Precision medicine in cancer treatment increasingly relies on advanced radiotherapies, such as proton beam radiotherapy, to enhance efficacy of the treatment. When the proton beam in this treatment interacts with patient matter, the excited nuclei may emit prompt gamma ray interactions that can be captured by a Compton camera. The image reconstruction from this captured data faces the issue of mischaracterizing the sequences of incoming scattering events, leading to excessive background noise. To address this problem, several machine learning models such as Feedfoward Neural Networks (FNN) and Recurrent Neural Networks (RNN) were developed in PyTorch to properly characterize the scattering sequences on simulated datasets, including newly-created patient medium data, which were generated by using a pipeline comprised of the GEANT4 and Monte-Carlo Detector Effects (MCDE) softwares. These models were implemented using the novel ‘Big-data REU Integrated Development and Experimentation’ (BRIDE) platform, a modular pipeline that streamlines preprocessing, feature engineering, and model development and evaluation on parallelized GPU processors. Hyperparameter studies were done on the novel patient data as well as on water phantom datasets used during previous research. Patient data was more difficult than water phantom data to classify for both FNN and RNN models. FNN models had higher accuracy on patient medium data but lower accuracy on water phantom data when compared to RNN models. Previous results on several different datasets were reproduced on BRIDE and multiple new models achieved greater performance than in previous research.

Proton beam therapy is a unique form of radiotherapy that utilizes protons to treat cancer by irradiating cancerous tumors, while avoiding unnecessary radiation exposure to surrounding healthy tissues. Real-time imaging of the proton beam can make this form of therapy more precise and safer for the patient during delivery. The use of Compton cameras is one proposed method for the real-time imaging of prompt gamma rays that are emitted by the proton beams as they travel through a patient’s body. Unfortunately, some of the Compton camera data is flawed and the reconstruction algorithm yields noisy and insufficiently detailed images to evaluate the proton delivery for the patient. Previous work used a deep residual fully connected neural network. The use of recurrent neural networks (RNNs) has been proposed, since they use recurrence relationships to make potentially better predictions. In this work, RNN architectures using two different recurrent layers are tested, the LSTM and the GRU. Although the deep residual fully connected neural network achieves over 75% testing accuracy and our models achieve only over 73% testing accuracy, the simplicity of our RNN models containing only 6 hidden layers as opposed to 512 is a significant advantage. Importantly in a clinical setting, the time to load the model from disk is significantly faster, potentially enabling the use of Compton camera image reconstruction in real-time during patient treatment.

Proton therapy is a unique form of radiotherapy that utilizes protons to treat cancer by irradiating cancerous tumors while avoiding unnecessary radiation exposure to surrounding healthy tissues. Real-time imaging of prompt gamma rays can be used as a tool to make this form of therapy more effective. The use of Compton cameras is one proposed method for the real-time imaging of prompt gamma rays that are emitted by the proton beams as they travel through a patient’s body. The non-zero time resolution of the Compton camera, during which all interactions are recorded as occurring simultaneously, causes the reconstructed images to be noisy and insufficiently detailed to evaluate the proton delivery for the patient. Deep Learning has been a promising method used to remove and correct the different problems existing within the Compton Camera’s data. Previous papers have demonstrated the effectiveness of using deep fully connected networks to correct improperly ordered gamma interactions within the data. We do a moderately large hyperparameter grid search to find a promising set which yields competitive performance but contains fewer neurons making it compact. The studies which have many neurons, many layers, and a non-zero dropout rate have the best testing accuracy. These many neuron and many layer networks still have significantly fewer total neurons than the current neural network implementation. If given considerably more training time these compact networks could yield equal, if not superior, testing accuracy when compared to larger networks. More improvements are still needed for clinical use and we are currently experimenting with recurrent neural networks to test the viability of this type of architecture for this application.

Predicting violent storms and dangerous weather conditions with current models can take a long time due to the immense complexity associated with weather simulation. Machine learning has the potential to classify tornadic weather patterns much more rapidly, thus allowing for more timely alerts to the public. To deal with class imbalance challenges in machine learning, different data augmentation approaches have been proposed. In this work, we examine the wall time difference between live data augmentation methods versus the use of preaugmented data when they are used in a convolutional neural network based training for tornado prediction. We also compare CPU and GPU based training over varying sizes of augmented data sets. Additionally we examine what impact varying the number of GPUs used for training will produce given a convolutional neural network.