Generative Invertible Networks (GIN): Pathophysiology-Interpretable Feature Mapping and Virtual Patient Generation

Abstract

Machine learning methods play increasingly important roles in pre-procedural planning for complex surgeries and interventions. Very often, however, researchers find the historical records of emerging surgical techniques, such as the transcatheter aortic valve replacement (TAVR), are highly scarce in quantity. In this paper, we address this challenge by proposing novel generative invertible networks (GIN) to select features and generate high-quality virtual patients that may potentially serve as an additional data source for machine learning. Combining a convolutional neural network (CNN) and generative adversarial networks (GAN), GIN discovers the pathophysiologic meaning of the feature space. Moreover, a test of predicting the surgical outcome directly using the selected features results in a high accuracy of 81.55%, which suggests little pathophysiologic information has been lost while conducting the feature selection. This demonstrates GIN can generate virtual patients not only visually authentic but also pathophysiologically interpretable.

Online detection of cyber‐incidents in additive manufacturing systems via analyzing multimedia signals

“…To quantify the difference between two spectorgrams, we propose to use the Wasserstein‐2 distance as the metric. Wasserstein distance is widely used in economy, 20 machine learning, 21 and geographics 22 . Different from $$L_2$$ metric, Wasserstein metric measures the translation shift in a more robust manner.…”

Active Image Synthesis for Efficient Labeling

“…However, this may not hold in, e.g., medical images. Taking CT scans as an example, different substances of human tissues correspond to different ranges of image intensity, alterations of which may lead to a completely different interpretation of the pathophysiological condition [15]. This significantly limits the augmentation methods that can be used for learning tasks in manufacturing and healthcare.…”

Deep Learning in the Biomedical Applications: Recent and Future Status

“…In [304], a conditional GAN is explored to augment artificially generated lung nodules to improve the robustness of the progressive holistically nested network (P-HNN) model for pathological lung segmentation of CT scans. In [305], a novel generative invertible networks (GIN), which is a combination of a convolutional neural network and generative adversarial networks, is proposed to extract the features from real patients and generate virtual patients, which are both visually and pathophysiologically plausible, using the features. In [306], a deep generative multi-task is developed to solve the problem of limited training data and data with lesion annotations because making the annotations is a very expensive and time consuming task.…”