Publications by authors named "Xiuhong Lin"

8 Publications

  • Page 1 of 1

DDA-Net: Unsupervised cross-modality medical image segmentation via dual domain adaptation.

Comput Methods Programs Biomed 2021 Nov 14;213:106531. Epub 2021 Nov 14.

Fujian Key Laboratory of Sensing and Computing for Smart Cities, Department of Computer Science, School of Informatics, Xiamen University, Xiamen 361005, China. Electronic address:

Background And Objective: Deep convolutional networks are powerful tools for single-modality medical image segmentation, whereas generally require semantic labelling or annotation that is laborious and time-consuming. However, domain shift among various modalities critically deteriorates the performance of deep convolutional networks if only trained by single-modality labelling data.

Methods: In this paper, we propose an end-to-end unsupervised cross-modality segmentation network, DDA-Net, for accurate medical image segmentation without semantic annotation or labelling on the target domain. To close the domain gap, different images with domain shift are mapped into a shared domain-invariant representation space. In addition, spatial position information, which benefits the spatial structure consistency for semantic information, is preserved by an introduced cross-modality auto-encoder.

Results: We validated the proposed DDA-Net method on cross-modality medical image datasets of brain images and heart images. The experimental results show that DDA-Net effectively alleviates domain shift and suppresses model degradation.

Conclusions: The proposed DDA-Net successfully closes the domain gap between different modalities of medical image, and achieves state-of-the-art performance in cross-modality medical image segmentation. It also can be generalized for other semi-supervised or unsupervised segmentation tasks in some other field.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.cmpb.2021.106531DOI Listing
November 2021

A deep learning model for detection and tracking in high-throughput images of organoid.

Comput Biol Med 2021 07 25;134:104490. Epub 2021 May 25.

Fujian Key Laboratory of Sensing and Computing for Smart City, School of Informatics, Xiamen University, Xiamen, 361005, China. Electronic address:

Organoid, an in vitro 3D culture, has extremely high similarity with its source organ or tissue, which creates a model in vitro that simulates the in vivo environment. Organoids have been extensively studied in cell biology, precision medicine, drug toxicity, efficacy tests, etc., which have been proven to have high research value. Periodic observation of organoids in microscopic images to obtain morphological or growth characteristics is essential for organoid research. It is difficult and time-consuming to perform manual screens for organoids, but there is no better solution in the prior art. In this paper, we established the first high-throughput organoid image dataset for organoids detection and tracking, which experienced experts annotate in detail. Moreover, we propose a novel deep neural network (DNN) that effectively detects organoids and dynamically tracks them throughout the entire culture. We divided our solution into two steps: First, the high-throughput sequential images are processed frame by frame to detect all organoids; Second, the similarities of the organoids in the adjacent frames are computed, and the organoids on the adjacent frames are matched in pairs. With the help of our proposed dataset, our model achieves organoids detection and tracking with fast speed and high accuracy, effectively reducing the burden on researchers. To our knowledge, this is the first exploration of applying deep learning to organoid tracking tasks. Experiments have demonstrated that our proposed method achieved satisfactory results on organoid detection and tracking, verifying the great potential of deep learning technology in this field.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.compbiomed.2021.104490DOI Listing
July 2021

An optimal glycemic load range is better for reducing obesity and diabetes risk among middle-aged and elderly adults.

Nutr Metab (Lond) 2021 Mar 22;18(1):31. Epub 2021 Mar 22.

Department of Endocrinology, Sun Yat-Sen Memorial Hospital of Sun Yat-Sen University, Guangzhou, 510120, China.

Background: Due to the lack of evidence, advice pertaining to glycemic load (GL) can be misleading. Does the excessive restriction of GL, mostly through an extreme reduction in carbohydrate intake, result in a relatively high intake of fat and protein and result in overweight and obesity? This study was performed to initially explore the optimal GL range.

Methods: A cross-sectional study involving 2029 participants aged 40 years or older in Guangzhou, China was conducted. Participants were divided into four groups according to cluster analysis. Dietary data were assessed using a previously validated 3-day food record.

Results: Instead of participants with the highest [cluster 1, median (interquartile ranges) GL was 112(107-119)/1000 kcal] and the lowest GL intake [cluster 4, 90(82-96)/1000 kcal], those with moderate GL intakes [clusters 2 and 3, 93(85-102) and 93(85-99)/1000 kcal, respectively] had a lower prevalence of overweight, obesity and diabetes. In addition, clusters 2 and 3 were more consistent with the macronutrient intake reference with adequate micronutrient intake. Therefore, the optimal GL range was determined to be (85-100)/1000 kcal, rather than "lower is better".

Conclusions: Reducing the GL intake to prevent diabetes deserves more attention in the context of a balanced diet. An appropriate GL may be better than excessive restriction.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12986-020-00504-5DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7983372PMC
March 2021

Yunvjian-Medicated Serum Protects INS-1 Cells against Glucolipotoxicity-Induced Apoptosis through Autophagic Flux Modulation.

Evid Based Complement Alternat Med 2020 14;2020:8878259. Epub 2020 Dec 14.

Department of Histology and Embryology, Fujian University of Traditional Chinese Medicine, Fuzhou, Fujian 350122, China.

Yunvjian (YNJ) is a traditional Chinese medicine formula adopted to prevent and treat diabetes. Our previous results from animal experiments showed that YNJ decreased blood glucose. This study aimed to examine the effect of high glucose and high lipid (HG/HL) conditions on the proliferation and apoptosis of INS-1 cells and the possible protective mechanism of YNJ-medicated serum on INS-1 cells exposed to HG/HL conditions. INS-1 cells were cultured in RPMI 1640 medium after being passaged. Then, INS-1 cells in the logarithmic growth phase were collected and divided into five groups: control, HG/HL, HG/HL+5% YNJ-medicated serum, HG/HL+10% YNJ-medicated serum, and HG/HL+20% YNJ-medicated serum. MTT assay and flow cytometry were used to detect proliferation and apoptosis of INS-1 cells, respectively. Protein profiles of INS-1 cells were analyzed using a tandem mass tag (TMT) label-based quantitative proteomic approach. Western blotting was performed to verify the proteomic results. YNJ-medicated serum significantly promoted INS-1 cell proliferation and inhibited apoptosis. Proteomic results from the INS-1 cells in the control, HG/HL, and HG/HL+10% YNJ-medicated serum groups showed that 7,468 proteins were identified, of which 6,423 proteins were quantified. Compared with the HG/HL group,430 differential proteins were upregulated, and 671 were downregulated in the HG/HL+10% YNJ-medicated serum group. Compared with the control group, 711 differential proteins were upregulated and 455 were downregulated in the HG/HL group, whereas 10 differential proteins were upregulated and 9 were downregulated in the HG/HL+10% YNJ-medicated serum group. Furthermore, several proteins related to autophagy, including ATG3, ATG2B, GABARAP, WIPI2, and p62/SQSTM1, were verified by western blotting, and these results were consistent with the results obtained from the proteomics analysis. These results confirmed that the autophagy pathway is critical to glucolipotoxicity in INS-1 cells. YNJ-medicated serum exhibited a protective effect on INS-1 cells cultured under HG/HL conditions by regulating autophagy genes' expression and restoring the autophagic flux.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1155/2020/8878259DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7752277PMC
December 2020

Optic disc and optic cup segmentation based on anatomy guided cascade network.

Comput Methods Programs Biomed 2020 Dec 27;197:105717. Epub 2020 Aug 27.

Fujian Key Laboratory of Sensing and Computing for Smart Cities, Department of Computer Science, School of Informatics, Xiamen University, Xiamen 361005, China.

Background And Objective: Glaucoma, a worldwide eye disease, may cause irreversible vision damage. If not treated properly at an early stage, glaucoma eventually deteriorates into blindness. Various glaucoma screening methods, e.g. Ultrasound Biomicroscopy (UBM), Optical Coherence Tomography (OCT), and Heidelberg Retinal Scanner (HRT), are available. However, retinal fundus image photography examination, because of its low cost, is one of the most common solutions used to diagnose glaucoma. Clinically, the cup-to-disk ratio is an important indicator in glaucoma diagnosis. Therefore, precise fundus image segmentation to calculate the cup-to-disk ratio is the basis for screening glaucoma.

Methods: In this paper, we propose a deep neural network that uses anatomical knowledge to guide the segmentation of fundus images, which accurately segments the optic cup and the optic disc in a fundus image to accurately calculate the cup-to-disk ratio. Optic disc and optic cup segmentation are typical small target segmentation problems in biomedical images. We propose to use an attention-based cascade network to effectively accelerate the convergence of small target segmentation during training and accurately reserve detailed contours of small targets.

Results: Our method, which was validated in the MICCAI REFUGE fundus image segmentation competition, achieves 93.31% dice score in optic disc segmentation and 88.04% dice score in optic cup segmentation. Moreover, we win a high CDR evaluation score, which is useful for glaucoma screening.

Conclusions: The proposed method successfully introduce anatomical knowledge into segmentation task, and achieve state-of-the-art performance in fundus image segmentation. It also can be used for both automatic segmentation and semiautomatic segmentation with human interaction.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.cmpb.2020.105717DOI Listing
December 2020

Association of Serum Retinol-Binding Protein 4 Levels and the Risk of Incident Type 2 Diabetes in Subjects With Prediabetes.

Diabetes Care 2019 08 11;42(8):1574-1581. Epub 2019 Jun 11.

Guangdong Provincial Key Laboratory of Food, Nutrition, and Health, Guangdong Engineering Technology Research Center of Nutrition Translation, Department of Nutrition, School of Public Health, Sun Yat-sen University (Northern Campus), Guangzhou, Guangdong Province, China

Objective: To explore the association of serum retinol-binding protein 4 (RBP4) levels and risk for the development of type 2 diabetes in individuals with prediabetes.

Research Design And Methods: A population-based prospective study was conducted among 1,011 Chinese participants with prediabetes (average age 55.6 ± 7.2 years). Incident type 2 diabetes was diagnosed according to the American Diabetes Association 2010 criteria. Serum RBP4 levels were measured using a commercially available ELISA. We analyzed the association of serum RBP4 levels with the risk of incident type 2 diabetes using the Cox proportional hazards model.

Results: During a median follow-up period of 3.1 years, 153 participants developed incident type 2 diabetes. A U-shaped association was observed between serum RBP4 levels and the risk of incident type 2 diabetes, with the lowest risk in the RBP4 range of 31-55 μg/mL. Multivariate Cox regression model analysis showed that serum RBP4 levels <31 μg/mL and RBP4 levels >55 μg/mL were associated with an increased risk of incident type 2 diabetes. The adjusted hazard ratios (95% CI) were 2.01 (1.31-3.09) and 1.97 (1.32-2.93), respectively, after adjusting for age, sex, BMI, waist circumference, γ-glutamyltransferase, HOMA of insulin resistance index, fasting plasma glucose, 2-h plasma glucose, and glycated hemoglobin (HbA) levels.

Conclusions: A U-shaped relationship exists between serum RBP4 levels and the risk of incident type 2 diabetes in subjects with prediabetes.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2337/dc19-0265DOI Listing
August 2019

Dietary glycemic load and metabolic status in newly diagnosed type 2 diabetes in southeastern China.

Asia Pac J Clin Nutr 2018;27(2):375-382

Department of Endocrinology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China.

Background And Objectives: Large-scale epidemiological investigations worldwide have shown that dietary glycemic load is associated with metabolic diseases, including diabetes. However, only a few studies have examined the correlations between glycemic load and blood glucose and lipids in Chinese diabetic patients. Therefore, this study aimed to determine these correlations in southeastern China.

Methods And Study Design: 201 patients with newly diagnosed type 2 diabetes and 126 participants with normal blood glucose were enrolled at the Sun Yat-sen Memorial Hospital, Guangdong Province. Carbohydrate intake and glycemic load were assessed based on 3-day dietary records. Using glycemic load as the dependent variable, a correlation analysis and multiple regression analyses were used to analyze the correlations between glycemic load and blood glucose and lipids.

Results: The mean glycemic load in diabetic patients was significantly higher than that in the control group (p<0.05). Correlation analysis showed that glycemic load was positively correlated with body mass index and glycated hemoglobin in diabetic patients (p<0.05) but negatively correlated with high-density lipoprotein cholesterol in all subjects (p<0.05). Multivariable regression analysis indicated that, among participants in southeastern China, a higher glycemic load increased the odds of having diabetes, a low high-density lipoprotein cholesterol, and higher Charlson weighted index of comorbidities score, as well as being overweight.

Conclusions: A high-glycemic load diet may be associated with a risk of diabetes, glycemic control, lipid metabolism, prognosis of diseases, and body composition. It is necessary to control dietary glycemic load for both patients with diabetes and healthy people in southeastern China.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.6133/apjcn.052017.03DOI Listing
September 2019

The intriguing effects of time to glycemic goal in newly diagnosed type 2 diabetes after short-term intensive insulin therapy.

Endocr J 2016 Aug 22;63(8):739-46. Epub 2016 Jun 22.

Department of Endocrinology, Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou 510120, China.

Short-term intensive insulin therapy is effective for type 2 diabetes because it offers the potential to achieve excellent glycemic control and improve β-cell function. We observed that the time to glycemic goal (TGG) was adjustable. Original data of 138 newly diagnosed type 2 diabetic patients received intensive insulin therapy by continuous subcutaneous insulin infusion for 2-3 weeks were retrospectively collected. Subjects underwent an intravenous glucose tolerance test (IVGTT) and an oral glucose tolerance test (OGTT) pre and post treatment. The glycemic goal was achieved within 6 (4-8) days. Patients were divided into two groups by TGG above (TGG-slow) and below (TGG-fast) the median value. Patients in both groups had significantly better glycemic control. Compared with TGG-fast, TGG-slow required a few more total insulin and performed more improvement of HOMA-β and IVGTT-AUCIns, but less improvement of HOMA-IR and QUICKI. Multiple linear regression analysis revealed that TGG was always an explanatory variable for the changes (HOMA-β, IVGTT-AUCIns, HOMA-IR and QUICKI). The hypoglycemia prevalence was lower in TGG-slow (1.48% vs. 3.40%, P<0.01). Multivariate logistic regression analysis indicated that individuals in TGG-slow had a lower risk of hypoglycemia (adjusted OR, 0.700; 95% CI, 0.567-0.864; P<0.05). Multiple linear regression analysis confirmed that the ratio of the incremental insulin to glucose responses over the first 30 min during OGTT (ΔIns30/ΔG30), average insulin dose before achieving targets, initial insulin dose and LDL-c were independent predictors for TGG. It is intriguing to hypothesize that patients with fast time to glycemic goal benefit more in improving insulin sensitivity, but patients with slow time benefit more in improving β-cell function and reducing the risk of hypoglycemia.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1507/endocrj.EJ16-0154DOI Listing
August 2016
-->