Date: 2015-07-08 Impact factor: 11.105
Authors: Laurent Huwart; Christine Sempoux; Najat Salameh; Jacques Jamart; Laurence Annet; Ralph Sinkus; Frank Peeters; Leon C ter Beek; Yves Horsmans; Bernard E Van Beers Journal: Radiology Date: 2007-11 Impact factor: 11.105
Authors: Curtis L Johnson; Matthew D J McGarry; Armen A Gharibans; John B Weaver; Keith D Paulsen; Huan Wang; William C Olivero; Bradley P Sutton; John G Georgiadis Journal: Neuroimage Date: 2013-05-01 Impact factor: 6.556
Authors: Mark Jenkinson; Christian F Beckmann; Timothy E J Behrens; Mark W Woolrich; Stephen M Smith Journal: Neuroimage Date: 2011-09-16 Impact factor: 6.556
Authors: Hillary Schwarb; Curtis L Johnson; Ana M Daugherty; Charles H Hillman; Arthur F Kramer; Neal J Cohen; Aron K Barbey Journal: Neuroimage Date: 2017-03-30 Impact factor: 6.556
Authors: Grace McIlvain; Alexander M Cerjanic; Anthony G Christodoulou; Matthew D J McGarry; Curtis L Johnson Journal: Magn Reson Med Date: 2022-06-01 Impact factor: 3.737
Authors: Jonathan M Scott; KowsalyaDevi Pavuluri; Joshua D Trzasko; Armando Manduca; Matthew L Senjem; John Huston; Richard L Ehman; Matthew C Murphy Journal: Magn Reson Med Date: 2022-04-05 Impact factor: 3.737