scholarly journals Interocular Symmetry Analysis of Corneal Elevation Using the Fellow Eye as the Reference Surface and Machine Learning

Healthcare ◽  
2021 ◽  
Vol 9 (12) ◽  
pp. 1738
Author(s):  
Shiva Mehravaran ◽  
Iman Dehzangi ◽  
Md Mahmudur Rahman
Keyword(s):  
Machine Learning ◽  
Mirror Symmetry ◽  
Single Point ◽  
Population Based ◽  
Reference Ranges ◽  
Reference Surface ◽  
Fellow Eye ◽  
Elevation Data ◽  
Clinically Significant ◽  
Middle Aged Adults

Unilateral corneal indices and topography maps are routinely used in practice, however, although there is consensus that fellow-eye asymmetry can be clinically significant, symmetry studies are limited to local curvature and single-point thickness or elevation measures. To improve our current practices, there is a need to devise algorithms for generating symmetry colormaps, study and categorize their patterns, and develop reference ranges for new global discriminative indices for identifying abnormal corneas. In this work, we test the feasibility of using the fellow eye as the reference surface for studying elevation symmetry throughout the entire corneal surface using 9230 raw Pentacam files from a population-based cohort of 4613 middle-aged adults. The 140 × 140 matrix of anterior elevation data in these files were handled with Python to subtract matrices, create color-coded maps, and engineer features for machine learning. The most common pattern was a monochrome circle (“flat”) denoting excellent mirror symmetry. Other discernible patterns were named “tilt”, “cone”, and “four-leaf”. Clustering was done with different combinations of features and various algorithms using Waikato Environment for Knowledge Analysis (WEKA). Our proposed approach can identify cases that may appear normal in each eye individually but need further testing. This work will be enhanced by including data of posterior elevation, thickness, and common diagnostic indices.

scholarly journals A Comparison of the Cyclic Variation in Serum Levels of CA125 Across the Menstrual Cycle Using Two Commercial Assays

2011 ◽  
Vol 14 (3) ◽  
pp. 250-256 ◽  
Author(s):  
Monica R. McLemore ◽  
Bradley E. Aouizerat ◽  
Kathryn A. Lee ◽  
Lee-may Chen ◽  
Bruce Cooper ◽  
...  
Keyword(s):  
Menstrual Cycle ◽  
Premenopausal Women ◽  
Serum Levels ◽  
Population Based ◽  
Reference Ranges ◽  
Healthy Women ◽  
Serum Ca125 ◽  
Caucasian Women ◽  
Clinically Significant ◽  
Good Concordance

Background: Clinicians use CA125, a tumor-associated antigen, primarily to monitor epithelial ovarian cancer. However, CA125 lacks the sensitivity and specificity necessary for population-based screening in healthy women. The purpose of this study was to determine if serum concentrations of CA125 differed across the three phases of the menstrual cycle in healthy, premenopausal women using two commercially available assays. Methods: Healthy, Caucasian women between the ages of 18 and 39 were enrolled using strict criteria to exclude factors known to contribute to CA125 fluctuations. Menstrual cycle regularity was determined using calendars maintained by participants for 3 months. After cycle regularity was established, blood was drawn at three time points for CA125 determination using two commercial assays (i.e., Siemens and Panomics). Results: Regardless of the assay used, CA125 values were highest during menses. The CA125 values decreased 0.2 U/ml per day from menses to the end of the same cycle, which resulted in a net decrease of 5.8 U/ml across the cycle. Conclusions: The two commercial assays for CA125 determination demonstrated good concordance in terms of reference ranges regardless of epitope differences. While CA125 levels changed over the course of the menstrual cycle, these changes may not be clinically significant in healthy women. This study is the first to control for factors known to contribute to CA125 elevations; to quantify a decrease in CA125 levels across the menstrual cycle; and to confirm concordance in the relative decreases in serum CA125 levels across the menstrual cycle between two frequently used commercial assays.

Early Detection of Bipolar Disorder Four Years Before Onset in a 22-Year Population-Based Birth Cohort Using Advanced Machine Learning Techniques

2020 ◽  
Author(s):  
Francisco Diego Rabelo-da-Ponte ◽  
Jacson Gabriel Feiten ◽  
Benson Mwangi ◽  
Fernando C. Barros ◽  
Fernando C. Wehrmeister ◽  
...  
Keyword(s):  
Machine Learning ◽  
Bipolar Disorder ◽  
Early Detection ◽  
Birth Cohort ◽  
Population Based ◽  
Machine Learning Techniques ◽  
Learning Techniques

scholarly journals Factors associated with sarcopenia screened by finger-circle test among middle-aged and older adults: a population-based multisite cross-sectional survey in Japan

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Daiki Watanabe ◽  
Tsukasa Yoshida ◽  
Takashi Nakagata ◽  
Naomi Sawada ◽  
Yosuke Yamada ◽  
...  
Keyword(s):  
Older Adults ◽  
Smoking Status ◽  
Multiple Logistic Regression Analysis ◽  
Population Based ◽  
Middle Aged ◽  
Cross Sectional ◽  
Factors Associated ◽  
Mail Surveys ◽  
Osaka Prefecture ◽  
Middle Aged Adults

AbstractBackgroundPrevious epidemiological studies have demonstrated the prevalence and relationship of various factors associated with sarcopenia in older adults; however, few have examined the status of sarcopenia in middle-aged adults. In this study, we aimed to, 1) evaluate the validity of the finger-circle test, which is potentially a useful screening tool for sarcopenia, and 2) determine the prevalence and factors associated with sarcopenia in middle-aged and older adults.MethodsWe conducted face-to-face surveys of 525 adults, who were aged 40–91 years and resided in Settsu City, Osaka Prefecture, Japan to evaluate the validity of finger-circle test. The finger-circle test evaluated calf circumference by referring to an illustration printed on the survey form. The area under the receiver operating characteristic curves (AUROC) was plotted to evaluate the validity of the finger-circle test for screening sarcopenia and compared to that evaluated by skeletal muscle mass index (SMI) measured using bioimpedance. We also conducted multisite population-based cross-sectional anonymous mail surveys of 9337 adults, who were aged 40–97 years and resided in Settsu and Hannan Cities, Osaka Prefecture, Japan. Participants were selected through stratified random sampling by sex and age in the elementary school zones of their respective cities. We performed multiple logistic regression analysis to explore associations between characteristics and prevalence of sarcopenia.ResultsSarcopenia, defined by SMI, was moderately predicted by a finger-circle test response showing that the subject’s calf was smaller than their finger-circle (AUROC: 0.729, < 65 years; 0.653, ≥65 years); such subjects were considered to have sarcopenia. In mail surveys, prevalence of sarcopenia screened by finger-circle test was higher in older subjects (approximately 16%) than in middle-aged subjects (approximately 8–9%). In a multiple regression model, the factors associated with sarcopenia were age, body mass index, smoking status, self-reported health, and number of meals in all the participants.ConclusionsSarcopenia, screened by the finger-circle test, was present not only among older adults but also among middle-aged adults. These results may provide useful indications for developing public health programs, not only for the prevention, but especially for the management of sarcopenia.Trial registrationUMIN000036880, registered prospectively May 29, 2019, https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000042027

scholarly journals Distributed Learning Applications in Power Systems: A Review of Methods, Gaps, and Challenges

Energies ◽  
2021 ◽  
Vol 14 (12) ◽  
pp. 3654
Author(s):  
Nastaran Gholizadeh ◽  
Petr Musilek
Keyword(s):  
Machine Learning ◽  
Power Systems ◽  
Learning Algorithm ◽  
Single Point ◽  
Distributed Learning ◽  
Large Data ◽  
Multi Agent Systems ◽  
Power Quality Monitoring ◽  
Multi Agent ◽  
Learning Frameworks

In recent years, machine learning methods have found numerous applications in power systems for load forecasting, voltage control, power quality monitoring, anomaly detection, etc. Distributed learning is a subfield of machine learning and a descendant of the multi-agent systems field. Distributed learning is a collaboratively decentralized machine learning algorithm designed to handle large data sizes, solve complex learning problems, and increase privacy. Moreover, it can reduce the risk of a single point of failure compared to fully centralized approaches and lower the bandwidth and central storage requirements. This paper introduces three existing distributed learning frameworks and reviews the applications that have been proposed for them in power systems so far. It summarizes the methods, benefits, and challenges of distributed learning frameworks in power systems and identifies the gaps in the literature for future studies.

scholarly journals Applying Machine Learning to Predict the Exportome of Bovine and Canine Babesia Species That Cause Babesiosis

Pathogens ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 660
Author(s):  
Stephen J. Goodswen ◽  
Paul J. Kennedy ◽  
John T. Ellis
Keyword(s):  
Machine Learning ◽  
Severe Disease ◽  
Economic Loss ◽  
Babesia Bovis ◽  
Training Data ◽  
Babesia Bigemina ◽  
Bovine Babesiosis ◽  
Canine Babesiosis ◽  
Clinically Significant ◽  
Significant Disease

Babesia infection of red blood cells can cause a severe disease called babesiosis in susceptible hosts. Bovine babesiosis causes global economic loss to the beef and dairy cattle industries, and canine babesiosis is considered a clinically significant disease. Potential therapeutic targets against bovine and canine babesiosis include members of the exportome, i.e., those proteins exported from the parasite into the host red blood cell. We developed three machine learning-derived methods (two novel and one adapted) to predict for every known Babesia bovis, Babesia bigemina, and Babesia canis protein the probability of being an exportome member. Two well-studied apicomplexan-related species, Plasmodium falciparum and Toxoplasma gondii, with extensive experimental evidence on their exportome or excreted/secreted proteins were used as important benchmarks for the three methods. Based on 10-fold cross validation and multiple train–validation–test splits of training data, we expect that over 90% of the predicted probabilities accurately provide a secretory or non-secretory indicator. Only laboratory testing can verify that predicted high exportome membership probabilities are creditable exportome indicators. However, the presented methods at least provide those proteins most worthy of laboratory validation and will ultimately save time and money.

Development of a machine learning algorithm to accurately detect clinically significant prostate cancer on MRI

2019 ◽  
Vol 18 (11) ◽  
pp. e3493
Author(s):  
J.W.M. Dillon ◽  
T.I. Whish-Wilson ◽  
S.J. Riddell ◽  
L-M. Wong ◽  
P. Brotchie ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Machine Learning ◽  
Learning Algorithm ◽  
Machine Learning Algorithm ◽  
Clinically Significant

scholarly journals Assessing Conformer Energies using Electronic Structure and Machine Learning Methods

2020 ◽  
Author(s):  
Dakota Folmsbee ◽  
Geoffrey Hutchison
Keyword(s):  
Machine Learning ◽  
Electronic Structure ◽  
Density Functional ◽  
Large Scale ◽  
Single Point ◽  
Semiempirical Method ◽  
Coupled Cluster ◽  
Scale Evaluation ◽  
Machine Learning Methods ◽  
Electronic Structure Methods

We have performed a large-scale evaluation of current computational methods, including conventional small-molecule force fields, semiempirical, density functional, ab initio electronic structure methods, and current machine learning (ML) techniques to evaluate relative single-point energies. Using up to 10 local minima geometries across ~700 molecules, each optimized by B3LYP-D3BJ with single-point DLPNO-CCSD(T) triple-zeta energies, we consider over 6,500 single points to compare the correlation between different methods for both relative energies and ordered rankings of minima. We find promise from current ML methods and recommend methods at each tier of the accuracy-time tradeoff, particularly the recent GFN2 semiempirical method, the B97-3c density functional approximation, and RI-MP2 for accurate conformer energies. The ANI family of ML methods shows promise, particularly the ANI-1ccx variant trained in part on coupled-cluster energies. Multiple methods suggest continued improvements should be expected in both performance and accuracy.

scholarly journals Assessing Conformer Energies using Electronic Structure and Machine Learning Methods

2020 ◽  
Author(s):  
Dakota Folmsbee ◽  
Geoffrey Hutchison
Keyword(s):  
Machine Learning ◽  
Electronic Structure ◽  
Density Functional ◽  
Large Scale ◽  
Single Point ◽  
Semiempirical Method ◽  
Coupled Cluster ◽  
Scale Evaluation ◽  
Machine Learning Methods ◽  
Electronic Structure Methods

We have performed a large-scale evaluation of current computational methods, including conventional small-molecule force fields, semiempirical, density functional, ab initio electronic structure methods, and current machine learning (ML) techniques to evaluate relative single-point energies. Using up to 10 local minima geometries across ~700 molecules, each optimized by B3LYP-D3BJ with single-point DLPNO-CCSD(T) triple-zeta energies, we consider over 6,500 single points to compare the correlation between different methods for both relative energies and ordered rankings of minima. We find promise from current ML methods and recommend methods at each tier of the accuracy-time tradeoff, particularly the recent GFN2 semiempirical method, the B97-3c density functional approximation, and RI-MP2 for accurate conformer energies. The ANI family of ML methods shows promise, particularly the ANI-1ccx variant trained in part on coupled-cluster energies. Multiple methods suggest continued improvements should be expected in both performance and accuracy.

Sign in / Sign up

Export Citation Format

Share Document