ABSTRACT
RATIONALE AND OBJECTIVES: We aimed at developing and validating a nomogram and machine learning (ML) models based on radiomics score (Radscore), morphology, and PHASES to predict intracranial aneurysm (IA) rupture. MATERIALS AND METHODS: We collected 440 patients with IAs in our hospital from 2015 to 2023, totaling 475 IAs (214 ruptured and 261 unruptured). A 7:3 random split was utilized to allocate participants into training and testing sets. To optimize the selection of radiomics features extracted from digital subtraction angiography, we employed t-tests and LASSO regression. Subsequently, we built single-factor and multifactor logistic regression (LR) models, alongside a nomogram. Furthermore, we employed four ML algorithms. After a comprehensive evaluation, including area under the curve (AUC), calibration curves, decision curve analysis (DCA), and other metrics, the best model was determined. RESULTS: The AUCs for LR models P (PHASES), M (Morphology), and R (Radscore) in the testing set were 0.859, 0.755, and 0.803, respectively, while those for multifactor models R+M (Radscore and Morphology), R+P (Radscore and PHASES), and R+M+P (Radscore, Morphology, and PHASES) were 0.818, 0.899, and 0.887, respectively. The AUCs of random forest, extreme gradient boosting, gradient boosting machine, and light gradient boosting machine were 0.880, 0.888, 0.891, and 0.892 in testing set, respectively. In the training set, the LR model showed significant differences in AUCs compared with the four ML models (all p < 0.05). However, in the testing set, no statistically significant differences were found between them (all p > 0.05). Both ML models and the nomogram exhibit excellent performance in DCA and calibration curves. CONCLUSION: Nomogram and ML models based on Radscore, morphology, and PHASES show high precision in predicting aneurysm rupture.
ABSTRACT
Sound source localization aims to localize objects emitting the sound in visual scenes. Recent works obtaining impressive results typically rely on contrastive learning. However, the common practice of randomly sampling negatives in prior arts can lead to the false negative issue, where the sounds semantically similar to visual instance are sampled as negatives and incorrectly pushed away from the visual anchor/query. As a result, this misalignment of audio and visual features could yield inferior performance. To address this issue, we propose a novel audio-visual learning framework which is instantiated with two individual learning schemes: self-supervised predictive learning (SSPL) and semantic-aware contrastive learning (SACL). SSPL explores image-audio positive pairs alone to discover semantically coherent similarities between audio and visual features, while a predictive coding module for feature alignment is introduced to facilitate the positive-only learning. In this regard SSPL acts as a negative-free method to eliminate false negatives. By contrast, SACL is designed to compact visual features and remove false negatives, providing reliable visual anchor and audio negatives for contrast. Different from SSPL, SACL releases the potential of audio-visual contrastive learning, offering an effective alternative to achieve the same goal. Comprehensive experiments demonstrate the superiority of our approach over the state-of-the-arts. Furthermore, we highlight the versatility of the learned representation by extending the approach to audio-visual event classification and object detection tasks. Code and models are available at: https://github.com/zjsong/SACL.
ABSTRACT
Training with more data has always been the most stable and effective way of improving performance in the deep learning era. The Open Images dataset, the largest object detection dataset, presents significant opportunities and challenges for general and sophisticated scenarios. However, its semi-automatic collection and labeling process, designed to manage the huge data scale, leads to label-related problems, including explicit or implicit multiple labels per object and highly imbalanced label distribution. In this work, we quantitatively analyze the major problems in large-scale object detection and provide a detailed yet comprehensive demonstration of our solutions. First, we design a concurrent softmax to handle the multi-label problems in object detection and propose a soft-balance sampling method with a hybrid training scheduler to address the label imbalance. This approach yields a notable improvement of 3.34 points, achieving the best single-model performance with a mAP of 60.90% on the public object detection test set of Open Images. Then, we introduce a well-designed ensemble mechanism that substantially enhances the performance of the single model, achieving an overall mAP of 67.17%, which is 4.29 points higher than the best result from the Open Images public test 2018. Our result is published on https://www.kaggle.com/c/open-images-2019-object-detection/leaderboard.
ABSTRACT
BACKGROUND: The superomedial pedicle reduction mammoplasty has gained popularity and is an important alternative approach for reduction mammoplasty, while the inferior pedicle reduction mammaplasty remains by far the most performed as it is considered to provide the best vascularization to the nipple-areola complex, allowing safe removal of large amount of redundant tissue. The authors conducted the first systematic review and meta-analysis in an attempt to declare the differences of the superomedial pedicle versus the inferior pedicle reduction technique by comparing the postoperative complications. METHODS: PubMed, MEDLINE, and Cochrane Library for clinical studies were queried from inception to January 1, 2024. Review Manager Version 5.4 was used for this meta-analysis. A random effects model was applied to OR, and 95%CI were determined using the Mantel-Haenszel method. The I2 test was used to assess heterogeneity, and the Newcastle-Ottawa scale was used to assess the risk of bias in the nonrandomized studies. RESULTS: Twelve observational comparative studies were included. The superomedial pedicle technique had a statistically lower rate of overall complications (OR 0.59, 95% CI 0.47-0.75; p < 0.0001) and delayed wound healing (OR 0.46, 95% CI 0.33-0.64; p < 0.00001) than the inferior pedicle technique. No significant differences in wound dehiscence, infection, seroma, hematoma, skin necrosis, fat necrosis, NAC necrosis, nipple sensation decrease or loss, asymmetry, hypertrophic scarring, and reoperation were noted. CONCLUSIONS: Both two techniques are equally safe and reliable, while the superomedial pedicle technique resulted in a statistically lower rate of overall complications and delayed wound healing. LEVEL OF EVIDENCE III: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .
ABSTRACT
BACKGROUND: Gynecomastia is a progressive disease characterized by enlarged breasts, affecting a significant proportion of men. Persistent gynecomastia negatively affects the psychological and emotional development of patients; therefore, surgical intervention is required. In this article, we describe a surgical technique, where liposuction through an axillary incision is used in combination with a single small periareolar incision, to obtain the most minimal scars in the treatment of gynecomastia. METHODS: Between June 2021 and June 2023, 125 patients with different Simon grades of gynecomastia were enrolled. The patients' basic conditions and operation processes were recorded. Following surgery, a score was assigned according to the five main aesthetic aspects of the surgical procedure. RESULTS: In total, 125 patients with gynecomastia were treated with a pre-axillary fold incision combined with a small areolar incision. There were 17 cases of Simon grade I, 46 grade IIA, 42 grade IIB, and 20 grade III. The average operation time was 45.8 min, the average liposuction volume was 250.5 mL, the average glandular tissue volume was 50.5 g, intraoperative blood loss ranged from 15 to 60 mL, and the average hospital stay was 3.2 days. Regarding the postoperative aesthetic effect, doctors scored > 4 points, and the patient satisfaction score was > 7.5, which fully affirmed the aesthetic effect of this method. CONCLUSIONS: Treatment of gynecomastia through an anterior axillary fold incision combined with a small areolar incision is safe and feasible, involving a simple procedure, short operation time, and few complications. Its efficacy and cosmetic effects could lead to its use as a primary surgical method to treat gynecomastia. LEVEL OF EVIDENCE IV: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine Ratings, please refer to Table of Contents or online Instructions to Authors www.springer.com/00266 .
ABSTRACT
Source-free domain adaptation has developed rapidly in recent years, where the well-trained source model is adapted to the target domain instead of the source data, offering the potential for privacy concerns and intellectual property protection. However, a number of feature alignment techniques in prior domain adaptation methods are not feasible in this challenging problem setting. Thereby, we resort to probing inherent domain-invariant feature learning and propose a curriculum-style self-training approach for source-free domain adaptive semantic segmentation. In particular, we introduce a curriculum-style entropy minimization method to explore the implicit knowledge from the source model, which fits the trained source model to the target data using certain information from easy-to-hard predictions. We then train the segmentation network by the proposed complementary curriculum-style self-training, which utilizes the negative and positive pseudo labels following the curriculum-learning manner. Although negative pseudo-labels with high uncertainty cannot be identified with the correct labels, they can definitely indicate absent classes. Moreover, we employ an information propagation scheme to further reduce the intra-domain discrepancy within the target domain, which could act as a standard post-processing method for the domain adaptation field. Furthermore, we extend the proposed method to a more challenging black-box source model scenario where only the source model's predictions are available. Extensive experiments validate that our method yields state-of-the-art performance on source-free semantic segmentation tasks for both synthetic-to-real and adverse conditions datasets. The code and corresponding trained models are released at https://github.com/yxiwang/ATP.
ABSTRACT
Metabolic reprogramming is frequently accompanied by hepatocellular carcinoma (HCC) progression. Disrupted metabolites act as potential biomarkers and drug therapeutic targets for HCC. Peptide extract of scorpion venom (PESV) induces cytotoxic anti-proliferative effects and apoptosis in tumors. However, the action mechanisms of PESV remain unknown. This study aimed to explore the serum metabolic profiles of tumor-bearing mouse model. We generated an orthotopic HCC xenograft mouse model by implanting H22 cells into the left hepatic lobe of male C57BL/6 mice. After surgery, the mice were assigned to two groups randomly: PESV (PESV-treated 40 mg/kg daily, i.g.; n = 6) and control (treated with the solvent equally for 14 d, n = 6) groups. Based on an untargeted metabolomics approach using ultra-high-performance liquid chromatography/quadrupole time-of-flight mass spectrometry, differential metabolites were screened via univariate and multivariate data analyses. A total of 48 differential metabolites in negative ion mode and 63 in positive ion mode were identified in the serum samples. Furthermore, metabolic pathway analysis revealed that aminoacyl-tRNA biosynthesis, amino acid pathway, glutathione metabolism, protein transports, protein digestion and absorption, and cAMP signaling pathways play vital roles in PESV-induced inhibition of tumors. These findings highlight the distinct changes in the metabolic profiles of HCC-bearing mice after PESV treatment, suggesting the potential of the identified metabolic molecules as therapeutic targets for HCC.
Subject(s)
Carcinoma, Hepatocellular , Liver Neoplasms , Metabolomics , Mice, Inbred C57BL , Scorpion Venoms , Animals , Carcinoma, Hepatocellular/drug therapy , Carcinoma, Hepatocellular/metabolism , Male , Liver Neoplasms/drug therapy , Liver Neoplasms/metabolism , Mice , Cell Line, Tumor , Humans , Xenograft Model Antitumor Assays , Antineoplastic Agents/therapeutic use , Antineoplastic Agents/pharmacology , Metabolome/drug effects , Disease Models, AnimalABSTRACT
BACKGROUND: Breast reduction surgery has witnessed significant advancements in recent years; however, it continues to pose challenges for both surgeons and patients when dealing with cases involving excessive breast volume and severe breast ptosis. This study aimed to assess the aesthetic outcomes and the impact on the quality of life, as measured by the BREAST-Q questionnaire, in patients with gigantomastia and severe breast ptosis who underwent reduction mammaplasty using the superomedial-based pedicle technique. METHODS: We present a retrospective series comprising 84 patients who underwent reduction mammoplasty utilizing the superomedial pedicle technique. The surgical resections exceeded 1 kg per breast, with a mean resection weight of 1506.58 g (right breast) and 1500.32 g (left breast). The preoperative mean suprasternal notch to nipple distance measured 40.50 cm (right breast) and 40.38 cm (left breast). Postoperatively, the patients were followed up for a minimum of 6 months. Both preoperative and postoperative BREAST-Q surveys were administered to the participants, and scores were analyzed using descriptive statistics. RESULTS: Complications were observed in 3 patients (3.57%), characterized by partial loss of the areola, which resolved spontaneously over time. Additionally, 2 cases of hematoma and 2 instances of minor delayed wound healing were reported. All patients expressed satisfaction with their aesthetic outcomes, as they achieved a natural breast shape and minimal scarring, along with symptomatic relief. CONCLUSIONS: The superomedial pedicle reduction mammaplasty technique has demonstrated its ability to produce satisfactory aesthetic outcomes and long-term benefits in patients with excessively large breasts. Careful patient selection and postoperative management are vital for achieving optimal results. Further investigations involving larger sample sizes and longer follow-up periods are warranted to validate our findings. LEVEL OF EVIDENCE: IV.
ABSTRACT
BACKGROUND: Skin incision scars are cosmetically displeasing; the effects of current treatments are limited, and new methods to reduce scar formation need to be found. OBJECTIVE: We sought to determine whether immediate postoperative injection of stromal vascular fraction gel (SVF-gel) could reduce scar formation at skin incision sites. METHODS: A prospective, randomized, double-blind, self-controlled trial was conducted in patients who underwent breast reduction. SVF-gel was intradermally injected into the surgical incision on one randomly selected side, with the other side receiving saline as a control. At the 6-month follow-up, the incision scars were evaluated using the Vancouver scar scale (VSS) and visual analog scale (VAS). Antera 3D camera was used for objective evaluation. RESULTS: The VSS score and VAS score were significantly different between the SVF-gel-treated side (3.80 ± 1.37, 3.37±1.25) and the control side (5.25 ± 1.18, 4.94 ± 1.28). Moreover, the SVF-gel-treated side showed statistically significant improvements in scar appearance, based on evidences from Antera 3D camera. LIMITATIONS: This was a single-center, single-race, and single-gender study. Furthermore, the results were available only for the 6-month interim follow-up period. CONCLUSION: Postoperative immediate SVF-gel injection in surgical incisions can reduce scar formation, and exert a preventive effect on scars. LEVEL OF EVIDENCE I: Evidence obtained from at least one properly designed randomized controlled trial. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .
ABSTRACT
BACKGROUND: Vertical mammoplasty techniques have been widely used for breast reduction. The authors present the combination of superior pedicle vertical mammoplasty with liposuction in different regions in the treatment of severe breast hypertrophy in obese patients. We also propose some innovative methods in terms of surgical approach, breast parenchymal anatomy pattern and liposuction. METHODS: A retrospective study of 50 female patients with severe hypertrophic breasts and obesity who underwent breast reduction in our department from February 2019 to February 2022 was performed. Pre- and postoperative photographs, breast parenchyma distribution and postoperative patient satisfaction were recorded. RESULTS: Fifty patients underwent breast reduction. Through clinical examination, patient photo evaluation and satisfaction survey results. Good breast shape and projection, full upper pole of the breast, and high satisfaction results were obtained. There were no serious complications. CONCLUSION: This technique is acceptable and reproducible. It is suitable for patients with varying degrees of breast hypertrophy, especially those with severe hypertrophic breasts and obesity. There are fewer associated complications and a lower rate of re-repair. LEVEL OF EVIDENCE V: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .
ABSTRACT
Electroacupuncture (EA) stimulation has been shown to be beneficial in stroke rehabilitation; however, little is known about the neurological mechanism by which this peripheral stimulation approach treats for stroke. This study showed that both pyramidal and parvalbumin (PV) neuronal activity increased in the contralesional primary motor cortex forelimb motor area (M1FL) after ischemic stroke induced by focal unilateral occlusion in the M1FL. EA stimulation reduced pyramidal neuronal activity and increased PV neuronal activity. These results were obtained by a combination of fiber photometry recordings, in vivo and in vitro electrophysiological recordings, and immunofluorescence. Moreover, EA was found to regulate the expression/function of N-methyl-D-aspartate receptors (NMDARs) altered by stroke pathology. In summary, our findings suggest that EA could restore disturbed neuronal activity through the regulation of the activity of pyramidal and PV neurons. Furthermore, NMDARs we shown to play an important role in EA-mediated improvements in sensorimotor ability during stroke rehabilitation.
ABSTRACT
Currently prevalent multi-modal 3D detection methods rely on dense detectors that usually use dense Bird's-Eye-View (BEV) feature maps. However, the cost of such BEV feature maps is quadratic to the detection range, making it not scalable for long-range detection. Recently, LiDAR-only fully sparse architecture has been gaining attention for its high efficiency in long-range perception. In this paper, we study how to develop a multi-modal fully sparse detector. Specifically, our proposed detector integrates the well-studied 2D instance segmentation into the LiDAR side, which is parallel to the 3D instance segmentation part in the LiDAR-only baseline. The proposed instance-based fusion framework maintains full sparsity while overcoming the constraints associated with the LiDAR-only fully sparse detector. Our framework showcases state-of-the-art performance on the widely used nuScenes dataset, Waymo Open Dataset, and the long-range Argoverse 2 dataset. Notably, the inference speed of our proposed method under the long-range perception setting is 2.7× faster than that of other state-of-the-art multimodal 3D detection methods. Code is released at https://github.com/BraveGroup/FullySparseFusion.
ABSTRACT
The remarkable performance of recent stereo depth estimation models benefits from the successful use of convolutional neural networks to regress dense disparity. Akin to most tasks, this needs gathering training data that covers a number of heterogeneous scenes at deployment time. However, training samples are typically acquired continuously in practical applications, making the capability to learn new scenes continually even more crucial. For this purpose, we propose to perform continual stereo matching where a model is tasked to 1) continually learn new scenes, 2) overcome forgetting previously learned scenes, and 3) continuously predict disparities at inference. We achieve this goal by introducing a Reusable Architecture Growth (RAG) framework. RAG leverages task-specific neural unit search and architecture growth to learn new scenes continually in both supervised and self-supervised manners. It can maintain high reusability during growth by reusing previous units while obtaining good performance. Additionally, we present a Scene Router module to adaptively select the scene-specific architecture path at inference. Comprehensive experiments on numerous datasets show that our framework performs impressively in various weather, road, and city circumstances and surpasses the state-of-the-art methods in more challenging cross-dataset settings. Further experiments also demonstrate the adaptability of our method to unseen scenes, which can facilitate end-to-end stereo architecture learning and practical deployment.
ABSTRACT
Inspired by human language, machine language is a novel discrete representation learned from visual data only through playing the speak, guess, and draw game.
ABSTRACT
AIM: The goal of this study was to evaluate whether topical administration of tacrolimus (TAC) and mycophenolic acid (MPA) at the transplant site enables vascularized composite allograft (VCA) survival with significant minimization of the dose and adverse effects of systemic TAC (STAC) immunosuppression. MATERIALS AND METHODS: Lewis (Lew) rats received orthotopic hind limb allotransplants from fully mismatched Brown Norway (BN) donors. Group 1 (Controls) received no treatment. Other groups were treated with STAC at a dose of 1 mg/kg/day for 7 days. On post-operative day (POD) 8, the STAC dose was dropped to 0.1 mg/kg/day for Group 2 and maintained at 1 mg/kg for Group 3. Group 4 received topical application of TAC and MPA on the transplanted (Tx) limb starting POD 8 without STAC. Group 5 received topical TAC and MPA on the contralateral non-Tx limb and Group 6 received topical TAC and MPA on the Tx limb starting POD 8 along with low dose STAC (0.1 mg/kg/day). Treatment was continued until the study end point was reached, defined as either grade 3 rejection or allograft survival exceeding 100 days. .We conducted sequential LC-MS/MS measurements to assess TAC and MPA concentrations in both blood/plasma and allograft tissues. Additionally, we evaluated markers indicative of organ toxicity associated with STAC immunosuppression. RESULTS: Compared to controls, topical therapy with TAC+MPA significantly prolonged allograft survival beyond 100 daysat very low dose STAC (0.1 mg/kg/day) (Group 6). The histopathological assessment of the grafts was consistent with the clinical outcomes. .Drug levels in blood/plasma remained low or undetectable, while allograft tissues showed higher drug concentrations compared to contralateral limb tissues (P<0.05). . Urinary creatinine clearance remained within the normal range at 2.5 mL/min. CONCLUSION: Combination therapy with topical TAC and MPA synergizes with a very low dose, corticosteroid- free-STAC regimen and facilitates rejection-free, prolonged VCA survival without morbidity.
Subject(s)
Administration, Topical , Graft Survival , Immunosuppressive Agents , Mycophenolic Acid , Rats, Inbred BN , Rats, Inbred Lew , Tacrolimus , Animals , Tacrolimus/administration & dosage , Tacrolimus/pharmacokinetics , Mycophenolic Acid/administration & dosage , Mycophenolic Acid/pharmacokinetics , Immunosuppressive Agents/administration & dosage , Immunosuppressive Agents/pharmacokinetics , Graft Survival/drug effects , Rats , Male , Graft Rejection/prevention & control , Graft Rejection/immunology , Immunosuppression Therapy/methods , Vascularized Composite Allotransplantation/methods , Drug Synergism , Composite Tissue Allografts/drug effects , AllograftsABSTRACT
Data association is at the core of many computer vision tasks, e.g., multiple object tracking, image matching, and point cloud registration. however, current data association solutions have some defects: they mostly ignore the intra-view context information; besides, they either train deep association models in an end-to-end way and hardly utilize the advantage of optimization-based assignment methods, or only use an off-the-shelf neural network to extract features. In this paper, we propose a general learnable graph matching method to address these issues. Especially, we model the intra-view relationships as an undirected graph. Then data association turns into a general graph matching problem between graphs. Furthermore, to make optimization end-to-end differentiable, we relax the original graph matching problem into continuous quadratic programming and then incorporate training into a deep graph neural network with KKT conditions and implicit function theorem. In MOT task, our method achieves state-of-the-art performance on several MOT datasets. For image matching, our method outperforms state-of-the-art methods on a popular indoor dataset, ScanNet. For point cloud registration, we also achieve competitive results.
ABSTRACT
BACKGROUND: Autologous adipose tissue often experiences ischemia and hypoxia after transplantation, leading to low retention rates and unstable operative impacts due to necrotic absorption. Platelet-rich plasma (PRP) can enhance fat regeneration and increase the fat retention rate after transplantation. However, the quick release of growth factors (GFs) in PRP decreases therapeutic efficiency. This study aimed to achieve a slow release of PRP to promote fat retention. METHODS: We prepared a dual-network hydrogel (DN gel) based on FDA-approved PRP and sodium alginate (SA) through a simple "one-step" activation process. In vivo study, adipose tissue with saline (control group), SA gel (SA gel group), PRP gel (PRP gel group), and DN gel (DN gel group) was injected subcutaneously into the dorsum of nude mice. At 4 and 12 weeks after injection, tissues were assessed for volume and weight. Hematoxylin and eosin staining (HE) and immunofluorescence staining were performed for histological assessment. RESULTS: DN gel exhibits long-lasting growth factor effects, surpassing conventional clinical PRP gel regarding vascularization potential. In fat transplantation experiments, DN gel demonstrated improved vascularization of transplanted fat and increased retention rates, showing promise for clinical applications. CONCLUSIONS: DN gel-assisted lipofilling can significantly improve the retention rate and quality of transplanted fat. DN gel-assisted lipofilling, which is considered convenient, is a promising technique to improve neovascularization and fat survival. NO LEVEL ASSIGNED: This journal requires that authors assign a level of evidence to each submission to which Evidence-Based Medicine rankings are applicable. This excludes Review Articles, Book Reviews, and manuscripts that concern Basic Science, Animal Studies, Cadaver Studies, and Experimental Studies. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266.
Subject(s)
Adipose Tissue , Platelet-Rich Plasma , Animals , Mice , Mice, Nude , Adipose Tissue/transplantation , InjectionsABSTRACT
Recent developments in the field of regenerative surgeries and medical applications have led to a renewed interest in adipose tissue-enriched mesenchymal stem cell scaffolds. Various advantages declared for the decellularized adipose matrix (DAM) have caused its extensive use in the transfer of stem cells or growth factors for soft tissue regeneration induction. Meanwhile, the long-term application of detergents toward DAM regeneration has been assumed as a risky obstacle in this era. Herein, a rapid, mechanical protocol was developed to prepare DAM (M-DAM) without chemicals/enzymes and was comprehensively compared with the ordinary DAM (traditional chemical method). Accordingly, this method could effectively hinder oils and cells, sustain the structural and biological elements, and contain a superior level of collagen content. In addition, more protein numbers, as well as higher basement membrane elements, glycoproteins, and extracellular matrix-related proteins were detected in the regenerated M-DAM. Also, superior adipogenesis and angiogenesis proteins were distinguished. The noncytotoxicity of the M-DAM was also approved, and a natural ecological niche was observed for the proliferation and differentiation of stem cells, confirming its great potential for vascularization and adipogenesis in vivo. The suggested technique could effectively prepare the modified DAM in variant constructions of tablets, powders, emulsions, hydrogels, and different three-dimensional-printed structures. Hence, this rapid, mechanical process can produce bioactive DAM, which has the potential to be widely used in various research fields of regenerative medicine.
Subject(s)
Adipogenesis , Tissue Scaffolds , Humans , Tissue Scaffolds/chemistry , Extracellular Matrix/metabolism , Adipose Tissue , Cell Differentiation , Obesity/metabolism , Tissue Engineering/methodsABSTRACT
Interactive object segmentation aims to produce object masks with user interactions, such as clicks, bounding boxes, and scribbles. Click point is the most popular interactive cue for its efficiency, and related deep learning methods have attracted lots of interest in recent years. Most works encode click points as gaussian maps and concatenate them with images as the model's input. However, the spatial and semantic information of gaussian maps would be noised through multiple convolution layers and won't be fully exploited by top layers for mask prediction. To pass click information to top layers exactly and efficiently, we propose a coarse mask guided model (CMG) which predicts coarse masks with a coarse module to guide the object mask prediction. Specifically, the coarse module encodes user clicks as query features and enriches their semantic information with backbone features through transformer layers, coarse masks are generated based on the enriched query feature and fed into CMG's decoder. Benefiting from the efficiency of transformer, CMG's coarse module and decoder module are lightweight and computationally efficient, making the interaction process more smooth. Experiments on several segmentation benchmarks demonstrate the effectiveness of our method, and we get new state-of-the-art results compared with previous works.
ABSTRACT
Few-shot learning aims to recognize novel categories solely relying on a few labeled samples, with existing few-shot methods primarily focusing on the categories sampled from the same distribution. Nevertheless, this assumption cannot always be ensured, and the actual domain shift problem significantly reduces the performance of few-shot learning. To remedy this problem, we investigate an interesting and challenging cross-domain few-shot learning task, where the training and testing tasks employ different domains. Specifically, we propose a Meta-Memory scheme to bridge the domain gap between source and target domains, leveraging style-memory and content-memory components. The former stores intra-domain style information from source domain instances and provides a richer feature distribution. The latter stores semantic information through exploration of knowledge of different categories. Under the contrastive learning strategy, our model effectively alleviates the cross-domain problem in few-shot learning. Extensive experiments demonstrate that our proposed method achieves state-of-the-art performance on cross-domain few-shot semantic segmentation tasks on the COCO-20 i, PASCAL-5 i, FSS-1000, and SUIM datasets and positively affects few-shot classification tasks on Meta-Dataset.