A database for MR-based electrical properties tomography with in silico brain data-ADEPT.

PURPOSE: Several reconstruction methods for MR-based electrical properties tomography (EPT) have been developed. However, the lack of common data makes it difficult to objectively compare their performances. This is, however, a necessary precursor for standardizing and introducing this technique in the clinical setting. To enable objective comparison of the performances of reconstruction methods and provide common data for their training and testing, we created ADEPT, a database of simulated data for brain MR-EPT reconstructions. METHODS: ADEPT is a database containing in silico data for brain EPT reconstructions. This database was created from 25 different brain models, with and without tumors. Rigid geometric augmentations were applied, and different electrical properties were assigned to white matter, gray matter, CSF, and tumors to generate 120 different brain models. These models were used as input for finite-difference time-domain simulations in Sim4Life, used to compute the electromagnetic fields needed for MR-EPT reconstructions. RESULTS: Electromagnetic fields from 84 healthy and 36 tumor brain models were simulated. The simulated fields relevant for MR-EPT reconstructions (transmit and receive RF fields and transceive phase) and their ground-truth electrical properties are made publicly available through ADEPT. Additionally, nonattainable fields such as the total magnetic field and the electric field are available upon request. CONCLUSION: ADEPT will serve as reference database for objective comparisons of reconstruction methods and will be a first step toward standardization of MR-EPT reconstructions. Furthermore, it provides a large amount of data that can be exploited to train data-driven methods. It can be accessed from  https://doi.org/10.34894/V0HBJ8.

A deep learning method for image-based subject-specific local SAR assessment.

PURPOSE: Local specific absorption rate (SAR) cannot be measured and is usually evaluated by offline numerical simulations using generic body models that of course will differ from the patient's anatomy. An additional safety margin is needed to include this intersubject variability. In this work, we present a deep learning-based method for image-based subject-specific local SAR assessment. We propose to train a convolutional neural network to learn a "surrogate SAR model" to map the relation between subject-specific B1+ maps and the corresponding local SAR. METHOD: Our database of 23 subject-specific models with an 8-transmit channel body array for prostate imaging at 7 T was used to build 5750 training samples. These synthetic complex B1+ maps and local SAR distributions were used to train a conditional generative adversarial network. Extra penalization for local SAR underestimation errors was included in the loss function. In silico and in vivo validation were performed. RESULTS: In silico cross-validation shows a good qualitative and quantitative match between predicted and ground-truth local SAR distributions. The peak local SAR estimation error distribution shows a mean overestimation error of 15% with 13% probability of underestimation. The higher accuracy of the proposed method allows the use of less conservative safety factors compared with standard procedures. In vivo validation shows that the method is applicable with realistic measurement data with impressively good qualitative and quantitative agreement to simulations. CONCLUSION: The proposed deep learning method allows online image-based subject-specific local SAR assessment. It greatly reduces the uncertainty in current state-of-the-art SAR assessment methods, reducing the time in the examination protocol by almost 25%.