IMC-Denoise: a content aware denoising pipeline to enhance Imaging Mass Cytometry

Lu, Peng; Oetjen, Karolyn A.; Bender, Diane; Ruzinova, Marianna B.; Fisher, Daniel A.C.; Shim, Kevin G.; Pachynski, Russell K.; Brennen, W. Nathaniel; Oh, Stephen T.; Link, Daniel C.; Thorek, Daniel L.J.

doi:10.1101/2022.07.21.501021

Cited by 10 publications

(18 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MD reflects the overall diffusion of water molecules in the tissues, the bigger MD indicates the less restriction for diffusion in tissues. FA and MD are defined as follows

\begin{equation}FA = \sqrt {\frac{1}{2}} \frac{{\sqrt {{{({\lambda _1} - {\lambda _2})}^2} + {{({\lambda _1} - {\lambda _3})}^2} + {{({\lambda _2} - {\lambda _3})}^2}} }}{{\sqrt {\lambda _1^2 + \lambda _2^2 + \lambda _3^2} }}\end{equation}

\begin{equation}MD = \frac{{({\lambda _1} + {\lambda _2} + {\lambda _3})}}{3}\end{equation}

where λ 1 , λ 2 , and λ 3 indicate three eigenvalues of the DT, respectively 29,30 . FO reflects the arrangement pattern of cardiac myocytes, which is determined by the direction of the eigenvector associated with the largest eigenvalue of the DT.…”

Section: Denoising Principle Of Ssecnn Methodsmentioning

confidence: 99%

Self‐supervised structural similarity‐based convolutional neural network for cardiac diffusion tensor image denoising

Wang

et al. 2023

Medical Physics

View full text Add to dashboard Cite

BackgroundDiffusion tensor imaging (DTI) is a promising technique for non‐invasively investigating the myocardial fiber structures of human heart. However, low signal‐to‐noise ratio (SNR) has been a major limit of cardiac DTI to prevent us from detecting myocardium structure accurately. Therefore, it is important to remove the effect of noise on diffusion weighted (DW) images.PurposeAlthough the conventional and deep learning‐based denoising methods have shown the potential to deal with effectively the noise in DW images, most of them are redundant information dependent or require the noise‐free images as golden standard. In addition, the existed DW image denoising methods often suffer from problems of over‐smoothing. To address these issues, we propose a self‐supervised learning model, structural similarity based convolutional neural network with edge‐weighted loss (SSECNN), to remove the noise effectively in cardiac DTI.MethodsConsidering that the DW images acquired along different diffusion directions have structural similarity, and the noise in these DW images is independent and identically distributed, the structural similarity‐based matching algorithm is proposed to search for the most similar DW images. Such similar noisy DW image pairs are then used as the input and target of the denoising network SSECNN, which consists of several convolutional and residual blocks. Through the self‐supervised training with these image pairs, the network can restore the clean DW images and retain the correlations between the denoised DW images along different directions. To avoid the over‐smoothing problem, we design a novel edge‐weighted loss which enables the network to adaptively adjust the loss weights with iterations and therefore to improve the detail preserve ability of the model. To verify the superiority of the proposed method, comparisons with state‐of‐the‐art (SOTA) denoising methods are performed on both synthetic and real acquired DTI datasets.ResultsExperimental results show that SSECNN can effectively reduce the noise in the DW images while preserving detailed texture and edge information and therefore achieve better performance in DTI reconstruction. For synthetic dataset, compared to the SOTA method, the root mean square error (RMSE), peak signal‐to‐noise ratio (PSNR), and structure similarity index measure (SSIM) between the denoised DW images obtained with SSECNN and noise‐free DW images are improved by 6.94%, 1.98%, and 0.76% respectively when the noise level is 10%. As for the acquired cardiac DTI dataset, the SSECNN method could significantly improve SNR and contrast to noise ratio (CNR) of cardiac DW images and achieve more regular helix angle (HA) and transverse angle (TA) maps. The ablation experimental results validate that using the structure similarity‐based method to search the similar DW image pairs yield the smallest loss, and with the help of the edge‐weighted loss, the denoised DW images and diffusion metric maps can preserve more details.ConclusionsThe proposed SSECNN method can fully explore the similarity between the DW images along different diffusion directions. Using such similarity and an edge‐weighted loss enable us to denoise cardiac DTI effectively in a self‐supervised manner. Our method can overcome the redundancy information dependence and over‐smoothing problem of the SOTA methods.

show abstract

“…MD reflects the overall diffusion of water molecules in the tissues, the bigger MD indicates the less restriction for diffusion in tissues. FA and MD are defined as follows

\begin{equation}FA = \sqrt {\frac{1}{2}} \frac{{\sqrt {{{({\lambda _1} - {\lambda _2})}^2} + {{({\lambda _1} - {\lambda _3})}^2} + {{({\lambda _2} - {\lambda _3})}^2}} }}{{\sqrt {\lambda _1^2 + \lambda _2^2 + \lambda _3^2} }}\end{equation}

\begin{equation}MD = \frac{{({\lambda _1} + {\lambda _2} + {\lambda _3})}}{3}\end{equation}

Section: Denoising Principle Of Ssecnn Methodsmentioning

confidence: 99%

Self‐supervised structural similarity‐based convolutional neural network for cardiac diffusion tensor image denoising

Wang

et al. 2023

Medical Physics

View full text Add to dashboard Cite

show abstract

“…These thresholds were then used in R script to binarize the KI67 and ZEB1 intensities into ZEB1 and KI67 positive and negative cell populations that were then used in further analysis. The images were prepared for visualization by using IMC denoise pipeline as described in work by Lu et al( 5 ).…”

Section: Supplementary Methods and Materialsmentioning

confidence: 99%

HER2 Cellular Localization, Cell-Cell Contacts, and Cell Density Regulate Cancer Cell Plasticity in HER2+ Breast Cancer

Sameri,

Das,

Shahrivari

et al. 2023

Preprint

View full text Add to dashboard Cite

Phenotypic plasticity in HER2+ breast cancer presents a significant challenge in comprehending and treating early-stage metastasis. We conducted an extensive study employingin vitromodels and patient samples to investigate the relationships among cell density, HER2 expression, HER2 cellular localization, and their impact on the plasticity program of HER2+ breast cancer cells.Our research reveals a previously unexplored facet of HER2+ breast cancer biology. In conditions of low cellular density, the endoplasmic reticulum (ER) undergoes a remarkable transformation, shifting its function from protein trafficking to serving as a regulator of calcium flux via the PLC pathway. This alteration in ER function leads to the accumulation of HER2 protein within the perinuclear ER compartment.Furthermore, we discovered that in situations of low cellular density or loss of cell-cell contact among epithelial cells, an epithelial-to-mesenchymal transition (EMT) program is activated, resulting in a significant upregulation of desmosome junctions. Desmosome junctions persist as the longest-lasting epithelial cell junctions on the cell surface and represent the primary means for solitary epithelial cells to establish initial contact. Notably, our research revealed a physical binding between HER2 and Desmoplakin (DSP), facilitating HER2 membrane localization upon early cell-cell contact that initiates the proliferation of cancer cells. The switching from migration to proliferation hinges on a competition between the HER2 membrane pathway and the PLC pathway for the shared target molecule, PIP2. Upon HER2 membrane localization, the PI3K pathway is activated, converting PIP2 to PIP3. Simultaneously, HER2 activation upregulates PDPK1, whose PH domain exhibits a high binding affinity to PIP2 molecules. Consequently, membrane localization of HER2 substantially diminishes PIP2 availability for the PLC pathway, redirecting cells toward proliferation. Analysis of patient samples confirms the loss of membrane bound HER2 in circulating tumor cells (CTCs) and highlights phenotypic transitions in solitary cancer cells. These cells are characterized by increased plakoglobin, elevated ZEB1, and reduced KI67 protein levels.In summary, our findings introduce a novel mechanism of phenotypic plasticity in HER2+ breast cancer that regulates phenotype switching. This mechanism holds promise for the interception of early-stage metastasis, potentially leading to improved outcomes for HER2+ breast cancer patients.

show abstract

“…There are a number of published "end to end" pipelines for IMC data analysis (8)(9)(10)(11)(12) that utilise open source software for segmentation such as Ilastik (13) and CellProfiler (14,15), as well as StarDist (16) and IMC-specific approaches that utilise deep learning (17). There have also been attempts to use matched fluorescent images of the nuclei using DAPI co-staining to improve segmentation accuracy (18) as well as removing image noise (19,20). Nonetheless, it has been shown that, due to the nature of tissue imaging, simple approaches to single cell segmentation are often highly effective (21).…”

Section: Introductionmentioning

confidence: 99%

OPTIMAL: An OPTimised Imaging Mass cytometry AnaLysis framework for benchmarking segmentation and data exploration

Hunter

Nicorescu

Foster

et al. 2023

Preprint

View full text Add to dashboard Cite

Analysis of Imaging Mass Cytometry (IMC) data and other low-resolution multiplexed tissue imaging technologies is often confounded by poor single cell segmentation and sub-optimal approaches for data visualisation and exploration. This can lead to inaccurate identification of cell phenotypes, states or spatial relationships compared to reference data from single cell suspension technologies. To this end we have developed the "OPTIMAL" framework to determine the best approaches for cell segmentation, parameter transformation, batch effect correction, data visualisation/clustering and spatial neighbourhood analysis. Using a panel of 27 metal-tagged antibodies recognising well characterised phenotypic and functional markers to stain the same FFPE human tonsil sample Tissue Microarray (TMA) over 12 temporally distinct batches we tested a total of four cell segmentation models, a range of different arcsinh cofactor parameter transformation values, five different dimensionality reduction algorithms and two clustering methods. Finally we assessed the optimal approach for performing neighbourhood analysis. We found that single cell segmentation was improved by the use of an Ilastik-derived probability map but that issues with poor segmentation were only really evident after clustering and cell type/state identification and not always evident when using "classical" bi-variate data display techniques. The optimal arcsinh cofactor for parameter transformation was 1 as it maximised the statistical separation between negative and positive signal distributions and a simple Z-score normalisation step after arcsinh transformation eliminated batch effects. Of the five different dimensionality reduction approaches tested, PacMap gave the best data structure with FLOWSOM clustering out-performing Phenograph in terms of cell type identification. We also found that neighbourhood analysis was influenced by the method used for finding neighbouring cells with a "disc" pixel expansion outperforming a "bounding box" approach combined with the need for filtering objects based on size and image-edge location. Importantly OPTIMAL can be used to assess and integrate with any existing approach to IMC data analysis and, as it creates .FCS files from the segmentation output, allows for single cell exploration to be conducted using a wide variety of accessible software and algorithms familiar to conventional flow cytometrists.

show abstract

IMC-Denoise: a content aware denoising pipeline to enhance Imaging Mass Cytometry

Cited by 10 publications

References 54 publications

Self‐supervised structural similarity‐based convolutional neural network for cardiac diffusion tensor image denoising

Self‐supervised structural similarity‐based convolutional neural network for cardiac diffusion tensor image denoising

HER2 Cellular Localization, Cell-Cell Contacts, and Cell Density Regulate Cancer Cell Plasticity in HER2+ Breast Cancer

OPTIMAL: An OPTimised Imaging Mass cytometry AnaLysis framework for benchmarking segmentation and data exploration

Contact Info

Product

Resources

About