Learnable Stroke Models for Example-based Portrait Painting

With the advent of the deep learning method, portrait video stylization has become more popular. In this paper, we present a robust method for automatically stylizing portrait videos that contain small human faces. By extending the Mask Regions with Convolutional Neural Network features (R-CNN) with a CNN branch which detects the contour landmarks of the face, we divided the input frame into three regions: the region of facial features, the region of the inner face surrounded by 36 face contour landmarks, and the region of the outer face. Besides keeping the facial features region as it is, we used two different stroke models to render the other two regions. During the non-photorealistic rendering (NPR) of the animation video, we combined the deformable strokes and optical flow estimation between adjacent frames to follow the underlying motion coherently. The experimental results demonstrated that our method could not only effectively reserve the small and distinct facial features, but also follow the underlying motion coherently.

Download Full-text

THE SOCIETY PORTRAIT: PAINTING, PRESTIGE AND THE PURSUIT OF ELEGANCE BY GABRIEL BADEA-PÄUN

The Art Book ◽

10.1111/j.1467-8357.2009.01010_9.x ◽

2009 ◽

Vol 16 (1) ◽

pp. 46-47

Author(s):

DAVID A H B TAYLOR

Keyword(s):

Portrait Painting

Download Full-text

Therapeutically Targeting Neuroinflammation and Microglia after Acute Ischemic Stroke

BioMed Research International ◽

10.1155/2014/297241 ◽

2014 ◽

Vol 2014 ◽

pp. 1-9 ◽

Cited By ~ 40

Author(s):

Youngjeon Lee ◽

Sang-Rae Lee ◽

Sung S. Choi ◽

Hyeon-Gu Yeo ◽

Kyu-Tae Chang ◽

...

Keyword(s):

Ischemic Stroke ◽

Acute Ischemic Stroke ◽

Inflammatory Responses ◽

Secondary Injury ◽

Initial Event ◽

Stroke Models ◽

Anti Inflammatory ◽

Preclinical And Clinical Studies ◽

Multiple Receptor

Inflammation has a pivotal role in the pathogenesis of ischemic stroke, and recent studies posit that inflammation acts as a double-edged sword, not only detrimentally augmenting secondary injury, but also potentially promoting recovery. An initial event of inflammation in ischemic stroke is the activation of microglia, leading to production of both pro- and anti-inflammatory mediators acting through multiple receptor signaling pathways. In this review, we discuss the role of microglial mediators in acute ischemic stroke and elaborate on preclinical and clinical studies focused on microglia in stroke models. Understanding how microglia can lead to both pro- and anti-inflammatory responses may be essential to implement therapeutic strategies using immunomodulatory interventions in ischemic stroke.

Download Full-text

Abstract WP281: Incorporating Biological Variability in Experimental Stroke to Better Mimic Human Stroke

Stroke ◽

10.1161/str.48.suppl_1.wp281 ◽

2017 ◽

Vol 48 (suppl_1) ◽

Author(s):

Thomas A Kent ◽

Harriett C Rea ◽

William Dalmeida ◽

Roderic H Fabian ◽

Cenk Ayata ◽

...

Keyword(s):

Middle Surface ◽

Lower Surface ◽

Clinical Results ◽

Experimental Stroke ◽

Biological Variability ◽

Occlusion Time ◽

Physiological Variability ◽

Functional Measure ◽

Stroke Models ◽

Human Stroke

Introduction: Failures to translate pre-clinical results have been discouraging. We have contended that stroke is too heterogeneous with respect to factors influencing outcome to expect small studies to be balanced. It is not only difficult to control for biological and methodological variability but efforts to improve homogeneity, such as minimizing physiological variability, may render results less applicable to humans. Here, we report a predictive outcome model in experimental stroke which incorporates baseline variability and provides statistical thresholds a treatment must exceed to be efficacious in a broad population. Methods: We generated a mathematical model to predict outcome using transient MCA occlusion in 23 unfasted rats. To create baseline variability, we varied occlusion times from 90-120 min, altered baseline glucose with streptozotocin, and assessed neurological outcome 3 days later with a modified Bederson Score (BS; 0-6 functional measure, 7 death). Statistical surfaces in 3 dimensions were generated using Jacobian matrices flanking the model to provide a screening threshold (1 SD) for comparing new therapies against this model. Results: We successfully generated an outcome model from occlusion time, glucose and BS (Fig; R 2 =.49, p=.0003; middle surface is the model surrounded by ±SD surfaces). Outcome was sensitive to change in glucose and time, suggesting small imbalances in these factors between groups may influence outcome, and hence the perceived efficacy of a new therapeutic intervention. At normoglycemia and 90 mins, the lower surface overlapped with no deficit, indicating it would be difficult to reliably demonstrate benefit under those conditions. Conclusions: These results indicate it is feasible to incorporate biological variability to generate more clinically relevant conditions. The method will be tested with other stroke models and modifiers towards a generalized model to screen for therapies worthy of further study.

Download Full-text

Abstract MP56: Nitric Oxide-Therapy but Not Remote Ischemic Conditioning is Cerebroprotective in Stroke: A Study in Mouse Model of Endothelial Dysfunction and Hypertension

Stroke ◽

10.1161/str.52.suppl_1.mp56 ◽

2021 ◽

Vol 52 (Suppl_1) ◽

Author(s):

Syed M Zaidi ◽

Md Nasrul Hoda ◽

Farid Ahmed ◽

Heba Alkhatabi ◽

Muhammed H Al Qahtani

Keyword(s):

Nitric Oxide ◽

Endothelial Dysfunction ◽

Statistical Significance ◽

Intravenous Thrombolysis ◽

Once Daily ◽

Remote Ischemic Conditioning ◽

Ischemic Conditioning ◽

Stroke Models ◽

Endogenous Protection ◽

Endothelial Nitric Oxide

Background: Remote Ischemic Conditioning (RIC) was found effective in stroke models, likely via increased endothelial nitric oxide (NO); yet RIC failed to improve clinical outcomes ( NCT02342522 ; NCT02189928 ). We anticipated that comorbidities neutralize the benefits of RIC in stroke. Hypothesis: NO-therapy but not RIC is vasculoprotective in hypertensive stroke. Methods: Aged (18±1-mo old) S100A1-hets mutant (S100A +/- ) and wild-type (WT) mice were used. As needed, mice were treated with RIC, s-nitrosoglutathione (GSNO) reductase inhibitor (GRI; 5 mg/kg nebulized once daily for 2-wks), GSNO (100-ug/kg; nebulized once daily at 2h post-TES), and/or intravenous thrombolysis (IVT; 10mg/kg at 4h post-TES). Stroke and outcome measures were performed as mentioned below. Statistical significance was determined at P &lt 0.05. Results: S100A +/- compared to WT-type mice showed significantly higher mean arterial pressure (MAP) and lower plasma-NO, supporting a hypertensive phenotype with endothelial dysfunction in S100A +/- mice. In photothrombotic stroke (PTS), RIC significantly improved cerebral blood flow (CBF), behavior and reduced infarction in WT but not in S100A +/- mice at 48h. GRI in S100A +/- mice enhanced plasma NO, improved behavior and CBF, and reduced infarction significantly as compared to vehicle-treated S100A +/- at 48h post-PTS. RIC with GRI did not produce additive protection in S100A +/- mice at 48h post-PTS, demonstrating that the preservation of NO-pool with GRI protects against stroke, but RIC is not effective to enhance this endogenous protection in hypertensive mice. Moreover, GSNO nebulization but not RIC enhanced PbtO 2 , reduced BBB-leakage and brain hemoglobin (Hb)-content at 24h after thromboembolic stroke with and without IVT. Conclusions: NO-therapies but not RIC is effective in hypertensive stroke. RIC- and NO- therapies need further validation in comorbid stroke before embarking on the clinical trial.

Download Full-text

Noninvasive Brain Imaging in Small Animal Stroke Models: MRI and PET

Neuromethods - Rodent Models of Stroke ◽

10.1007/978-1-60761-750-1_11 ◽

2010 ◽

pp. 139-165 ◽

Cited By ~ 1

Author(s):

Anna M. Planas

Keyword(s):

Brain Imaging ◽

Small Animal ◽

Stroke Models

Download Full-text

Impact of C57BL/6 substrain on sex-dependent differences in mouse stroke models

Neurochemistry International ◽

10.1016/j.neuint.2018.11.011 ◽

2019 ◽

Vol 127 ◽

pp. 12-21 ◽

Cited By ~ 3

Author(s):

Thaddeus S. Nowak ◽

Megan K. Mulligan

Keyword(s):

Stroke Models

Download Full-text

Atlas registration for edema-corrected MRI lesion volume in mouse stroke models

Journal of Cerebral Blood Flow & Metabolism ◽

10.1177/0271678x17726635 ◽

2017 ◽

Vol 39 (2) ◽

pp. 313-323 ◽

Cited By ~ 17

Author(s):

Stefan Koch ◽

Susanne Mueller ◽

Marco Foddis ◽

Thomas Bienert ◽

Dominik von Elverfeldt ◽

...

Keyword(s):

Vasogenic Edema ◽

Lesion Size ◽

Artery Occlusion ◽

Brain Atlas ◽

Mathematical Framework ◽

Lesion Volume ◽

Atlas Registration ◽

Stroke Models ◽

Volume Measurements ◽

Automated Image Registration

Lesion volume measurements with magnetic resonance imaging are widely used to assess outcome in rodent models of stroke. In this study, we improved a mathematical framework to correct lesion size for edema which is based on manual delineation of the lesion and hemispheres. Furthermore, a novel MATLAB toolbox to register mouse brain MR images to the Allen brain atlas is presented. Its capability to calculate edema-corrected lesion size was compared to the manual approach. Automated image registration performed equally well in in a mouse middle cerebral artery occlusion model (Pearson r = 0.976, p = 2.265e-11). Information encapsulated in the registration was used to generate maps of edema induced tissue volume changes. These showed discrepancies to simplified tissue models underlying the manual approach. The presented techniques provide biologically more meaningful, voxel-wise biomarkers of vasogenic edema after stroke.

Download Full-text