Shushui Wang

Dr. Shushui Wang

Department of Pediatric Cardiology
+4

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 August 2021 in Orphanet Journal of Rare Diseases
Verified
Journal Article
Open access
PDF file
DHDian Hong
Dian Hong
YZYing-Yi Zheng
Ying-Yi Zheng
YXYing Xin
Ying Xin
LSLing Sun
Ling Sun
HYHang Yang
Hang Yang
SWShushui Wang
Shushui Wang
Background: Many genetic syndromes (GSs) have distinct facial dysmorphism, and facial gestalts can be used as a diagnostic tool for recognizing a syndrome. Facial recognition technology has advanced in recent years, and the screening of GSs by facial recognition technology has become feasible. This study constructed an automatic facial recognition model for the identification of children with GSs. Results: A total of 456 frontal facial photos were collected from 228 children with GSs and 228 healthy children in Guangdong Provincial People's Hospital from Jun 2016 to Jan 2021. Only one frontal facial image was selected for each participant. The VGG-16 network (named after its proposal lab, Visual Geometry Group from Oxford University) was pretrained by transfer learning methods, and a facial recognition model based on the VGG-16 architecture was constructed. The performance of the VGG-16 model was evaluated by five-fold cross-validation. Comparison of VGG-16 model to five physicians were also performed. The VGG-16 model achieved the highest accuracy of 0.8860 ± 0.0211, specificity of 0.9124 ± 0.0308, recall of 0.8597 ± 0.0190, F1-score of 0.8829 ± 0.0215 and an area under the receiver operating characteristic curve of 0.9443 ± 0.0276 (95% confidence interval: 0.9210–0.9620) for GS screening, which was significantly higher than that achieved by human experts. Conclusions: This study highlighted the feasibility of facial recognition technology for GSs identification. The VGG-16 recognition model can play a prominent role in GSs screening in clinical practice.

13 June 2022 in Frontiers in Genetics
Verified
Journal Article
Open access
LSLing Sun
Ling Sun
YXYu-Mei Xie
Yu-Mei Xie
SWShushui Wang
Shushui Wang
ZZZhi-Wei Zhang&Ast
Zhi-Wei Zhang&Ast
Background: Common cardiac abnormalities in Noonan syndrome (NS) include congenital heart diseases (CHD), pulmonary valve stenosis and hypertrophic cardiomyopathy (HCM). Molecular diagnoses are enabling earlier and more precise diagnosis of patients who have a subtle or atypical presentation. The aims of this study were to investigate genotype-phenotype associations with respect to Noonan syndrome (NS)-associated cardiac abnormalities and catheter or surgery-based interventions conditions.Methods: From January 2019 to December 2021, 22 children with a confirmed molecular diagnosis of NS combined with cardiovascular abnormalities were consecutively enrolled into the current study. A comprehensive review was carried out of echocardiography and electrocardiogram results, second-generation whole-exome sequencing results and catheter or surgery-based interventions conditions.Results: The main manifestations of electrocardiogram abnormalities were QTc prolongation, abnormal Q wave in the precordial lead and limb lead, right ventricular hypertrophy and left or right deviation of the electrical axis. The most commonly detected abnormality was pulmonary valve dysplasia with stenosis, seen in 15 (68.2%) patients, followed by atrial septal defect in 11 (50%) patients. Seven genes (RAF1, RIT1, SOS1, PTPN11, BRAF, SOS2, and LZTR1) were found to contain disease-associated variants. The most commonly observed genetic mutations were PTPN11 (27%) and RAF1 (27%). Each genotype was associated with specific phenotypic findings. RIT1, SOS1, PTPN11, and SOS2 had common echocardiography features characterized by pulmonary valve stenosis, while RAF1 was characterized by HCM. Interestingly, patients with BRAF mutations were not only characterized by HCM, but also by pulmonary valve stenosis. In the cohort there was only one patient carrying a LZTR1 mutation characterized by left ventricle globose dilation. Ten cases underwent catheter or surgery-based interventions. All the operations had immediate results and high success rates. However, some of the cases had adverse outcomes during extended follow-up. Based on the genotype-phenotype associations observed during follow-up, BRAF and RAF1 genotypes seem to be poor prognostic factors, and multiple interventions may be required for NS patients with severe pulmonary stenosis or myectomy for HCM.Conclusions: The identification of causal genes in NS patients has enabled the evaluation of genotype-cardiac phenotype relationships and prognosis of the disease. This may be beneficial for the development of therapeutic approaches.

19 May 2021 in Frontiers in Pediatrics
Verified
Journal Article
Open access
PDF file
HLHui Liu
Hui Liu
ZMZi-Hua Mo
Zi-Hua Mo
HYHang Yang
Hang Yang
ZZZheng-Fu Zhang
Zheng-Fu Zhang
DHDian Hong
Dian Hong
SWShushui Wang
Shushui Wang
Background: Williams-Beuren syndrome (WBS) is a rare genetic syndrome with a characteristic “elfin” facial gestalt. The “elfin” facial characteristics include a broad forehead, periorbital puffiness, flat nasal bridge, short upturned nose, wide mouth, thick lips, and pointed chin. Recently, deep convolutional neural networks (CNNs) have been successfully applied to facial recognition for diagnosing genetic syndromes. However, there is little research on WBS facial recognition using deep CNNs.Objective: The purpose of this study was to construct an automatic facial recognition model for WBS diagnosis based on deep CNNs.Methods: The study enrolled 104 WBS children, 91 cases with other genetic syndromes, and 145 healthy children. The photo dataset used only one frontal facial photo from each participant. Five face recognition frameworks for WBS were constructed by adopting the VGG-16, VGG-19, ResNet-18, ResNet-34, and MobileNet-V2 architectures, respectively. ImageNet transfer learning was used to avoid over-fitting. The classification performance of the facial recognition models was assessed by five-fold cross validation, and comparison with human experts was performed.Results: The five face recognition frameworks for WBS were constructed. The VGG-19 model achieved the best performance. The accuracy, precision, recall, F1 score, and area under curve (AUC) of the VGG-19 model were 92.7 ± 1.3%, 94.0 ± 5.6%, 81.7 ± 3.6%, 87.2 ± 2.0%, and 89.6 ± 1.3%, respectively. The highest accuracy, precision, recall, F1 score, and AUC of human experts were 82.1, 65.9, 85.6, 74.5, and 83.0%, respectively. The AUCs of each human expert were inferior to the AUCs of the VGG-16 (88.6 ± 3.5%), VGG-19 (89.6 ± 1.3%), ResNet-18 (83.6 ± 8.2%), and ResNet-34 (86.3 ± 4.9%) models.Conclusions: This study highlighted the possibility of using deep CNNs for diagnosing WBS in clinical practice. The facial recognition framework based on VGG-19 could play a prominent role in WBS diagnosis. Transfer learning technology can help to construct facial recognition models of genetic syndromes with small-scale datasets.

27 May 2022 in eClinicalMedicine
Verified
Journal Article
Open access
WCWenxiu Chan
Wenxiu Chan
SYShiwei Yang
Shiwei Yang
JWJian Wang
Jian Wang
STShilu Tong
Shilu Tong
MLMinyin Lin
Minyin Lin
LFLijun Fu
Lijun Fu
Few data on paediatric hypertrophic cardiomyopathy (HCM) are available in developing countries. A multicentre, retrospective, cohort study was conducted to profile the clinical characteristics and survival of children with HCM in China. We collected longitudinal data on children with HCM aged 0–18 years at three participating institutions between January 1, 2010 and December 31, 2019. Patients were identified by searching for the diagnosis using ICD-10 codes from the electronic medical records database. HCM was diagnosed morphologically with echocardiography or cardiovascular magnetic resonance imaging. The exclusion criteria were secondary aetiologies of myocardial hypertrophy. The primary outcomes were all-cause death or heart transplantation. The Kaplan–Meier method was used to estimate the survival rate of different groups. A total of 564 children were recruited, with a median age at diagnosis of 1.0 year (interquartile range, IQR: 0.4–8.0 years), followed for a median of 2.6 years (1977 patient-years, IQR:0.5, 5.9 years). The underlying aetiology was sarcomeric (382, 67.7%), inborn errors of metabolism (IEMs) (108, 19.2%), and RASopathies (74, 13.1%). A total of 149 patients (26.4%) died and no patients underwent heart transplantation during follow-up. The survival probability was 71.1% (95% confidence interval [CI], 66.3%–75.3%) at 5 years. Patients with IEMs or those diagnosed during infancy had the poorest outcomes, with an estimated 5-year survival rate of 16.9% (95% CI, 7.7%–29.1%) and 56.0% (95% CI, 48.8%–62.5%), respectively. Heart failure was the leading cause of death in the cohort (90/149, 60.4%), while sudden cardiac death was the leading cause in patients with sarcomeric HCM (32/66, 48.5%). There is a high proportion of patients with IEM and a low proportion of patients with neuromuscular disease in children with HCM in China. Overall, mortality remains high in China, especially in patients with IEMs and those diagnosed during infancy. National Natural Science Fund of China (81770380, 81974029), China Project of Shanghai Municipal Science and Technology Commission (20MC1920400, 21Y31900301).

30 July 2021 in Journal of Interventional Cardiology
Verified
Journal Article
Open access
PDF file
YLYifan Li
Yifan Li
YXYumei Xie
Yumei Xie
BLBoning Li
Boning Li
ZXZhaofeng Xie
Zhaofeng Xie
JSJunjun Shen
Junjun Shen
ZZZhiwei Zhang
Zhiwei Zhang
Objective. We reported the 3-year follow-up results of initial clinical experience with the AbsnowTM device, a novel biodegradable occluder for percutaneous closure of atrial septal defect (ASD). Background. The AbsnowTM device is a total biodegradable septal occluder with double-disc poly-L-lactic acid (PLLA) framework and PLLA membranes intergraded into the device to ensure its biodegradability, clinical safety, and efficacy. Methods. Five pediatric patients were enrolled from May to June 2018 in our institution and were followed up for 3 years. A clinical evaluation and transthoracic echocardiography were performed at 24hr, 1 month, 3 months, 6 months, 12 months, and yearly after implantation. Primary endpoints were a composite clinical success, comprising of clinical closure success and safety at the 36-month follow-up evaluation. Secondary endpoints included technical success, procedure success, closure success, and safety at each of the follow-up visits. Results. The median subject age was 3.6 years (range 3.16.5 years). The mean ASD diameter was (13.7±2.9) mm. The median device size was 20mm (range 14 to 24mm). Technical and procedure success was achieved in 100 (5/5) of the patients. At 2-year follow-up, 3 of the 5 patients developed new-onset residual shunts and 2 of them reached a moderate degree. At 3-year follow-up, the residual shunt size increased over time in all the 3 patients, and 1 of them had right ventricular enlargement. All of the 5 patients were free from serious adverse events during the 3-year follow-up, with no device embolization, thromboembolization, or reintervention to the target defect. Conclusion. This 3-year follow-up result of initial experience with the biodegradable AbsnowTM device has demonstrated acceptable safety with no procedural complications. Notably, the high rate of residual shunt significantly affected its efficacy. The long-term safety and efficacy of the device should be further evaluated in a large cohort of patients in future studies.

1 August 2015 in Respiratory Medicine
Verified
Journal Article
Closed access
JLJing Li
Jing Li
XGXinglin Gao
Xinglin Gao
JCJingjing Chen
Jingjing Chen
MLMiaochan Lao
Miaochan Lao
SWShushui Wang
Shushui Wang
GZGuohong Zeng
Guohong Zeng