Self-Supervised Poisson-Gaussian Denoising.
Ontology highlight
ABSTRACT:
SUBMITTER: Khademi W
PROVIDER: S-EPMC8294668 | biostudies-literature | 2021 Jan
REPOSITORIES: biostudies-literature
Publications
Self-Supervised Poisson-Gaussian Denoising.
Khademi Wesley W Rao Sonia S Minnerath Clare C Hagen Guy G Ventura Jonathan J
IEEE Winter Conference on Applications of Computer Vision. IEEE Winter Conference on Applications of Computer Vision 20210101
We extend the blindspot model for self-supervised denoising to handle Poisson-Gaussian noise and introduce an improved training scheme that avoids hyperparameters and adapts the denoiser to the test data. Self-supervised models for denoising learn to denoise from only noisy data and do not require corresponding clean images, which are difficult or impossible to acquire in some application areas of interest such as low-light microscopy. We introduce a new training strategy to handle Poisson-Gauss ...[more]
Similar Datasets
Project description:By providing three-dimensional visualization of tissues and instruments at high resolution, live volumetric optical coherence tomography (4D-OCT) has the potential to revolutionize ophthalmic surgery. However, the necessary imaging speed is accompanied by increased noise levels. A high data rate and the requirement for minimal latency impose major limitations for real-time noise reduction. In this work, we propose a low complexity neural network for denoising, directly incorporated into the image reconstruction pipeline of a microscope-integrated 4D-OCT prototype with an A-scan rate of 1.2 MHz. For this purpose, we trained a blind-spot network on unpaired OCT images using a self-supervised learning approach. With an optimized U-Net, only a few milliseconds of additional latency were introduced. Simultaneously, these architectural adaptations improved the numerical denoising performance compared to the basic setup, outperforming non-local filtering algorithms. Layers and edges of anatomical structures in B-scans were better preserved than with Gaussian filtering despite comparable processing time. By comparing scenes with and without denoising employed, we show that neural networks can be used to improve visual appearance of volumetric renderings in real time. Enhancing the rendering quality is an important step for the clinical acceptance and translation of 4D-OCT as an intra-surgical guidance tool.
| S-EPMC10082772 | biostudies-literature
Project description:Fluorescence imaging with high signal-to-noise ratios has become the foundation of accurate visualization and analysis of biological phenomena. However, the inevitable noise poses a formidable challenge to imaging sensitivity. Here we provide the spatial redundancy denoising transformer (SRDTrans) to remove noise from fluorescence images in a self-supervised manner. First, a sampling strategy based on spatial redundancy is proposed to extract adjacent orthogonal training pairs, which eliminates the dependence on high imaging speed. Second, we designed a lightweight spatiotemporal transformer architecture to capture long-range dependencies and high-resolution features at low computational cost. SRDTrans can restore high-frequency information without producing oversmoothed structures and distorted fluorescence traces. Finally, we demonstrate the state-of-the-art denoising performance of SRDTrans on single-molecule localization microscopy and two-photon volumetric calcium imaging. SRDTrans does not contain any assumptions about the imaging process and the sample, thus can be easily extended to various imaging modalities and biological applications.
| S-EPMC10766531 | biostudies-literature
Project description:Voltage imaging is a powerful technique for studying neuronal activity, but its effectiveness is often constrained by low signal-to-noise ratios (SNR). Traditional denoising methods, such as matrix factorization, impose rigid assumptions about noise and signal structures, while existing deep learning approaches fail to fully capture the rapid dynamics and complex dependencies inherent in voltage imaging data. Here, we introduce CellMincer, a novel self-supervised deep learning method specifically developed for denoising voltage imaging datasets. CellMincer operates by masking and predicting sparse pixel sets across short temporal windows and conditions the denoiser on precomputed spatiotemporal auto-correlations to effectively model long-range dependencies without large temporal contexts. We developed and utilized a physics-based simulation framework to generate realistic synthetic datasets, enabling rigorous hyperparameter optimization and ablation studies. This approach highlighted the critical role of conditioning on spatiotemporal auto-correlations, resulting in an additional 3-fold SNR gain. Comprehensive benchmarking on both simulated and real datasets, including those validated with patch-clamp electrophysiology (EP), demonstrates CellMincer's state-of-the-art performance, with substantial noise reduction across the frequency spectrum, enhanced subthreshold event detection, and high-fidelity recovery of EP signals. CellMincer consistently outperforms existing methods in SNR gain (0.5-2.9 dB) and reduces SNR variability by 17-55%. Incorporating CellMincer into standard workflows significantly improves neuronal segmentation, peak detection, and functional phenotype identification, consistently surpassing current methods in both SNR gain and consistency.
| S-EPMC11618097 | biostudies-literature
Project description:Diffusion tensor magnetic resonance imaging (DTI) is a widely adopted neuroimaging method for the in vivo mapping of brain tissue microstructure and white matter tracts. Nonetheless, the noise in the diffusion-weighted images (DWIs) decreases the accuracy and precision of DTI derived microstructural parameters and leads to prolonged acquisition time for achieving improved signal-to-noise ratio (SNR). Deep learning-based image denoising using convolutional neural networks (CNNs) has superior performance but often requires additional high-SNR data for supervising the training of CNNs, which reduces the feasibility of supervised learning-based denoising in practice. In this work, we develop a self-supervised deep learning-based method entitled "SDnDTI" for denoising DTI data, which does not require additional high-SNR data for training. Specifically, SDnDTI divides multi-directional DTI data into many subsets of six DWI volumes and transforms DWIs from each subset to along the same diffusion-encoding directions through the diffusion tensor model, generating multiple repetitions of DWIs with identical image contrasts but different noise observations. SDnDTI removes noise by first denoising each repetition of DWIs using a deep 3-dimensional CNN with the average of all repetitions with higher SNR as the training target, following the same approach as normal supervised learning based denoising methods, and then averaging CNN-denoised images for achieving higher SNR. The denoising efficacy of SDnDTI is demonstrated in terms of the similarity of output images and resultant DTI metrics compared to the ground truth generated using substantially more DWI volumes on two datasets with different spatial resolutions, b-values and numbers of input DWI volumes provided by the Human Connectome Project (HCP) and the Lifespan HCP in Aging. The SDnDTI results preserve image sharpness and textural details and substantially improve upon those from the raw data. The results of SDnDTI are comparable to those from supervised learning-based denoising and outperform those from state-of-the-art conventional denoising algorithms including BM4D, AONLM and MPPCA. By leveraging domain knowledge of diffusion MRI physics, SDnDTI makes it easier to use CNN-based denoising methods in practice and has the potential to benefit a wider range of research and clinical applications that require accelerated DTI acquisition and high-quality DTI data for mapping of tissue microstructure, fiber tracts and structural connectivity in the living human brain.
| S-EPMC9511973 | biostudies-literature
Project description:Cryogenic electron tomography (cryo-ET) has rapidly advanced as a high-resolution imaging tool for visualizing subcellular structures in 3D with molecular detail. Direct image inspection remains challenging due to inherent low signal-to-noise ratios (SNR). We introduce CryoSamba, a self-supervised deep learning-based model designed for denoising cryo-ET images. CryoSamba enhances single consecutive 2D planes in tomograms by averaging motion-compensated nearby planes through deep learning interpolation, effectively mimicking increased exposure. This approach amplifies coherent signals and reduces high-frequency noise, substantially improving tomogram contrast and SNR. CryoSamba operates on 3D volumes without needing pre-recorded images, synthetic data, labels or annotations, noise models, or paired volumes. CryoSamba suppresses high-frequency information less aggressively than do existing cryo-ET denoising methods, while retaining real information, as shown both by visual inspection and by Fourier shell correlation analysis of icosahedrally symmetric virus particles. Thus, CryoSamba enhances the analytical pipeline for direct 3D tomogram visual interpretation.
| S-EPMC11276013 | biostudies-literature
Project description:Cryogenic electron microscopy (cryo-EM) is a pivotal technique for imaging macromolecular structures. However, despite extensive processing of large image sets collected in cryo-EM experiments to amplify the signal-to-noise ratio, the reconstructed 3D protein-density maps are often limited in quality due to residual noise, which in turn affects the accuracy of the macromolecular representation. Here, crefDenoiser is introduced, a denoising neural network model designed to enhance the signal in 3D cryo-EM maps produced with standard processing pipelines. The crefDenoiser model is trained without the need for `clean' ground-truth target maps. Instead, a custom dataset is employed, composed of real noisy protein half-maps sourced from the Electron Microscopy Data Bank repository. Competing with the current state-of-the-art, crefDenoiser is designed to optimize for the theoretical noise-free map during self-supervised training. We demonstrate that our model successfully amplifies the signal across a wide variety of protein maps, outperforming a classic map denoiser and following a network-based sharpening model. Without biasing the map, the proposed denoising method leads to improved visibility of protein structural features, including protein domains, secondary structure elements and modest high-resolution feature restoration.
| S-EPMC11364040 | biostudies-literature
Project description:SignificanceVoltage imaging is a powerful tool for studying the dynamics of neuronal activities in the brain. However, voltage imaging data are fundamentally corrupted by severe Poisson noise in the low-photon regime, which hinders the accurate extraction of neuronal activities. Self-supervised deep learning denoising methods have shown great potential in addressing the challenges in low-photon voltage imaging without the need for ground-truth but usually suffer from the trade-off between spatial and temporal performances.AimWe present DeepVID v2, a self-supervised denoising framework with decoupled spatial and temporal enhancement capability to significantly augment low-photon voltage imaging.ApproachDeepVID v2 is built on our original DeepVID framework, which performs frame-based denoising by utilizing a sequence of frames around the central frame targeted for denoising to leverage temporal information and ensure consistency. Similar to DeepVID, the network further integrates multiple blind pixels in the central frame to enrich the learning of local spatial information. In addition, DeepVID v2 introduces a new spatial prior extraction branch to capture fine structural details to learn high spatial resolution information. Two variants of DeepVID v2 are introduced to meet specific denoising needs: an online version tailored for real-time inference with a limited number of frames and an offline version designed to leverage the full dataset, achieving optimal temporal and spatial performances.ResultsWe demonstrate that DeepVID v2 is able to overcome the trade-off between spatial and temporal performances and achieve superior denoising capability in resolving both high-resolution spatial structures and rapid temporal neuronal activities. We further show that DeepVID v2 can generalize to different imaging conditions, including time-series measurements with various signal-to-noise ratios and extreme low-photon conditions.ConclusionsOur results underscore DeepVID v2 as a promising tool for enhancing voltage imaging. This framework has the potential to generalize to other low-photon imaging modalities and greatly facilitate the study of neuronal activities in the brain.
| S-EPMC11519979 | biostudies-literature
Project description:Volumetric functional imaging is widely used for recording neuron activities in vivo, but there exist tradeoffs between the quality of the extracted calcium traces, imaging speed, and laser power. While deep-learning methods have recently been applied to denoise images, their applications to downstream analyses, such as recovering high-SNR calcium traces, have been limited. Further, these methods require temporally-sequential pre-registered data acquired at ultrafast rates. Here, we demonstrate a supervised deep-denoising method to circumvent these tradeoffs for several applications, including whole-brain imaging, large-field-of-view imaging in freely moving animals, and recovering complex neurite structures in C. elegans. Our framework has 30× smaller memory footprint, and is fast in training and inference (50-70 ms); it is highly accurate and generalizable, and further, trained with only small, non-temporally-sequential, independently-acquired training datasets (∼500 pairs of images). We envision that the framework will enable faster and long-term imaging experiments necessary to study neuronal mechanisms of many behaviors.
| S-EPMC9440141 | biostudies-literature
Project description:BackgroundLow-dose computed tomography (LDCT) is a diagnostic imaging technique designed to minimize radiation exposure to the patient. However, this reduction in radiation may compromise computed tomography (CT) image quality, adversely impacting clinical diagnoses. Various advanced LDCT methods have emerged to mitigate this challenge, relying on well-matched LDCT and normal-dose CT (NDCT) image pairs for training. Nevertheless, these methods often face difficulties in distinguishing image details from nonuniformly distributed noise, limiting their denoising efficacy. Additionally, acquiring suitably paired datasets in the medical domain poses challenges, further constraining their applicability. Hence, the objective of this study was to develop an innovative denoising framework for LDCT images employing unpaired data.MethodsIn this paper, we propose a LDCT denoising network (DNCNN) that alleviates the need for aligning LDCT and NDCT images. Our approach employs generative adversarial networks (GANs) to learn and model the noise present in LDCT images, establishing a mapping from the pseudo-LDCT to the actual NDCT domain without the need for paired CT images.ResultsWithin the domain of weakly supervised methods, our proposed model exhibited superior objective metrics on the simulated dataset when compared to CycleGAN and selective kernel-based cycle-consistent GAN (SKFCycleGAN): the peak signal-to-noise ratio (PSNR) was 43.9441, the structural similarity index measure (SSIM) was 0.9660, and the visual information fidelity (VIF) was 0.7707. In the clinical dataset, we conducted a visual effect analysis by observing various tissues through different observation windows. Our proposed method achieved a no-reference structural sharpness (NRSS) value of 0.6171, which was closest to that of the NDCT images (NRSS =0.6049), demonstrating its superiority over other denoising techniques in preserving details, maintaining structural integrity, and enhancing edge contrast.ConclusionsThrough extensive experiments on both simulated and clinical datasets, we demonstrated the superior efficacy of our proposed method in terms of denoising quality and quantity. Our method exhibits superiority over both supervised techniques, including block-matching and 3D filtering (BM3D), residual encoder-decoder convolutional neural network (RED-CNN), and Wasserstein generative adversarial network-VGG (WGAN-VGG), and over weakly supervised approaches, including CycleGAN and SKFCycleGAN.
| S-EPMC11320552 | biostudies-literature
Project description:Machine learning analysis of longitudinal neuroimaging data is typically based on supervised learning, which requires large number of ground-truth labels to be informative. As ground-truth labels are often missing or expensive to obtain in neuroscience, we avoid them in our analysis by combing factor disentanglement with self-supervised learning to identify changes and consistencies across the multiple MRIs acquired of each individual over time. Specifically, we propose a new definition of disentanglement by formulating a multivariate mapping between factors (e.g., brain age) associated with an MRI and a latent image representation. Then, factors that evolve across acquisitions of longitudinal sequences are disentangled from that mapping by self-supervised learning in such a way that changes in a single factor induce change along one direction in the representation space. We implement this model, named Longitudinal Self-Supervised Learning (LSSL), via a standard autoencoding structure with a cosine loss to disentangle brain age from the image representation. We apply LSSL to two longitudinal neuroimaging studies to highlight its strength in extracting the brain-age information from MRI and revealing informative characteristics associated with neurodegenerative and neuropsychological disorders. Moreover, the representations learned by LSSL facilitate supervised classification by recording faster convergence and higher (or similar) prediction accuracy compared to several other representation learning techniques.
| S-EPMC8184636 | biostudies-literature