Journal: Scientific Data

Article Title: A Real-World Benchmark for Sentinel-2 Multi-Image Super-Resolution

doi: 10.1038/s41597-023-02538-9

Figure Lengend Snippet: Reconstruction accuracy obtained for our MuS2 benchmark measured with PSNR (in dB), SSIM, LPIPS, and the balanced score \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathscr{B}}$$\end{document} B , obtained for different interpolation techniques alongside HighRes-net and RAMS networks trained using real-world and simulated images.

Article Snippet: However, it can be noticed that these qualitative observations are quantitatively reflected only in the LPIPS values, while PSNR and SSIM are slightly worse for HighRes-net and RAMS–overall, both LPIPS and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathscr{B}}$$\end{document} B indicate that SR networks perform better than interpolation and they penalize for the artefacts.

Techniques:

Journal: Scientific Data

Article Title: A Real-World Benchmark for Sentinel-2 Multi-Image Super-Resolution

doi: 10.1038/s41597-023-02538-9

Figure Lengend Snippet: Reconstruction outcome (band B08) obtained with RAMS and HighRes-net trained from real-life PROBA-V NIR and Red images and from simulated data, compared with image interpolation techniques and HR reference. PSNR (in dB), SSIM, LPIPS and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathscr{B}}$$\end{document} B scores are presented above each example. The images present an area of 5.5 × 6.5 km near Wageningen, Netherlands–the WV-2 image was acquired in June 2011, and S-2 images were acquired in April 2019–March 2021.

Article Snippet: However, it can be noticed that these qualitative observations are quantitatively reflected only in the LPIPS values, while PSNR and SSIM are slightly worse for HighRes-net and RAMS–overall, both LPIPS and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathscr{B}}$$\end{document} B indicate that SR networks perform better than interpolation and they penalize for the artefacts.

Techniques:

Journal: Scientific Data

Article Title: A Real-World Benchmark for Sentinel-2 Multi-Image Super-Resolution

doi: 10.1038/s41597-023-02538-9

Figure Lengend Snippet: The MOS survey outcome showing how often each metric was consistent with the answers (in %), stratified into SR networks trained with PROBA-V and simulated images, and interpolation.

Article Snippet: However, it can be noticed that these qualitative observations are quantitatively reflected only in the LPIPS values, while PSNR and SSIM are slightly worse for HighRes-net and RAMS–overall, both LPIPS and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathscr{B}}$$\end{document} B indicate that SR networks perform better than interpolation and they penalize for the artefacts.

Techniques:

Journal: Journal of Nuclear Medicine

Article Title: Projection Space Implementation of Deep Learning–Guided Low-Dose Brain PET Imaging Improves Performance over Implementation in Image Space

doi: 10.2967/jnumed.119.239327

Figure Lengend Snippet: Summary of Previous Studies Focusing on LD-to-FD PET Conversion in Brain PET Imaging Using Deep Learning Techniques

Article Snippet: The convolutional network trained in PIS applied simplistic noise reduction, thus leading to blurred, highly smoothed, and slightly biased FD images. table ft1 table-wrap mode="anchored" t5 TABLE 5 caption a7 Study Patients ( n ) Input Network architecture Injected activity (MBq) LD ratio (%) Tracer Scanning time (min) Time from injection to scan (min) Scanner model Evaluation method and metrics ( 16 ) 11 LD PET, MRI (T1) Segmented brain tissues from MRI to build tissue-specific models to predict standard-dose plus iterative refinement strategy 203 ± 12 25 18 F-FDG 12 36 Siemens Biograph mMR Leave-one-out cross-validation, SUV bias ( 8 ) 8 LD PET, MRI (T1, DTI) Sparse representation based on mapping strategy and incremental refinement scheme 203 ± 12 25 18 F-FDG 12 60 Siemens Biograph mMR NMSE, PSNR, contrast recovery, quantification bias ( 19 ) 16 LD PET MRI (T1) Deep auto-context CNN architecture with 4 layers 203 ± 12 25 18 F-FDG 12 60 Siemens Biograph mMR NMSE, PSNR, training and testing time comparison, training loss vs. iteration ( 22 ) 9 LD PET 2.5D U-Net (modified) 370 0.5 18 F-FDG 40 45 GE Signa PET/MRI NRMSE, error maps ( 15 ) 40 LD PET MRI (T1, T2, 3D) 2D U-Net 330 ± 30 1 18 F-florbetaben 20 90–110 GE Signa PET/MRI SSIM, PSNR, RMSE, clinical image quality scoring, Bland–Altman analysis ( 20 ) 16 LD PET 3D conditional GANs 203 ± 12 25 18 F-FDG 12 36 Siemens Biograph mMR SSIM, PSNR, RMSE, quantification bias ( 21 ) 40 LD PET 2.5D GAN 330 ± 30 1 18 F-florbetaben 20 90–111 GE SIGNA PET/MRI SSIM, PSNR, RMSE, clinical image quality scoring, error maps This work in PIS 140 LD PET image 3D U-Net (modified) 205 ± 10 5 18 F-FDG 20 34 Siemens Biograph mCT SSIM, PSNR, RMSE, clinical image quality scoring, Bland–Altman analysis, quantification bias, radiomic features, joint histogram, error maps This work in PSS 140 LD PET sinograms 3D U-Net (modified) 205 ± 10 5 18 F-FDG 20 34 Siemens Biograph mCT SSIM, PSNR, RMSE, clinical image quality scoring, Bland–Altman analysis, quantification bias, radiomic features, joint histogram, error maps Open in a separate window DTI = diffusion tensor imaging; GAN = generative adversarial network; NMSE = normalized mean square error; NRMSE = normalized root mean square error.

Techniques: Imaging, Comparison, Modification

Journal: Neurotherapeutics

Article Title: Regulatory T cells attenuate astrogliosis via IL-10/STAT3/PKC/vimentin signaling and promote neurological recovery after spinal cord injury

doi: 10.1016/j.neurot.2025.e00827

Figure Lengend Snippet: IL-10 supplementation and PKC phosphorylation inhibition recapitulate Treg-mediated downregulation of vimentin in vitro . A Representative immunofluorescence images of Foxp3 (green) and IL-10 (red) of magnetically sorted Tregs from mouse spleen. B ELISA quantification of IL-10 levels in supernatants of isolated Tregs. Cohen's d = 4.7545. C Experimental timeline of in vitro IL-10 intervention in astrocytes (rIL-10: recombinant IL-10; αIL-10, IL-10 neutralizing antibody). D-H Western blot quantification of p-STAT3/STAT3 (η 2 = 0.8939), SOCS3 (η 2 = 0.9199), p -PKC/PKC (η 2 = 0.9414), and vimentin (η 2 = 0.9252) protein levels in astrocytes following IL-10 intervention. I Representative immunofluorescence images of vimentin in astrocytes following IL-10 intervention. J Experimental timeline of in vitro PKC phosphorylation modulation in astrocytes (Go6976: PKC inhibitor; PMA: PKC activator). K, L Western blot quantification of vimentin protein levels in astrocytes following PKC modulation. η 2 = 0.8597. M, N Immunofluorescence quantification of vimentin levels in astrocytes following PKC modulation. η 2 = 0.6114. Data are mean ± SEM (n = 6 independent cultures). Statistical analyses were performed using two-tailed Student's t -test (B) or one-way ANOVA (E-H, L, N). P values are indicated on the graphs. n.s., not significant.

Article Snippet: The following primary antibodies were used for Western blot assays: GAPDH (Cell Signaling Technology, 2118S), Vimentin (Abcam, AB92547), PTPσ (Solarbio, K109903P), PKCα (Proteintech, 21991-1-AP), Phospho-PKCα (Thr638) (Proteintech, 29123-1-AP), STAT3 (Cell Signaling Technology, 12640), p-STAT3 (Cell Signaling Technology, 9145), and SOCS3 (Proteintech, 66797-1-lg).

Techniques: Phospho-proteomics, Inhibition, In Vitro, Immunofluorescence, Enzyme-linked Immunosorbent Assay, Isolation, Recombinant, Western Blot, Two Tailed Test

Journal: IEEE Winter Conference on Applications of Computer Vision. IEEE Winter Conference on Applications of Computer Vision

Article Title: PathLDM: Text conditioned Latent Diffusion Model for Histopathology

doi: 10.1109/wacv57701.2024.00510

Figure Lengend Snippet: Comparison of reconstruction quality of VAE used in Stable Diffusion, medfusion and PathLDM, evaluated on TCGA-BRCA. The input image size is 256× 256×3. A smaller downsampling factor (f) significantly boosts the SSIM.

Article Snippet: VAE Latent size f SSIM ↑ MSE ↓ Stable Diffusion 32x32x4 8 0.874 25.795 Medfusion 32x32x8 8 0.891 21.401 Ours 64x64x3 4 0.961 11.503 Open in a separate window Comparison of reconstruction quality of VAE used in Stable Diffusion, medfusion and PathLDM, evaluated on TCGA-BRCA.

Techniques: Comparison, Diffusion-based Assay

Journal: IEEE Winter Conference on Applications of Computer Vision. IEEE Winter Conference on Applications of Computer Vision

Article Title: PathLDM: Text conditioned Latent Diffusion Model for Histopathology

doi: 10.1109/wacv57701.2024.00510

Figure Lengend Snippet: Effect of latent size on the reconstruction quality of Satellite Images from Deepglobe [ 6 ] dataset. f: downsampling factor. The choice of VAE is significant even in diverse scenarios containing small objects.

Article Snippet: VAE Latent size f SSIM ↑ MSE ↓ Stable Diffusion 32x32x4 8 0.874 25.795 Medfusion 32x32x8 8 0.891 21.401 Ours 64x64x3 4 0.961 11.503 Open in a separate window Comparison of reconstruction quality of VAE used in Stable Diffusion, medfusion and PathLDM, evaluated on TCGA-BRCA.

Techniques:

Journal: Cell Death & Disease

Article Title: The helicase HAGE prevents interferon- α -induced PML expression in ABCB5+ malignant melanoma-initiating cells by promoting the expression of SOCS1

doi: 10.1038/cddis.2014.29

Figure Lengend Snippet: HAGE enhances SOCS1 RNA unwinding and protein expression. ( a ) Schematic representation of the mechanism of action of SOCS1. ( b ) Left panel: IF with antibodies to SOCS1 (green) and HAGE (red) on FM82 control and FM82 shRNA1 and 2 (HAGE knocked down). Right panel: IB with antibodies to HAGE, SOCS1 and actin (loading control) using whole lysates from FM82 and FM55 control and FM82 and FM55 shRNA1 and 2. Scale bar 50 μ m. ( c ) Representative image of immunohistochemistry with antibodies to SOCS1 (green) and HAGE (red) on normal skin and malignant melanoma sections. Scale bar 100 μ m. ( d ) IB with antibodies to HAGE, SOCS1, p-Jak1, p-Tyk2, PML and actin using whole cell extracts from FM82 and FM55 control and FM82 and FM55 shRNA1 and 2 transfected with siRNA control and siRNA for SOCS1 mRNA. ( e ) Transient silencing of SOCS1 in HAGE stable knockdown and control FM82 cells followed by rescue of HAGE expression using a HAGE cDNA expression vector. Actin was measured as a loading control. ( f ) Semi-quantitative real-time PCR of SOCS1 mRNAs isolated from FM82 control and FM82 shRNA1. ( g ) Unwinding of biotinylated SOCS1 N-terminal complimentary RNA sequence-SOCS1 RNA duplexes in the presence of increasing HAGE protein concentrations (lane 1: 0 μ g, lane 2: 0.6 μ g, lane 3: 1.2 μ g). Biotinylated SOCS1 N-terminal complimentary RNA sequence was used as a loading control in lane 4. ( h ) IP with antibody to SOCS1 of biotin alkyle metabolically labelled (30, 60 and 120 min) and unlabelled SOCS1 proteins followed by IB with streptavidin antibody from FM82 control whole cell extracts transfected with SOCS1cDNA and with or without transfected HAGE cDNA

Article Snippet: The plasmid pCMV6-XL4 containing SOCS1 was purchased from OriGene (SC111081) and used for in vitro transcription using T7 promoter to generate SOCS1 transcript (Ambion, Paisley, UK).

Techniques: Expressing, Immunohistochemistry, Transfection, Plasmid Preparation, Real-time Polymerase Chain Reaction, Isolation, Sequencing, Metabolic Labelling

Journal: Cell Death & Disease

Article Title: The helicase HAGE prevents interferon- α -induced PML expression in ABCB5+ malignant melanoma-initiating cells by promoting the expression of SOCS1

doi: 10.1038/cddis.2014.29

Figure Lengend Snippet: SOCS1 induces ubiquitination of p-Jak1 and p-Tyk2 in FM82 control malignant melanoma cell lines. ( a ) IP with antibody to SOCS1 or rabbit IgG followed by IB with antibodies to SOCS1, p-Jak1 and p-Tyk2 from FM82 control whole cell extracts. ( b and c ) IP with antibody to p-Jak1 or p-Tyk2 and corresponding IgG (rabbit or goat, respectively) followed by IB with antibodies to p-Jak1 or p-Tyk2 and ubiquitin from FM82 control and FM82 shRNA1 whole cell extracts. ( d ) IP with antibody to p-Jak1 followed by whole cell extracts IB with an antibody to ubiquitin from FM82 shRNA1 transfected with HAGE cDNA or SOCS1 cDNA. The protein expression of HAGE and SOCS1 was determined by IB with the corresponding antibodies

Article Snippet: The plasmid pCMV6-XL4 containing SOCS1 was purchased from OriGene (SC111081) and used for in vitro transcription using T7 promoter to generate SOCS1 transcript (Ambion, Paisley, UK).

Techniques: Transfection, Expressing

Journal: Cell Death & Disease

Article Title: The helicase HAGE prevents interferon- α -induced PML expression in ABCB5+ malignant melanoma-initiating cells by promoting the expression of SOCS1

doi: 10.1038/cddis.2014.29

Figure Lengend Snippet: HAGE counteracts the anti-proliferative effect of IFN α in ABCB5+ MMICs. ( a , b and c ) IF with antibodies to HAGE (red), ABCB5 (green), SOCS1 (green) and PML (green) on FM82 control, FM82 shRNA1 and 2 melanoma spheres. ( d ) IF and IB with an antibody to PML (or IgG) on FM82 cell lines and corresponding melanoma spheres stably-expressing PML or the empty vector. ( e ) FM82 empty vector and stably expressing PML spheres formation in the presence or absence of cDNA transfected HAGE or SOCS1. ANOVA: *** P =<0.0001. Scale bar 100 μ m (for spheres) and 20 μ m (for cells). ( f ) IF and IB with an antibody to PML on FM82 shRNA/PMLshRNA CTL and FM82shRNA/PMLshRNA 1 and 2. ( g ) Spheres formation assay from FM82shRNA/PMLshRNA CTL and FM82shRNA/PML shRNA 1 and 2. ANOVA: *** P =<0.0001

Article Snippet: The plasmid pCMV6-XL4 containing SOCS1 was purchased from OriGene (SC111081) and used for in vitro transcription using T7 promoter to generate SOCS1 transcript (Ambion, Paisley, UK).

Techniques: Stable Transfection, Expressing, Plasmid Preparation, Transfection, shRNA, Tube Formation Assay

Journal: Cell Death & Disease

Article Title: The helicase HAGE prevents interferon- α -induced PML expression in ABCB5+ malignant melanoma-initiating cells by promoting the expression of SOCS1

doi: 10.1038/cddis.2014.29

Figure Lengend Snippet: HAGE counteracts the anti-proliferative effect of IFN α in vivo . ( a , b and c ) FM82 and FM55 controls and FM82 and FM55 shRNA 1 and 2 melanoma sphere formation in the presence or absence of IFN α . Student t -test: ** P =<0.01. ( d ) Summary of the in vivo experimental procedure. ( e ) Measurement of tumour growth in mice injected with either FM82 control cells or FM82 shRNA1 cells and treated or untreated with IFN α ANOVA: ** P =<0.0028; Mann–Whitney U -test: * P =<0.0108 (when comparing FM82 control + PBS and FM82 shRNA + PBS; Mann–Whitney U -test: ** P =<0.0079 (when comparing FM82 control + IFN α and FM82 shRNA + IFN α treated). ( f ) Immunohistochemistry/fluorescence staining with antibodies to HAGE (red), ABCB5 (green), SOCS1 (green) and on sections from FM82 control and FM82 shRNA1 NOD/SCID xenotransplanted tumours and treated or untreated with IFN α . Scale bar 100 μ m

Article Snippet: The plasmid pCMV6-XL4 containing SOCS1 was purchased from OriGene (SC111081) and used for in vitro transcription using T7 promoter to generate SOCS1 transcript (Ambion, Paisley, UK).

Techniques: In Vivo, shRNA, Injection, MANN-WHITNEY, Immunohistochemistry, Fluorescence, Staining

Journal: Advances in neural information processing systems

Article Title: CELL-E 2: Translating Proteins to Pictures and Back with a Bidirectional Text-to-Image Transformer

doi:

Figure Lengend Snippet: Validation Set Image Prediction Accuracy. MAPE: mean absolute percentage error, MAE: mean absolute error, SSIM: structural similarity index measure, FID: Fréchet inception distance, IS: inception score.

Article Snippet: Table 3: Fine-Tuned Threshold Image Encoder Nucleus Proportion MAPE Image MAE PDF MAE SSIM FID IS No HPA 0.0181 ± 0.0168 0.4154 ± 0.0594 0.3887 ± 0.1270 0.1250 ± 0.1149 3.9509 2.1739 ± 0.1255 No OpenCell 0.0161 ± 0.0148 0.4953 ± 0.0064 0.3620 ± 0.1168 0.1220 ± 0.1188 1.5844 2.6069 ± 0.1175 Yes HPA 0.0166 ± 0.0151 0.3776 ± 0.0834 0.3477 ± 0.1268 0.1869 ± 0.1503 17.4075 2.9113 ± 0.1199 Yes OpenCell 0.0159 ± 0.0156 0.4996 ± 0.0006 0.3506 ± 0.1208 0.1574 ± 0.1372 2.5026 2.7168 ± 0.1137 Yes HPA Finetuned 0.0170 ± 0.0160 0.3449 ± 0.1305 0.3487 ± 0.1340 0.1881 ± 0.1541 19.2683 3.6083 ± 0.2013 Open in a separate window OpenCell Validation Set Image Prediction Accuracy after Finetuning.

Techniques: Biomarker Discovery

Journal: Advances in neural information processing systems

Article Title: CELL-E 2: Translating Proteins to Pictures and Back with a Bidirectional Text-to-Image Transformer

doi:

Figure Lengend Snippet: OpenCell Validation Set Image Prediction Accuracy after Finetuning

Article Snippet: Table 3: Fine-Tuned Threshold Image Encoder Nucleus Proportion MAPE Image MAE PDF MAE SSIM FID IS No HPA 0.0181 ± 0.0168 0.4154 ± 0.0594 0.3887 ± 0.1270 0.1250 ± 0.1149 3.9509 2.1739 ± 0.1255 No OpenCell 0.0161 ± 0.0148 0.4953 ± 0.0064 0.3620 ± 0.1168 0.1220 ± 0.1188 1.5844 2.6069 ± 0.1175 Yes HPA 0.0166 ± 0.0151 0.3776 ± 0.0834 0.3477 ± 0.1268 0.1869 ± 0.1503 17.4075 2.9113 ± 0.1199 Yes OpenCell 0.0159 ± 0.0156 0.4996 ± 0.0006 0.3506 ± 0.1208 0.1574 ± 0.1372 2.5026 2.7168 ± 0.1137 Yes HPA Finetuned 0.0170 ± 0.0160 0.3449 ± 0.1305 0.3487 ± 0.1340 0.1881 ± 0.1541 19.2683 3.6083 ± 0.2013 Open in a separate window OpenCell Validation Set Image Prediction Accuracy after Finetuning.

Techniques: Biomarker Discovery