Subtle Data Crimes: Naively training machine learning algorithms could lead to overly-optimistic results

Shimron, Efrat; Tamir, Jonathan I.; Wang, Ke; Lustig, Michael

doi:10.48550/arxiv.2109.08237

Cited by 4 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar sentiments have been expressed in the case of microscopy [18], fluorescence microscopy [19], PET (Position Emission Tomography) [20] and computed tomography [21,22]. See also [23][24][25][26][27][28] for further discussion and related issues. See also [14] for some theoretical analysis of instabilities and hallucinations in deep learning.…”

Section: Issues With Deep Learning For Inverse Problemsmentioning

confidence: 67%

NESTANets: Stable, accurate and efficient neural networks for analysis-sparse inverse problems

Neyra-Nesterenko¹,

Adcock²

2022

Preprint

View full text Add to dashboard Cite

Solving inverse problems is a fundamental component of science, engineering and mathematics. With the advent of deep learning, deep neural networks have significant potential to outperform existing state-of-theart, model-based methods for solving inverse problems. However, it is known that current data-driven approaches face several key issues, notably instabilities and hallucinations, with potential impact in critical tasks such as medical imaging. This raises the key question of whether or not one can construct stable and accurate deep neural networks for inverse problems. In this work, we present a novel construction of an accurate, stable and efficient neural network for inverse problems with general analysis-sparse models. To construct the network, we unroll NESTA, an accelerated first-order method for convex optimization. Combined with a compressed sensing analysis, we prove accuracy and stability. Finally, a restart scheme is employed to enable exponential decay of the required network depth, yielding a shallower, and consequently more efficient, network. We showcase this approach in the case of Fourier imaging, and verify its stability and performance via a series of numerical experiments. The key impact of this work is to provide theoretical guarantees for computing and developing stable neural networks in practice.

show abstract

Section: Issues With Deep Learning For Inverse Problemsmentioning

confidence: 67%

NESTANets: Stable, accurate and efficient neural networks for analysis-sparse inverse problems

Neyra-Nesterenko¹,

Adcock²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This is important to consider as FSE is the backbone of clinical MRI and most publicly available raw MRI datasets are all FSE based. This bias is strongly related to other implicit data crimes arising from the misuse of public data for MRI reconstruction [20]. Fortunately, this problem is easily mitigated by accounting for the data ordering [15].…”

Section: Resultsmentioning

confidence: 99%

FSE Compensated Motion Correction for MRI Using Data Driven Methods

Levac¹,

Kumar²,

Kardonik³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Magnetic Resonance Imaging (MRI) is a widely used medical imaging modality boasting great soft tissue contrast without ionizing radiation, but unfortunately suffers from long acquisition times. Long scan times can lead to motion artifacts, for example due to bulk patient motion such as head movement and periodic motion produced by the heart or lungs. Motion artifacts can degrade image quality and in some cases render the scans nondiagnostic. To combat this problem, prospective and retrospective motion correction techniques have been introduced. More recently, data driven methods using deep neural networks have been proposed. As a large number of publicly available MRI datasets are based on Fast Spin Echo (FSE) sequences, methods that use them for training should incorporate the correct FSE acquisition dynamics. Unfortunately, when simulating training data, many approaches fail to generate accurate motion-corrupt images by neglecting the effects of the temporal ordering of the k-space lines as well as neglecting the signal decay throughout the FSE echo train. In this work, we highlight this consequence and demonstrate a training method which correctly simulates the data acquisition process of FSE sequences with higher fidelity by including sample ordering and signal decay dynamics. Through numerical experiments, we show that accounting for the FSE acquisition leads to better motion correction performance during inference.

show abstract

“…Although neural networks have the ability to produce high-quality reconstructions, their usage for this task has been shown to sometimes lead to the appearance of spurious image content from the fully-sampled reference images the models have been originally trained on (Hammernik et al, 2021;Muckley et al, 2020;Shimron et al, 2021). DP could counteract such hallucination as it is designed to limit the effect of individual training examples on model training.…”

Section: Mri Reconstructionmentioning

confidence: 99%

Complex-valued deep learning with differential privacy

Ziller,

Usynin,

Knolle

et al. 2021

Preprint

View full text Add to dashboard Cite

We present ζ-DP, an extension of differential privacy (DP) to complex-valued functions. After introducing the complex Gaussian mechanism, whose properties we characterise in terms of pε, δq-DP and Rényi-DP, we present ζ-DP stochastic gradient descent (ζ-DP-SGD), a variant of DP-SGD for training complex-valued neural networks. We experimentally evaluate ζ-DP-SGD on three complexvalued tasks, i.e. electrocardiogram classification, speech classification and magnetic resonance imaging (MRI) reconstruction. Moreover, we provide ζ-DP-SGD benchmarks for a large variety of complex-valued activation functions and on a complex-valued variant of the MNIST dataset. Our experiments demonstrate that DP training of complex-valued neural networks is possible with rigorous privacy guarantees and excellent utility.

show abstract

Subtle Data Crimes: Naively training machine learning algorithms could lead to overly-optimistic results

Cited by 4 publications

References 0 publications

NESTANets: Stable, accurate and efficient neural networks for analysis-sparse inverse problems

NESTANets: Stable, accurate and efficient neural networks for analysis-sparse inverse problems

FSE Compensated Motion Correction for MRI Using Data Driven Methods

Complex-valued deep learning with differential privacy

Contact Info

Product

Resources

About