ABSTRACT
Purpose: Early-stage glaucoma diagnosis has been a challenging problem in ophthalmology. The current state-of-the-art glaucoma diagnosis techniques do not completely leverage the functional measures' such as electroretinogram's immense potential; instead, focus is on structural measures like optical coherence tomography. The current study aims to take a foundational step toward the development of a novel and reliable predictive framework for early detection of glaucoma using machine-learning-based algorithm capable of leveraging medically relevant information that ERG signals contain. Methods: ERG signals from 60 eyes of DBA/2 mice were grouped for binary classification based on age. The signals were also grouped based on intraocular pressure (IOP) for multiclass classification. Statistical and wavelet-based features were engineered and extracted. Important predictors (ERG tests and features) were determined, and the performance of five machine learning-based methods were evaluated. Results: Random forest (bagged trees) ensemble classifier provided the best performance in both binary and multiclass classification of ERG signals. An accuracy of 91.7 and 80% was achieved for binary and multiclass classification, respectively, suggesting that machine-learning-based models can detect subtle changes in ERG signals if trained using advanced features such as those based on wavelet analyses. Conclusions: The present study describes a novel, machine-learning-based method to analyze ERG signals providing additional information that may be used to detect early-stage glaucoma. Based on promising performance metrics obtained using the proposed machine-learning-based framework leveraging an established ERG data set, we conclude that the novel framework allows for detection of functional deficits of early/various stages of glaucoma in mice.
ABSTRACT
Creation of a submucosal plane to separate the lesion from the deeper muscle layer in gastrointestinal tract is an integral and essential part of endoscopic resection therapies such as endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD). Thereby, an optimized submucosal injection technique is required to ensure a successful process. In this study, the computational fluid dynamics (CFD) technique is employed as a foundational step towards the development of a framework that can provide useful directions to optimize the injection process. Three different lifting agents, including Glycerol, Eleview®, and ORISE® gel have been used for this study. The role of different injection angles, injection dynamics, and effect of temperature are studied to understand the lifting characteristic of each agent. The study shows that Eleview® provides the highest lifting effect, including the initial injection period. To evaluate the impact of the injection process, two cases are simulated, termed static injection and dynamic injection. Under static injection, the injection angle is investigated from lower to higher angles of injection. In the dynamic injection, two cases are modulated, where a continuous change of injection angle from lower to higher degrees (denoted as clockwise) and vice-versa in the anti-clockwise direction are investigated. Increased lifting characteristics are observed at decreasing/lower angle of injection. Further, the correlation between temperature of the lifting agents and their lifting characteristics is investigated.
ABSTRACT
The present study aims to assess the effects of two different underlying RANS models on overall behavior of the IDDES methodology when applied to different flow configurations ranging from fully attached (plane channel flow) to separated flows (periodic hill flow). This includes investigating prediction accuracy of first and second order statistics, response to grid refinement, grey area dynamics and triggering mechanism. Further, several criteria have been investigated to assess reliability and quality of the methodology when operating in scale resolving mode. It turns out that irrespective of the near wall modeling strategy, the IDDES methodology does not satisfy all criteria required to make this methodology reliable when applied to various flow configurations at different Reynolds numbers with different grid resolutions. Further, it is found that using more advanced underlying RANS model to improve prediction accuracy of the near wall dynamics results in extension of the grey area, which may delay the transition to scale resolving mode. This systematic study for attached and separated flows suggests that the shortcomings of IDDES methodology mostly lie in inaccurate prediction of the dynamics inside the grey area and demands further investigation in this direction to make this methodology capable of dealing with different flow situations reliably.
