4 March 2020 in European Heart Journal
Providing therapies tailored to each patient is the vision of precision medicine, enabled by the increasing ability to capture extensive data about individual patients. In this position paper, we argue that the second enabling pillar towards this vision is the increasing power of computers and algorithms to learn, reason, and build the ‘digital twin’ of a patient. Computational models are boosting the capacity to draw diagnosis and prognosis, and future treatments will be tailored not only to current health status and data, but also to an accurate projection of the pathways to restore health by model predictions. The early steps of the digital twin in the area of cardiovascular medicine are reviewed in this article, together with a discussion of the challenges and opportunities ahead. We emphasize the synergies between mechanistic and statistical models in accelerating cardiovascular research and enabling the vision of precision medicine.
14 January 2021 in IEEE Transactions on Medical Imaging
Deep learning can bring time savings and increased reproducibility to medical image analysis. However, acquiring training data is challenging due to the time-intensive nature of labeling and high inter-observer variability in annotations. Rather than labeling images, in this work we propose an alternative pipeline where images are generated from existing high-quality annotations using generative adversarial networks (GANs). Annotations are derived automatically from previously built anatomical models and are transformed into realistic synthetic ultrasound images with paired labels using a CycleGAN. We demonstrate the pipeline by generating synthetic 2D echocardiography images to compare with existing deep learning ultrasound segmentation datasets. A convolutional neural network is trained to segment the left ventricle and left atrium using only synthetic images. Networks trained with synthetic images were extensively tested on four different unseen datasets of real images with median Dice scores of 91, 90, 88, and 87 for left ventricle segmentation. These results match or are better than inter-observer results measured on real ultrasound datasets and are comparable to a network trained on a separate set of real images. Results demonstrate the images produced can effectively be used in place of real data for training. The proposed pipeline opens the door for automatic generation of training data for many tasks in medical imaging as the same process can be applied to other segmentation or landmark detection tasks in any modality. The source code and anatomical models are available to other researchers. 1 1https://adgilbert.github.io/data-generation/
25 January 2018 in IEEE Transactions on Biomedical Engineering
JBJean-Marie C. Bouteiller
+2
Objective: The ideal form of a neural-interfacing device is highly dependent upon the anatomy of the region with which it is meant to interface. Multiple-electrode arrays provide a system that can be adapted to various neural geometries. Computational models of stimulating systems have proven useful for evaluating electrode placement and stimulation protocols, but have yet to be adequately adapted to the unique features of the hippocampus. Methods: As an approach to understanding potential memory restorative devices, an admittance method-NEURON model was constructed to predict the direct and synaptic response of a region of the rat dentate gyrus to electrical stimulation of the perforant path. Results: A validation of estimated local field potentials against experimental recordings is performed and results of a bilinear electrode placement and stimulation amplitude parameter search are presented. Conclusion: The parametric analysis presented herein suggests that stimulating electrodes placed between the lateral and medial perforant path, near the crest of the dentate gyrus, yield a larger relative population response to given stimuli. Significance: Beyond deepening understanding of the hippocampal tissue system, establishment of this model provides a method to evaluate candidate stimulating devices and protocols.
1 June 2021 in IEEE Journal of Biomedical and Health Informatics
Spectral Doppler measurements are an important part of the standard echocardiographic examination. These measurements give insight into myocardial motion and blood flow providing clinicians with parameters for diagnostic decision making. Many of these measurements are performed automatically with high accuracy, increasing the efficiency of the diagnostic pipeline. However, full automation is not yet available because the user must manually select which measurement should be performed on each image. In this work, we develop a pipeline based on convolutional neural networks (CNNs) to automatically classify the measurement type from cardiac Doppler scans. We show how the multi-modal information in each spectral Doppler recording can be combined using a meta parameter post-processing mapping scheme and heatmaps to encode coordinate locations. Additionally, we experiment with several architectures to examine the tradeoff between accuracy, speed, and memory usage for resource-constrained environments. Finally, we propose a confidence metric using the values in the last fully connected layer of the network and show that our confidence metric can prevent many misclassifications. Our algorithm enables a fully automatic pipeline from acquisition to Doppler spectrum measurements. We achieve 96% accuracy on a test set drawn from separate clinical sites, indicating that the proposed method is suitable for clinical adoption.
1 August 2016 in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
Owing to the dramatic rise in treatment of neurological disorders with electrical micro-stimulation it has become apparent that the major technological limitation in deploying effective devices lies in the process of designing efficient, safe, and outcome specific electrode arrays. The time-consuming and low-fidelity nature of gathering test data using experimental means and the immense control and flexibility of computational models, has prompted us and others to build models of electrical stimulation of neural networks that can be simulated in a computer. Because prior work has been focused on single cells, very small networks, or non-biological models of neural tissue, it was expedient that we take advantage of our, 4,040 processor, computing cluster to construct a large-scale 3-dimensional emulation of hippocampal tissue using detailed neuronal models with explicit and unique morphologies. This model, when paired with an equivalent circuit method of estimating voltage signal attenuation throughout anisotropic resistive tissue, can be used to predict tissue response to an exhaustive set of stimulation and tissue conditions: electrode geometry, array geometry, static dielectric properties of tissue, stimulation pulse features, etc. Preliminary experiments demonstrate that this system is capable of yielding neuronal responses with striking similarities to experimental results. This work provides an avenue to qualitative evaluation of electrode arrays, and more meaningful modeling of local field potentials in terms of their contributing sources and sinks.