Current methods for patient-specific voxel-level dosimetry in radionuclide therapy suffer from a trade-off between accuracy and computational efficiency. Monte Carlo (MC) radiation transport algorithms are considered the gold standard for voxel-level dosimetry but can be computationally expensive, whereas faster dose voxel kernel (DVK) convolution can be sub-optimal in the presence of tissue heterogeneities. Furthermore, the accuracies of both these methods are limited by the spatial resolution of the reconstructed emission image. To overcome these limitations, this paper considers a single deep convolutional neural network (CNN) with residual learning (named DblurDoseNet) that learns to produce dose-rate maps while compensating for the limited resolution of SPECT images.
We took the novel approach of constructing a convolutional neural network with residual learning to handle the accuracy-efficiency tradeoff while compensating for the limited resolution of SPECT images. We then test our CNN on clinically relevant phantoms and patients undergoing Lu-177 DOTATATE therapy in our clinic. Our network demonstrated superior results than Monte Carlo, the current gold standard for voxel dosimetry, but only takes a fraction of time. Thus, the DblurDoseNet has the potential for real-time patient-specific dosimetry in clinical treatment planning due to its demonstrated improvement in accuracy, resolution, noise and speed over the DVK/MC approaches.
Matlab is needed to access the phantoms and Python (with Numpy package installed) is needed to access the DVKs.
"DblurDoseNet: A Deep Residual Learning Network for Voxel Radionuclide Dosimetry Compensating for SPECT Imaging Resolution" by Zongyu Li, Jeffrey A. Fessler, Justin K. Mikell, Scott J. Wilderman and Yuni K. Dewaraja. Accepted by Medical Physics, 2021. DOI: 10.1002/mp.15397
