Peripapillary retinal artery in first diagnosed and untreated normal tension glaucoma.

BACKGROUND: Glaucoma, an important cause of visual impairment in many countries, remains a common eye condition due to difficulties in its early diagnosis. We analyzed the characteristics of retinal arteries to add a valuable technology for helping the normal tension glaucoma (NTG) diagnosis. METHODS: This study included 51 patients with newly diagnosed NTG with hemifield defects and 60 age-matched controls. Peripapillary retinal arteriolar calibers (PRACs) photoed by non-mydriatic retinal camera were measured using ImageJ by two masked readers. We also performed spectral-domain optical coherence tomography to evaluate retinal nerve fiber layer thickness (RNFLT) and optic disc parameters. Their relations to retinal arteriolar calibers were investigated by univariate and multivariate linear regression. The area under the receiver operating characteristic curve (AUROC) was used to confirm the powers to detect NTG by PRACs. RESULTS: PRACs in four quadrants were significantly reduced in individuals with first diagnosed NTG (82 ± 15.1 µm, 80 ± 13.6 µm, 71 ± 11.6 µm, and 64 ± 10.0 µm) compared with those in age-matched controls (101 ± 9.8 µm, 105 ± 8.7 µm, 90 ± 7.5 µm, and 82 ± 9.8 µm). Superotemporal and inferotemporal PRACs in the visual field-affected hemifield were narrower than those in the unaffected hemifield in NTG group (P ≤ 0.004). Temporal PRACs in the RNFL unaffected hemifield were significantly narrower than in healthy eyes (P < 0.001). Superotemporal PRAC showed a significant correlation with superior RNFLT (ß = 0.659, P < 0.001), and a similar relationship was found between inferotemporal PRAC and inferior RNFLT (ß = 0.227, P = 0.015). The diagnostic capability of temporal PRACs was satisfactory (superotemporal PRAC; AUROC 0.983, cut-off value 84.7 µm, inferotemporal PRAC; AUROC 0.946, cut-off value 94.2 µm). CONCLUSIONS: PRAC and inferotemporal PRAC are valid parameters for discriminating patients with NTG.

Effects of respiratory phases on the processing of emotional and non-emotional visual stimuli.

The number of studies investigating the relationship between respiratory phases and cognitive/neural processing of external events has been increasing, but the findings remain controversial. This registered report examined the effect of the respiratory phase on the discrimination accuracy of visual stimuli in the emotional and non-emotional domains. Forty-two healthy young participants were asked to choose fearful over neutral facial expressions and to choose high-contrast over low-contrast Gabor patches during spontaneous nasal respiration. Event-related potentials (ERPs) were also recorded for each type of stimulus presented during each respiratory phase. It was hypothesized that discrimination accuracy would be higher when the stimuli were presented during inhalation than during exhalation. It was also hypothesized that the amplitudes of ERPs elicited by the stimuli would be greater during inhalation than during exhalation. For comparison, the effect of the cardiac phase was examined, with the expectation that discrimination accuracy would be higher when the stimuli were presented during systole than during diastole. It was also hypothesized that the amplitudes of ERPs elicited by the stimuli would be greater during systole than during diastole. As expected, the results indicated that fear discrimination accuracy was higher during inhalation than exhalation and during systole than diastole. However, this was not the case for contrast discrimination. No differences in ERPs were observed between respiratory phases in either task. These results suggest that natural breathing in through the nose facilitates the discrimination of emotional stimuli, possibly via subcortical processes.

Deep Feature Engineering in Colposcopy Image Recognition: A Comparative Study.

Feature fusion techniques have been proposed and tested for many medical applications to improve diagnostic and classification problems. Specifically, cervical cancer classification can be improved by using such techniques. Feature fusion combines information from different datasets into a single dataset. This dataset contains superior discriminant power that can improve classification accuracy. In this paper, we conduct comparisons among six selected feature fusion techniques to provide the best possible classification accuracy of cervical cancer. The considered techniques are canonical correlation analysis, discriminant correlation analysis, least absolute shrinkage and selection operator, independent component analysis, principal component analysis, and concatenation. We generate ten feature datasets that come from the transfer learning of the most popular pre-trained deep learning models: Alex net, Resnet 18, Resnet 50, Resnet 10, Mobilenet, Shufflenet, Xception, Nasnet, Darknet 19, and VGG Net 16. The main contribution of this paper is to combine these models and then apply them to the six feature fusion techniques to discriminate various classes of cervical cancer. The obtained results are then fed into a support vector machine model to classify four cervical cancer classes (i.e., Negative, HISL, LSIL, and SCC). It has been found that the considered six techniques demand relatively comparable computational complexity when they are run on the same machine. However, the canonical correlation analysis has provided the best performance in classification accuracy among the six considered techniques, at 99.7%. The second-best methods were the independent component analysis, least absolute shrinkage and the selection operator, which were found to have a 98.3% accuracy. On the other hand, the worst-performing technique was the principal component analysis technique, which offered 90% accuracy. Our developed approach of analysis can be applied to other medical diagnosis classification problems, which may demand the reduction of feature dimensions as well as a further enhancement of classification performance.

Empirical evaluation of sub-cohort sampling designs for risk prediction modeling.

Sub-cohort sampling designs, such as nested case-control (NCC) and case-cohort (CC) studies, have been widely used to estimate biomarker-disease associations because of their cost effectiveness. These designs have been well studied and shown to maintain relatively high efficiency compared to full-cohort designs, but their performance of building risk prediction models has been less studied. Moreover, sub-cohort sampling designs often use matching (or stratifying) to further control for confounders or to reduce measurement error. Their predictive performance depends on both the design and matching procedures. Based on a dataset from the NYU Women's Health Study (NYUWHS), we performed Monte Carlo simulations to systematically evaluate risk prediction performance under NCC, CC, and full-cohort studies. Our simulations demonstrate that sub-cohort sampling designs can have predictive accuracy (i.e. discrimination and calibration) similar to that of the full-cohort design, but could be sensitive to the matching procedure used. Our results suggest that researchers can have the option of performing NCC and CC studies with huge potential benefits in cost and resources, but need to pay particular attention to the matching procedure when developing a risk prediction model in biomarker studies.

Species distribution models for invasive Eurasian watermilfoil highlight the importance of data quality and limitations of discrimination accuracy metrics.

AIM: Availability of uniformly collected presence, absence, and abundance data remains a key challenge in species distribution modeling (SDM). For invasive species, abundance and impacts are highly variable across landscapes, and quality occurrence and abundance data are critical for predicting locations at high risk for invasion and impacts, respectively. We leverage a large aquatic vegetation dataset comprising point-level survey data that includes information on the invasive plant Myriophyllum spicatum (Eurasian watermilfoil) to: (a) develop SDMs to predict invasion and impact from environmental variables based on presence-absence, presence-only, and abundance data, and (b) compare evaluation metrics based on functional and discrimination accuracy for presence-absence and presence-only SDMs. LOCATION: Minnesota, USA. METHODS: Eurasian watermilfoil presence-absence and abundance information were gathered from 468 surveyed lakes, and 801 unsurveyed lakes were leveraged as pseudoabsences for presence-only models. A Random Forest algorithm was used to model the distribution and abundance of Eurasian watermilfoil as a function of lake-specific predictors, both with and without a spatial autocovariate. Occurrence-based SDMs were evaluated using conventional discrimination accuracy metrics and functional accuracy metrics assessing correlation between predicted suitability and observed abundance. RESULTS: Water temperature degree days and maximum lake depth were two leading predictors influencing both invasion risk and abundance, but they were relatively less important for predicting abundance than other water quality measures. Road density was a strong predictor of Eurasian watermilfoil invasion risk but not abundance. Model evaluations highlighted significant differences: Presence-absence models had high functional accuracy despite low discrimination accuracy, whereas presence-only models showed the opposite pattern. MAIN CONCLUSION: Complementing presence-absence data with abundance information offers a richer understanding of invasive Eurasian watermilfoil's ecological niche and enables evaluation of the model's functional accuracy. Conventional discrimination accuracy measures were misleading when models were developed using pseudoabsences. We thus caution against the overuse of presence-only models and suggest directing more effort toward systematic monitoring programs that yield high-quality data.

Subjective memory complaint as a useful tool for the early detection of Alzheimer's disease.

PURPOSE: Despite their high prevalence in Alzheimer's disease (AD), and the increasing level of concern they have generated, subjective memory complaints (SMCs) are poorly understood. This study investigated the accuracy with which SMC can separate mild cognitive impairment (MCI) and early AD from cognitive normal (CN), and explored whether the discrimination ability is similar to or better than that of the Mini-Mental State Exam (MMSE). PATIENTS AND METHODS: This study recruited 175 CN subjects, 52 with MCI, and 66 with probable AD aged 60 years or older. To test the independent contributions of SMC and MMSE scores to the classification of cognitive status (CN vs MCI or early AD), logistic regression analyses were performed, adjusting for the following potential confounding variables: age, gender, Frontal Assessment Battery score, modified Hachinski Ischemic Scale score, and apolipoprotein E ε4 status. Receiver operating characteristic (ROC) curve analyses were used to determine the discrimination accuracy of SMC and MMSE scores, and area under the ROC curve (AUROC) was also calculated. RESULTS: In the highly educated (≥7 years), nondepressed (Geriatric Depression Scale ≤15) subgroup, SMC showed good accuracy in discriminating cognitively impaired subjects from CN after adjusting for potential confounding variables (the AUROC of the adjusted SMC was 0.841 for MCI discrimination, and it was 0.858 for MCI plus early AD discrimination). Both SMC and MMSE scores significantly contributed to differentiating between CN and MCI (OR=2.372, 95% CI=1.086-5.177; OR=0.730, 95% CI=0.566-0.941, respectively) after adjusting for the same covariates. However, in the highly educated and nondepressed subgroups, SMC showed significant predictive power for MCI from CN (OR=3.119, 95% CI=1.190-8.176; OR=3.328, 95% CI=1.320-8.396, respectively), whereas MMSE scores did not. CONCLUSION: Our findings support the usefulness of SMC, which was comparable or even superior to MMSE scores, for detecting MCI or early AD.

Comparing paired biomarkers in predicting quantitative health outcome subject to random censoring.

This paper uses a non-parametric test, based on consistently estimated discrimination accuracy defined as concordance probability between quantitative predictor and outcome, to compare paired biomarkers in predicting a health outcome, possibly subject to random censoring. Comparing with the Wilcoxon test for paired predictors based on Harrell's C-index, we found that the proposed test is better in presence of random censoring, although the two unbiased tests are equivalent for outcome either uncensored or censored by a constant. A simulation study also demonstrates that the bias in estimated difference in concordance probability, due to ignoring random censoring, results in overestimated power, especially when random censoring is heavy. The method was applied in two studies, where the biomarkers measured from the same study subjects are correlated. The first study on 299 school children in Bangladesh found the associations that higher blood arsenic and manganese were related to lower intellectual test scores, while the differences between the biomarkers in predicting the intellectual test scores were not statistically significant. The second study on 418 patients with primary biliary cirrhosis found that the baseline serum bilirubin had greater discrimination accuracy than the baseline serum albumin in predicting survival time.

Orbitofrontal cortex mediates the differential impact of signaled-reward probability on discrimination accuracy.

Orbitofrontal cortex (OFC) function is critical to decision making and behavior based on the value of expected outcomes. While some of the roles the OFC plays in value computations and behavior have been identified, the role of the OFC in modulating cognitive resources based on reward expectancy has not been explored. Here we assessed the involvement of OFC in the interaction between motivation and attention. We tested mice in a sustained-attention task in which explicitly signaling the probability of reward differentially modulates discrimination accuracy. Using pharmacogenetic methods, we generated mice in which neuronal activity in the OFC could be transiently and reversibly inhibited during performance of our signaled-probability task. We found that inhibiting OFC neuronal activity abolished the ability of reward-associated cues to differentially impact accuracy of sustained-attention performance. This failure to modulate attention occurred despite evidence that mice still processed the differential value of the reward-associated cues. These data indicate that OFC function is critical for the ability of a reward-related signal to impact other cognitive and decision-making processes and begin to delineate the neural circuitry involved in the interaction between motivation and attention.