Image Segmentation by Networks of Spiking Neurons

2005 ◽  
Vol 17 (5) ◽  
pp. 1010-1031 ◽  
Author(s):  
Joachim M. Buhmann ◽  
Tilman Lange ◽  
Ulrich Ramacher
Keyword(s):  
Image Segmentation ◽  
Real World ◽  
Network Architecture ◽  
Spiking Neurons ◽  
Local Smoothing ◽  
Class Membership ◽  
Local Competition ◽  
Integrate And Fire ◽  
Clustering Approach ◽  
The Mean

A network of leaky integrate-and-fire (IAF) neurons is proposed to segment gray-scale images. The network architecture with local competition between neurons that encode segment assignments of image blocks is motivated by a histogram clustering approach to image segmentation. Lateral excitatory connections between neighboring image sites yield a local smoothing of segments. The mean firing rate of class membership neurons encodes the image segmentation. A weight modification scheme is proposed that estimates segment-specific prototypical histograms. The robustness properties of the network implementation make it amenable to an analog VLSI realization. Results on synthetic and real-world images demonstrate the effectiveness of the architecture.

scholarly journals AB1346-HPR REAL-WORLD EFFECTIVENESS AND PERCEIVED USEFULNESS OF SYMPTOM CHECKERS IN RHEUMATOLOGY: INTERIM REPORT FROM THE PROSPECTIVE MULTICENTER BETTER STUDY

2020 ◽  
Vol 79 (Suppl 1) ◽  
pp. 1961.1-1961
Author(s):  
J. Knitza ◽  
J. Mohn ◽  
C. Bergmann ◽  
E. Kampylafka ◽  
M. Hagen ◽  
...  
Keyword(s):  
Diagnostic Accuracy ◽  
Real World ◽  
Search Engines ◽  
Correct Diagnosis ◽  
Perceived Usefulness ◽  
Symptom Duration ◽  
Research Support ◽  
Link Type ◽  
The Mean ◽  
Research Grant

Background:Symptom checkers (SC) promise to reduce diagnostic delay, misdiagnosis and effectively guide patients through healthcare systems. They are increasingly used, however little evidence exists about their real-life effectiveness.Objectives:The aim of this study was to evaluate the diagnostic accuracy, usage time, usability and perceived usefulness of two promising SC, ADA (www.ada.com) and Rheport (www.rheport.de). Furthermore, symptom duration and previous symptom checking was recorded.Methods:Cross-sectional interim clinical data from the first of three recruiting centers from the prospective, real-world, multicenter bETTeR-study (DKRS DRKS00017642) was used. Patients newly presenting to a secondary rheumatology outpatient clinic between September and December 2019 completed the ADA and Rheport SC. The time and answers were recorded and compared to the patient’s actual diagnosis. ADA provides up to 5 disease suggestions, Rheport calculates a risk score for rheumatic musculoskeletal diseases (RMDs) (≥1=RMD). For both SC the sensitivity, specificity was calculated regarding RMDs. Furthermore, patients completed a survey evaluating the SC usability using the system usability scale (SUS), perceived usefulness, previous symptom checking and symptom duration.Results:Of the 129 consecutive patients approached, 97 agreed to participate. 38% (37/97) of the presenting patients presented with an RMD (Figure 1). Mean symptom duration was 146 weeks and a mean number of 10 physician contacts occurred previously, to evaluate current symptoms. 56% (54/96) had previously checked their symptoms on the internet using search engines, spending a mean of 6 hours. Rheport showed a sensitivity of 49% (18/37) and specificity of 58% (35/60) concerning RMDs. ADA’s top 1 and top 5 disease suggestions concerning RMD showed a sensitivity of 43% (16/37) and 54% (20/37) and a specificity of 58% (35/60) and 52% (31/60), respectively. ADA listed the correct diagnosis of the patients with RMDs first or within the first 5 disease suggestions in 19% (7/37) and 30% (11/37), respectively. The average perceived usefulness for checking symptoms using ADA, internet search engines and Rheport was 3.0, 3.5 and 3.1 on a visual analog scale from 1-5 (5=very useful). 61% (59/96) and 64% (61/96) would recommend using ADA and Rheport, respectively. The mean SUS score of ADA and Rheport was 72/100 and 73/100. The mean usage time for ADA and Rheport was 8 and 9 minutes, respectively.Conclusion:This is the first prospective, real-world, multicenter study evaluating the diagnostic accuracy and other features of two currently used SC in rheumatology. These interim results suggest that diagnostic accuracy is limited, however SC are well accepted among patients and in some cases, correct diagnosis can be provided out of the pocket within few minutes, saving valuable time.Figure:Acknowledgments:This study was supported by an unrestricted research grant from Novartis.Disclosure of Interests:Johannes Knitza Grant/research support from: Research Grant: Novartis, Jacob Mohn: None declared, Christina Bergmann: None declared, Eleni Kampylafka Speakers bureau: Novartis, BMS, Janssen, Melanie Hagen: None declared, Daniela Bohr: None declared, Elizabeth Araujo Speakers bureau: Novartis, Lilly, Abbott, Matthias Englbrecht Grant/research support from: Roche Pharma, Chugai Pharma Europe, Consultant of: AbbVie, Roche Pharma, RheumaDatenRhePort GbR, Speakers bureau: AbbVie, Celgene, Chugai Pharma Europe, Lilly, Mundipharma, Novartis, Pfizer, Roche Pharma, UCB, David Simon Grant/research support from: Else Kröner-Memorial Scholarship, Novartis, Consultant of: Novartis, Lilly, Arnd Kleyer Consultant of: Lilly, Gilead, Novartis,Abbvie, Speakers bureau: Novartis, Lilly, Timo Meinderink: None declared, Wolfgang Vorbrüggen: None declared, Cay-Benedict von der Decken: None declared, Stefan Kleinert Shareholder of: Morphosys, Grant/research support from: Novartis, Consultant of: Novartis, Speakers bureau: Abbvie, Novartis, Celgene, Roche, Chugai, Janssen, Andreas Ramming Grant/research support from: Pfizer, Novartis, Consultant of: Boehringer Ingelheim, Novartis, Gilead, Pfizer, Speakers bureau: Boehringer Ingelheim, Roche, Janssen, Jörg Distler Grant/research support from: Boehringer Ingelheim, Consultant of: Boehringer Ingelheim, Paid instructor for: Boehringer Ingelheim, Speakers bureau: Boehringer Ingelheim, Peter Bartz-Bazzanella: None declared, Georg Schett Speakers bureau: AbbVie, BMS, Celgene, Janssen, Eli Lilly, Novartis, Roche and UCB, Axel Hueber Grant/research support from: Novartis, Lilly, Pfizer, Consultant of: Abbvie, BMS, Celgene, Gilead, GSK, Lilly, Novartis, Speakers bureau: GSK, Lilly, Novartis, Martin Welcker Grant/research support from: Abbvie, Novartis, UCB, Hexal, BMS, Lilly, Roche, Celgene, Sanofi, Consultant of: Abbvie, Actelion, Aescu, Amgen, Celgene, Hexal, Janssen, Medac, Novartis, Pfizer, Sanofi, UCB, Speakers bureau: Abbvie, Aescu, Amgen, Biogen, Berlin Chemie, Celgene, GSK, Hexal, Mylan, Novartis, Pfizer, UCB

scholarly journals Real-world effectiveness of app-based treatment for urinary incontinence: a cohort study

BMJ Open ◽  
2021 ◽  
Vol 11 (1) ◽  
pp. e040819
Author(s):  
Pontus Rygh ◽  
Ina Asklund ◽  
Eva Samuelsson
Keyword(s):  
Urinary Incontinence ◽  
Cohort Study ◽  
Real World ◽  
Short Form ◽  
Pelvic Floor Muscle ◽  
Controlled Trial ◽  
Healthcare Services ◽  
International Consultation ◽  
The Mean

ObjectivesThe efficacy of app-based treatment for stress urinary incontinence (SUI) has been demonstrated in a randomised controlled trial (RCT). In this study, we investigate the user characteristics and the effectiveness of the same app when freely available, and compare these results with the RCT.DesignProspective cohort study.ParticipantsDuring a 17-month period, 24 602 non-pregnant, non-postpartum women older than 18 years downloaded the app and responded anonymously to a questionnaire. Of these, 2672 (11%) responded to the 3-month follow-up.InterventionThree months’ use of the app Tät, containing information, a pelvic floor muscle training programme and lifestyle advice.Main outcome measuresChange in symptom severity (International Consultation on Incontinence Questionnaire-Urinary Incontinence Short Form (ICIQ-UI SF)) and subjective improvement (Patient Global Impression of Improvement (PGI-I)).ResultsOf the respondents, 88% lived in Sweden and 75% (18 384/24 602) were incontinent with a mean age of 45.5 (SD 14.1) years. The UI types, based on symptoms, were SUI (53%), urgency UI (12%), mixed UI (31%) and undefined (4%). The mean ICIQ-UI SF score was 8.2 (SD 4.0) at baseline. The mean ICIQ-UI SF score reduction at follow-up was 1.31 (95% CI: 1.19 to 1.44) with a larger reduction in those with more severe incontinence at baseline (severe/very severe 3.23 (95% CI: 2.85 to 3.61), moderate 1.41 (95% CI: 1.24 to 1.59) and slight 0.24 (95% CI 0.06 to 0.42). When the results were weighted to match the distribution of severity in the RCT, the ICIQ-UI SF score reduction was 2.2 compared with 3.9 in the RCT. Regarding PGI-I, 65% experienced improvement compared with 92% in the RCT.ConclusionsThe app Tät was effective for self-management of UI even in the real world. Although the reduction in incontinence symptoms was less than in the RCT, two-thirds of the users improved. App-based treatment reaches many women without requiring resources from ordinary healthcare services.

A Densely Connected Network Based on U-Net for Medical Image Segmentation

2021 ◽  
Vol 17 (3) ◽  
pp. 1-14
Author(s):  
Zhenzhen Yang ◽  
Pengfei Xu ◽  
Yongpeng Yang ◽  
Bing-Kun Bao
Keyword(s):  
Feature Extraction ◽  
Image Segmentation ◽  
Loss Function ◽  
Network Architecture ◽  
Medical Image ◽  
Medical Image Segmentation ◽  
Cross Entropy ◽  
Loss Functions ◽  
Feature Maps ◽  
Different Levels

The U-Net has become the most popular structure in medical image segmentation in recent years. Although its performance for medical image segmentation is outstanding, a large number of experiments demonstrate that the classical U-Net network architecture seems to be insufficient when the size of segmentation targets changes and the imbalance happens between target and background in different forms of segmentation. To improve the U-Net network architecture, we develop a new architecture named densely connected U-Net (DenseUNet) network in this article. The proposed DenseUNet network adopts a dense block to improve the feature extraction capability and employs a multi-feature fuse block fusing feature maps of different levels to increase the accuracy of feature extraction. In addition, in view of the advantages of the cross entropy and the dice loss functions, a new loss function for the DenseUNet network is proposed to deal with the imbalance between target and background. Finally, we test the proposed DenseUNet network and compared it with the multi-resolutional U-Net (MultiResUNet) and the classic U-Net networks on three different datasets. The experimental results show that the DenseUNet network has significantly performances compared with the MultiResUNet and the classic U-Net networks.

Baseline characteristics and clinical features of patients with heart failure with reduced ejection fraction: a European real-world, non-interventional registry study

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
U Zeymer ◽  
L.H Lund ◽  
V Barrios ◽  
C Fonseca ◽  
A.L Clark ◽  
...  
Keyword(s):  
Heart Failure ◽  
Ejection Fraction ◽  
Real World ◽  
Nyha Class ◽  
Funding Source ◽  
Class Iii ◽  
Reduced Ejection Fraction ◽  
Cardioverter Defibrillator ◽  
Real World Setting ◽  
The Mean

Abstract Background Heart failure (HF) is a major medical and economic burden that is often managed in office based practices. Recently, the angiotensin receptor neprilysin inhibitor (ARNI) sacubitril/valsartan (S/V) was introduced as novel therapeutic option into European guidelines for the management of HF. The ARIADNE registry aims to provide information on how outpatients with HF with reduced ejection fraction (HFrEF) are managed in Europe, in light of this novel treatment option. Methods ARIADNE was a prospective registry of patients with HFrEF treated by office-based cardiologists (OBC) or selected primary care physicians (recognized as HF specialists; PCP) in a real world setting. HFrEF patients were included prospectively, independently of whether treatment had been changed recently or not. 9069 patients were recruited from 687 centres in 17 European countries. Results The mean age of all patients was 68.1 years (S/V: 67.3 years, Non-S/V: 68.9 years). The majority of patients were in NYHA class II (61.3%), or NYHA class III (37.1%) overall, while more patients in the S/V group showed NYHA class III (S/V: 42.8%, Non-S/V: 30.9%). Mean LVEF was slightly lower in the S/V group than in the Non-S/V group (S/V: 32.7%, Non-S/V: 35.4%, overall 34.0%). The most frequently observed signs of HF were dyspnoea upon effort, followed by fatigue, palpitations on exertion at baseline. More patients tend to have more severe symptoms in the S/V groups (e.g. for dyspnoea on effort, Non-S/V: moderate 40.8%, severe 8.6%; S/V: moderate 46.4%, severe 14.1%). 44.0% of patients from the S/V group and 39.3% of non-S/V patients reported at least one hospitalization within 12 months prior to baseline, of which 73.3% in S/V and 69.9% in non-S/V patients were due to HF., At baseline, 44.7% of the patients used a CV device, of which most were implantable cardioverter defibrillator (ICD: Non-S/V 54.2%, S/V: 52.8%), implantable cardioverter defibrillator (CRT-ICD:Non-S/V 21.9%, S/V: 27.0%), and pacemaker (Non-S/V: 13.4%, S/V: 10.5%). The mean KCCQ overall summary score was 62.6 in the S/V group and 69.5 in the Non-S/V group at baseline. 83.9% of patients were treated with ARB or ACEi in Non-S/V group, (ACEi 57.3%, ARB 26.9%). The most frequently taken drug combinations in either group were ACEi/ ARB or S/V with β -blockers (Non-S/V 69.3%, S/V 67.3%). 40.2% in the Non-S/V group and 42.9% in S/V groups used a combination of ACEi/ARB or S/V, β-blocker and MRA. Conclusions The ARIADNE prospective registry provides insights and reflects variations in HF treatment practices in outpatients in Europe and the way S/V was introduced by OBCs and specialized PCPs in a real-world setting. In the observed population, S/V is more often prescribed to slightly younger patients with slightly lower LVEF, there was a greater observed percentage of S/V patients NYHA class III, with lower quality of life measurements and with more severe symptoms and recent hospitalizations for heart failure. Funding Acknowledgement Type of funding source: Private company. Main funding source(s): Novartis Pharma AG

A Modified CGM Algorithm Enhances Data Availability While Retaining iCGM Performance

2021 ◽  
pp. 193229682110075
Author(s):  
Rebecca A. Harvey Towers ◽  
Xiaohe Zhang ◽  
Rasoul Yousefi ◽  
Ghazaleh Esmaili ◽  
Liang Wang ◽  
...  
Keyword(s):  
Data Integration ◽  
Patient Experience ◽  
Real World ◽  
Data Availability ◽  
Post Processing ◽  
Sensor Error ◽  
The Mean ◽  
The Difference ◽  
Data Gap ◽  
Improve Patient

The algorithm for the Dexcom G6 CGM System was enhanced to retain accuracy while reducing the frequency and duration of sensor error. The new algorithm was evaluated by post-processing raw signals collected from G6 pivotal trials (NCT02880267) and by assessing the difference in data availability after a limited, real-world launch. Accuracy was comparable with the new algorithm—the overall %20/20 was 91.7% before and 91.8% after the algorithm modification; MARD was unchanged. The mean data gap due to sensor error nearly halved and total time spent in sensor error decreased by 59%. A limited field launch showed similar results, with a 43% decrease in total time spent in sensor error. Increased data availability may improve patient experience and CGM data integration into insulin delivery systems.

scholarly journals Real-World Clinical Outcomes Associated with Canagliflozin in Patients with Type 2 Diabetes Mellitus in Spain: The Real-Wecan Study

2020 ◽  
Vol 9 (7) ◽  
pp. 2275
Author(s):  
Juan J. Gorgojo-Martínez ◽  
Manuel A. Gargallo-Fernández ◽  
Alba Galdón Sanz-Pastor ◽  
Teresa Antón-Bravo ◽  
Miguel Brito-Sanfiel ◽  
...  
Keyword(s):  
Diabetes Mellitus ◽  
Real World ◽  
Primary Outcome ◽  
Baseline Hba1c ◽  
Real World Setting ◽  
Antihyperglycemic Therapy ◽  
The Mean ◽  
Hba1c Levels

The aims of this multicentric retrospective study were to assess in a real-world setting the effectiveness and safety of canagliflozin 100 mg/d (CANA100) as an add-on to the background antihyperglycemic therapy, and to evaluate the intensification of prior sodium–glucose co-transporter type 2 inhibitor (SGLT-2i) therapy by switching to canagliflozin 300 mg/d (CANA300) in patients with T2DM. One cohort of SGLT2i-naïve patients with T2DM who were initiated on CANA100 and a second cohort of patients with prior background SGLT-2i therapy who switched to CANA300 were included in the study. The primary outcome of the study was the mean change in HbA1c over the follow-up time. In total, 583 patients were included—279 in the cohort of CANA100 (HbA1c 8.05%, weight 94.9 kg) and 304 in the cohort of CANA300 (HbA1c 7.51%, weight 92.0 kg). Median follow-up periods in both cohorts were 9.1 and 15.4 months respectively. CANA100 was associated to significant reductions in HbA1c (−0.90%) and weight (−4.1 kg) at the end of the follow-up. In those patients with baseline HbA1c > 8% (mean 9.25%), CANA100 lowered HbA1c levels by 1.51%. In the second cohort, patients switching to CANA300 experienced a significant decrease in HbA1c (−0.35%) and weight (−2.1 kg). In those patients with baseline HbA1c > 8% (mean 8.94%), CANA300 lowered HbA1c levels by 1.12%. There were significant improvements in blood pressure in both cohorts. No unexpected adverse events were reported. In summary, CANA100 (as an add-on therapy) and CANA300 (switching from prior SGLT-2i therapy) significantly improved several cardiometabolic parameters in patients with T2DM.

A Hard C-Means Clustering Algorithm Incorporating Membership KL Divergence and Local Data Information for Noisy Image Segmentation

2017 ◽  
Vol 32 (04) ◽  
pp. 1850012 ◽  
Author(s):  
R. R. Gharieb ◽  
G. Gendy ◽  
H. Selim
Keyword(s):  
Image Segmentation ◽  
Membership Function ◽  
Clustering Algorithm ◽  
Clustering Algorithms ◽  
Cluster Center ◽  
Local Data ◽  
Cluster Membership ◽  
Kl Divergence ◽  
Clustering Approach ◽  
Center Distance

In this paper, the standard hard C-means (HCM) clustering approach to image segmentation is modified by incorporating weighted membership Kullback–Leibler (KL) divergence and local data information into the HCM objective function. The membership KL divergence, used for fuzzification, measures the proximity between each cluster membership function of a pixel and the locally-smoothed value of the membership in the pixel vicinity. The fuzzification weight is a function of the pixel to cluster-centers distances. The used pixel to a cluster-center distance is composed of the original pixel data distance plus a fraction of the distance generated from the locally-smoothed pixel data. It is shown that the obtained membership function of a pixel is proportional to the locally-smoothed membership function of this pixel multiplied by an exponentially distributed function of the minus pixel distance relative to the minimum distance provided by the nearest cluster-center to the pixel. Therefore, since incorporating the locally-smoothed membership and data information in addition to the relative distance, which is more tolerant to additive noise than the absolute distance, the proposed algorithm has a threefold noise-handling process. The presented algorithm, named local data and membership KL divergence based fuzzy C-means (LDMKLFCM), is tested by synthetic and real-world noisy images and its results are compared with those of several FCM-based clustering algorithms.

Real Time Interactive Image Segmentation Using User Indicated Real-World Seeds

2012 ◽  
Author(s):  
Rafael Beserra Gomes ◽  
Rafael Vidal Aroca ◽  
Bruno Motta de Carvalho ◽  
Luiz Marcos Garcia Goncalves
Keyword(s):  
Image Segmentation ◽  
Real Time ◽  
Real World ◽  
Interactive Image Segmentation

scholarly journals Recognition by Variance: Learning Rules for Spatiotemporal Patterns

2006 ◽  
Vol 18 (10) ◽  
pp. 2343-2358 ◽  
Author(s):  
Omri Barak ◽  
Misha Tsodyks
Keyword(s):  
Auditory Processing ◽  
Emergent Property ◽  
Spatiotemporal Patterns ◽  
Input Current ◽  
Time Warping ◽  
Time Constants ◽  
Integrate And Fire ◽  
Learning Rules ◽  
The Mean ◽  
Monotonic Relation

Recognizing specific spatiotemporal patterns of activity, which take place at timescales much larger than the synaptic transmission and membrane time constants, is a demand from the nervous system exemplified, for instance, by auditory processing. We consider the total synaptic input that a single readout neuron receives on presentation of spatiotemporal spiking input patterns. Relying on the monotonic relation between the mean and the variance of a neuron's input current and its spiking output, we derive learning rules that increase the variance of the input current evoked by learned patterns relative to that obtained from random background patterns. We demonstrate that the model can successfully recognize a large number of patterns and exhibits a slow deterioration in performance with increasing number of learned patterns. In addition, robustness to time warping of the input patterns is revealed to be an emergent property of the model. Using a leaky integrate-and-fire realization of the readout neuron, we demonstrate that the above results also apply when considering spiking output.

Abstract 1122‐000142: Safety and Efficacy of Surpass Streamline for Intracranial Aneurysms: A Multicenter US Real‐World Experience (SESSIA)

2021 ◽  
Vol 1 (S1) ◽  
Author(s):  
Juan Vivanco‐Suarez ◽  
Alan Mendez‐Ruiz ◽  
Farooqui Mudassir ◽  
Cynthia B Zevallos ◽  
Milagros Galecio‐Castillo ◽  
...  
Keyword(s):  
Real World ◽  
Intracranial Aneurysms ◽  
The United States ◽  
Chromium Alloy ◽  
Cobalt Chromium ◽  
Anterior Circulation ◽  
Safety And Efficacy ◽  
Post Marketing ◽  
The Mean

Introduction : Flow diversion has established itself as standard treatment of wide complex intracranial aneurysms (IA). Its recognition has been validated with positive occlusion rates and favorable clinical outcomes. The Surpass Streamline (SS) flow diverter (FD) is a braided cobalt/chromium alloy implant with 72 or 96 wires approved by the FDA in 2018. The aim of this study is to determine the safety and efficacy of the SS in a post‐marketing large US cohort. Methods : We performed a multicenter, retrospective study for consecutive patients treated with the SS FD for IA between January 2018 and June 2021 in the United States. Inclusion criteria for participants were: 1. Adults (≥ 18 years) and 2. Treatment with SS FD for IA. Primary safety end point was a major ipsilateral stroke (increase in National Institutes of Health Stroke Scale Score of ≥ 4) or neurological death within 12 months. Primary efficacy was assessed using the 3‐point Raymond‐Roy (RR) occlusion scale on digital subtraction angiography (DSA) at 6‐12‐month follow‐up. Results : A total of 276 patients with 313 aneurysms were enrolled. The median age was 59 years and 199 (72%) were females. The most common comorbidities included hypertension in 156 (57%) subjects followed by hyperlipidemia in 76 (28%) patients. One hundred and twenty‐two (44%) patients were asymptomatic while subarachnoid hemorrhage was present in only 10 (4%) patients. A total of 143 (46%) aneurysms were left‐sided. Aneurysms were located as follows: 274 (88%) were in the anterior circulation with paraophthalmic being the most common in 120 (38%) followed by petrocavernous ICA in 81 (26%); 33 (11%) aneurysms were located in the posterior circulation with basilar trunk being the most common in 14 (5%). The mean maximum aneurysm dome width was 5.77 ± 4.7 mm, neck width 4.22 ± 3.8 mm and dome to neck ratio was 1.63 ± 1.3 mm. The mean number of SS FD implanted per aneurysm was 1.06 (range 1–3) with more than one SS FD implanted in 21 (7%) aneurysms. Modified Rankin Scale (mRS) of 0–2 was present in 206/213 (97%) patients at 6–12 month follow‐up. The complete aneurysm occlusion (RR 1) rate was 145/175 (83%) among subjects who had angiographic follow‐up at 6–12 months. Major stroke and death was encountered in 7 (2%) and 5 (1.8%) of the patients respectively. Conclusions : Our data represent the largest real‐world study using SS FD. These results corroborate its post‐marketing safety and efficacy for the treatment of intracranial aneurysms showing more favorable rates to the initial experience during SCENT trial.

Sign in / Sign up

Export Citation Format

Share Document