scholarly journals Artificial intelligence classification model for macular degeneration images: a robust optimization framework for residual neural networks

2021 ◽  
Vol 22 (S5) ◽  
Author(s):  
Wen-Hsien Ho ◽  
Tian-Hsiang Huang ◽  
Po-Yuan Yang ◽  
Jyh-Horng Chou ◽  
Hong-Siang Huang ◽  
...  

Abstract Background The prevalence of chronic disease is growing in aging societies, and artificial-intelligence–assisted interpretation of macular degeneration images is a topic that merits research. This study proposes a residual neural network (ResNet) model constructed using uniform design. The ResNet model is an artificial intelligence model that classifies macular degeneration images and can assist medical professionals in related tests and classification tasks, enhance confidence in making diagnoses, and reassure patients. However, the various hyperparameters in a ResNet lead to the problem of hyperparameter optimization in the model. This study employed uniform design—a systematic, scientific experimental design—to optimize the hyperparameters of the ResNet and establish a ResNet with optimal robustness. Results An open dataset of macular degeneration images (https://data.mendeley.com/datasets/rscbjbr9sj/3) was divided into training, validation, and test datasets. According to accuracy, false negative rate, and signal-to-noise ratio, this study used uniform design to determine the optimal combination of ResNet hyperparameters. The ResNet model was tested and the results compared with results obtained in a previous study using the same dataset. The ResNet model achieved higher optimal accuracy (0.9907), higher mean accuracy (0.9848), and a lower mean false negative rate (0.015) than did the model previously reported. The optimal ResNet hyperparameter combination identified using the uniform design method exhibited excellent performance. Conclusion The high stability of the ResNet model established using uniform design is attributable to the study’s strict focus on achieving both high accuracy and low standard deviation. This study optimized the hyperparameters of the ResNet model by using uniform design because the design features uniform distribution of experimental points and facilitates effective determination of the representative parameter combination, reducing the time required for parameter design and fulfilling the requirements of a systematic parameter design process.

2011 ◽  
Vol 21 (9) ◽  
pp. 1679-1683 ◽  
Author(s):  
Tessa A. Ennik ◽  
David G. Allen ◽  
Ruud L.M. Bekkers ◽  
Simon E. Hyde ◽  
Peter T. Grant

BackgroundThere is a growing interest to apply the sentinel node (SN) procedure in the treatment of vulvar cancer. Previous vulvar surgery might disrupt lymphatic patterns and thereby decrease SN detection rates, lengthen scintigraphic appearance time (SAT), and increase SN false-negative rate. The aims of this study were to evaluate the SN detection rates at the Mercy Hospital for Women in Melbourne and to investigate whether previous vulvar surgery affects SN detection rates, SAT, and SN false-negative rate.MethodsData on all patients with vulvar cancer who underwent an SN procedure (blue dye, technetium, or combined technique) from November 2000 to July 2010 were retrospectively collected.ResultsSixty-five SN procedures were performed. Overall detection rate was 94% per person and 80% per groin. Detection rates in the group of patients who underwent previous excision of the primary tumor were not lower compared with the group without previous surgery or with just an incisional biopsy. There was no statistical significant difference in SAT between the previous excision group and the other patients. None of the patients with a false-negative SN had undergone previous excision.ConclusionsResults indicate that previous excision of a primary vulvar malignancy does not decrease SN detection rates or increase SN false-negative rate. Therefore, the SN procedure appears to be a reliable technique in patients who have previously undergone vulvar surgery. Previous excision did not significantly lengthen SAT, but the sample size in this subgroup analysis was small.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Katherine F. Jarvis ◽  
Joshua B. Kelley

AbstractColleges and other organizations are considering testing plans to return to operation as the COVID-19 pandemic continues. Pre-symptomatic spread and high false negative rates for testing may make it difficult to stop viral spread. Here, we develop a stochastic agent-based model of COVID-19 in a university sized population, considering the dynamics of both viral load and false negative rate of tests on the ability of testing to combat viral spread. Reported dynamics of SARS-CoV-2 can lead to an apparent false negative rate from ~ 17 to ~ 48%. Nonuniform distributions of viral load and false negative rate lead to higher requirements for frequency and fraction of population tested in order to bring the apparent Reproduction number (Rt) below 1. Thus, it is important to consider non-uniform dynamics of viral spread and false negative rate in order to model effective testing plans.


2021 ◽  
Vol 106 ◽  
pp. 106582
Author(s):  
Alex Niu ◽  
Bo Ning ◽  
Francisco Socola ◽  
Hana Safah ◽  
Tim Reynolds ◽  
...  

2021 ◽  
Vol 108 (Supplement_2) ◽  
Author(s):  
E Johnston ◽  
S Taylor ◽  
F Bannon ◽  
S McAllister

Abstract Introduction and Aims The aim of this systematic review is to provide an up-to-date evaluation of the role and test performance of sentinel lymph node biopsy (SLNB) in the head and neck. Method This review follows the PRISMA guidelines. Database searches for MEDLINE and EMBASE were constructed to retrieve human studies published between 1st January 2010 and 1st July 2020 assessing the role and accuracy of sentinel lymph node biopsy in cutaneous malignant melanoma of the head and neck. Articles were independently screened by two reviewers and critically appraised using the MINORS criteria. The primary outcomes consisted of the sentinel node identification rate and test-performance measures, including the false-negative rate and the posttest probability negative. Results A total of 27 studies, including 4688 patients, met the eligibility criteria. Statistical analysis produced weighted summary estimates for the sentinel node identification rate of 97.3% (95% CI, 95.9% to 98.6%), the false-negative rate of 21.3% (95% CI, 17.0% to 25.4%) and the posttest probability negative of 4.8% (95% CI, 3.9% to 5.8%). Discussion Sentinel lymph node biopsy is accurate and feasible in the head and neck. Despite technical improvements in localisation techniques, the false negative rate remains disproportionately higher than for melanoma in other anatomical sites.


2021 ◽  
Vol 10 (4) ◽  
pp. 602
Author(s):  
Antoine Tardieu ◽  
Lobna Ouldamer ◽  
François Margueritte ◽  
Lauranne Rossard ◽  
Aymeline Lacorre ◽  
...  

The objective of our study is to evaluate the diagnostic performance of positron emission tomography/computed tomography (PET-CT) for the assessment of lymph node involvement in advanced epithelial ovarian, fallopian tubal or peritoneal cancer (EOC). This was a retrospective, bicentric study. We included all patients over 18 years of age with a histological diagnosis of advanced EOC who had undergone PET-CT at the time of diagnosis or prior to cytoreduction surgery with pelvic or para-aortic lymphadenectomy. We included 145 patients with primary advanced EOC. The performance of PET-CT was calculated from the data of 63 patients. The sensitivity of PET-CT for preoperative lymph node evaluation was 26.7%, specificity was 90.9%, PPV was 72.7%, and NPV was 57.7%. The accuracy rate was 60.3%, and the false-negative rate was 34.9%. In the case of primary cytoreduction (n = 16), the sensitivity of PET-CT was 50%, specificity was 87.5%, PPV was 80%, and NPV was 63.6%. The accuracy rate was 68.8%, and the false negative rate was 25%. After neoadjuvant chemotherapy (n = 47), the sensitivity of PET-CT was 18.2%, specificity was 92%, PPV was 66.7%, and NPV was 56.1%. The accuracy rate was 57.5%, and the false negative rate was 38.3%. Due to its high specificity, the performance of a preoperative PET-CT scan could contribute to the de-escalation and reduction of lymphadenectomy in the surgical management of advanced EOC in a significant number of patients free of lymph node metastases.


2019 ◽  
Vol 58 (6) ◽  
pp. 671-676
Author(s):  
Amy M. West ◽  
Pierre A. d’Hemecourt ◽  
Olivia J. Bono ◽  
Lyle J. Micheli ◽  
Dai Sugimoto

The objective of this study was to determine diagnostic accuracy of magnetic resonance imaging (MRI) and computed tomography (CT) scans in young athletes diagnosed with spondylolysis. A cross-sectional study was used. Twenty-two young athletes (14.7 ± 1.5 years) were diagnosed as spondylolysis based on a single-photon emission CT. Following the diagnosis, participants underwent MRI and CT scan imaging tests on the same day. The sensitivity and false-negative rate of the MRI and CT scans were analyzed. MRI test confirmed 13 (+) and 9 (−) results while CT test showed 17 (+) and 5 (−) results. The sensitivity and false-negative rate of MRI were, respectively, 59.1% (95% confidence interval [CI] = 36.7% to 78.5%) and 40.9% (95% CI = 21.5% to 63.3%). Furthermore, the sensitivity and false-negative rate of CT scan were 77.3% (95% CI = 54.2% to 91.3%) and 22.7% (95% CI = 0.09% to 45.8%). Our results indicated that CT scan is a more accurate imaging modality to diagnose spondylolysis compared with MRI in young athletes.


2006 ◽  
Vol 82 (4) ◽  
pp. 1185-1190 ◽  
Author(s):  
Anthony Lemaire ◽  
Ivana Nikolic ◽  
Thomas Petersen ◽  
Jack C. Haney ◽  
Eric M. Toloza ◽  
...  

2012 ◽  
Vol 127 (3) ◽  
pp. 462-466 ◽  
Author(s):  
David Cibula ◽  
Nadeem R. Abu-Rustum ◽  
Ladislav Dusek ◽  
Jiri Slama ◽  
Michal Zikán ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document