-
Dynamic memory to alleviate catastrophic forgetting in continuous learning settings
Authors:
Johannes Hofmanninger,
Matthias Perkonigg,
James A. Brink,
Oleg Pianykh,
Christian Herold,
Georg Langs
Abstract:
In medical imaging, technical progress or changes in diagnostic procedures lead to a continuous change in image appearance. Scanner manufacturer, reconstruction kernel, dose, other protocol specific settings or administering of contrast agents are examples that influence image content independent of the scanned biology. Such domain and task shifts limit the applicability of machine learning algori…
▽ More
In medical imaging, technical progress or changes in diagnostic procedures lead to a continuous change in image appearance. Scanner manufacturer, reconstruction kernel, dose, other protocol specific settings or administering of contrast agents are examples that influence image content independent of the scanned biology. Such domain and task shifts limit the applicability of machine learning algorithms in the clinical routine by rendering models obsolete over time. Here, we address the problem of data shifts in a continuous learning scenario by adapting a model to unseen variations in the source domain while counteracting catastrophic forgetting effects. Our method uses a dynamic memory to facilitate rehearsal of a diverse training data subset to mitigate forgetting. We evaluated our approach on routine clinical CT data obtained with two different scanner protocols and synthetic classification tasks. Experiments show that dynamic memory counters catastrophic forgetting in a setting with multiple data shifts without the necessity for explicit knowledge about when these shifts occur.
△ Less
Submitted 7 July, 2020; v1 submitted 6 July, 2020;
originally announced July 2020.
-
Improving digital signal interpolation: L2-optimal kernels with kernel-invariant interpolation speed
Authors:
Oleg S. Pianykh
Abstract:
Interpolation is responsible for digital signal resampling and can significantly degrade the original signal quality if not done properly. For many years, optimal interpolation algorithms were sought within constrained classes of interpolation kernel functions. We derive a new family of unconstrained L2-optimal interpolation kernels, and compare their properties to the previously known. Although d…
▽ More
Interpolation is responsible for digital signal resampling and can significantly degrade the original signal quality if not done properly. For many years, optimal interpolation algorithms were sought within constrained classes of interpolation kernel functions. We derive a new family of unconstrained L2-optimal interpolation kernels, and compare their properties to the previously known. Although digital images are used to illustrate this work, our L2-optimal kernels can be applied to interpolate any digital signals.
△ Less
Submitted 21 April, 2011;
originally announced April 2011.
-
Bilateral filters: what they can and cannot do
Authors:
Oleg S. Pianykh
Abstract:
Nonlinear bilateral filters (BF) deliver a fine blend of computational simplicity and blur-free denoising. However, little is known about their nature, noise-suppressing properties, and optimal choices of filter parameters. Our study is meant to fill this gap-explaining the underlying mechanism of bilateral filtering and providing the methodology for optimal filter selection. Practical application…
▽ More
Nonlinear bilateral filters (BF) deliver a fine blend of computational simplicity and blur-free denoising. However, little is known about their nature, noise-suppressing properties, and optimal choices of filter parameters. Our study is meant to fill this gap-explaining the underlying mechanism of bilateral filtering and providing the methodology for optimal filter selection. Practical application to CT image denoising is discussed to illustrate our results.
△ Less
Submitted 6 July, 2010;
originally announced July 2010.
-
L2-optimal image interpolation and its applications to medical imaging
Authors:
Oleg Pianykh
Abstract:
Digital medical images are always displayed scaled to fit particular view. Interpolation is responsible for this scaling, and if not done properly, can significantly degrade diagnostic image quality. However, theoretically-optimal interpolation algorithms may also be the most time-consuming and impractical. We propose a new approach, adapted to the needs of digital medical imaging, to combine high…
▽ More
Digital medical images are always displayed scaled to fit particular view. Interpolation is responsible for this scaling, and if not done properly, can significantly degrade diagnostic image quality. However, theoretically-optimal interpolation algorithms may also be the most time-consuming and impractical. We propose a new approach, adapted to the needs of digital medical imaging, to combine high interpolation speed and superior L2-optimal image quality.
△ Less
Submitted 11 June, 2010;
originally announced June 2010.
-
Perfusion Linearity and Its Applications
Authors:
Oleg Pianykh
Abstract:
Perfusion analysis computes blood flow parameters (blood volume, blood flow, mean transit time) from the observed flow of contrast agent, passing through the patient's vascular system. Perfusion deconvolution has been widely accepted as the principal numerical tool for perfusion analysis, and is used routinely in clinical applications. This extensive use of perfusion in clinical decision-making ma…
▽ More
Perfusion analysis computes blood flow parameters (blood volume, blood flow, mean transit time) from the observed flow of contrast agent, passing through the patient's vascular system. Perfusion deconvolution has been widely accepted as the principal numerical tool for perfusion analysis, and is used routinely in clinical applications. This extensive use of perfusion in clinical decision-making makes numerical stability and robustness of perfusion computations vital for accurate diagnostics and patient safety. The main goal of this paper is to propose a novel approach for validating numerical properties of perfusion algorithms. The approach is based on Perfusion Linearity Property (PLP), which we find in perfusion deconvolution, as well as in many other perfusion techniques. PLP allows one to study perfusion values as weighted averages of the original imaging data. This, in turn, uncovers hidden problems with the existing deconvolution techniques, and may be used to suggest more reliable computational approaches and methodology.
△ Less
Submitted 1 June, 2010;
originally announced June 2010.
