Perfect Accuracy Research Articles

Abstract A064: Prediction of positive prostate biopsy is significantly improved in Black and Hispanic men when serum PVT1 exon 9 copy number is combined with serum prostate specific antigen

Abstract Background Men of African ancestry (moAA) experience the highest incidence and mortality rates of prostate Cancer (PCa) in the United States, where 1 in 8 men are expected to be diagnosed with the disease in their lifetime. Yet, most biomarkers used in PCa screening today, which are based on prostate specific antigen (PSA), are developed and validated in predominantly White populations with weak or nonexistent validation in moAA, making moAA more likely to be subjected to unnecessary prostate biopsies. To reduce this disparity in PCa detection and management, there is a critical need for new biomarkers that target the unique genetic make-up of moAA. We recently showed that plasmacytoma variant translocation 1 (PVT1), exons 4A, 4B, and 9 are overexpressed in the prostate tissues and detected in higher copy numbers in serum of moAA with PCa. In this study, we investigated the predictive power of these potential biomarkers in detecting PCa across different racial populations. Methodology Forty serum samples from White, Hispanic, and Black men, 50 percent of whom had PCa, were analyzed to obtain copy numbers for PVT1 exons 4A, 4B, and 9. Seven logistic regression models were developed to predict PCa using single biomarkers, all pairwise combinations of biomarkers and all three biomarkers. The median area under the receiver operator characteristic curve (AUC) was used to measure the predictive accuracy of each model after a repeated, stratified k-fold cross validation. Model accuracy was first evaluated using data from all races and then for just the Black and Hispanic subpopulation. Our best performing models were compared with PSA as a predictor of PCa. We also evaluated the predictive accuracy of composite models that combined our biomarkers with PSA. Results In predicting PCa for all races, our model with only PVT1 exon 9 achieved the highest median AUC of 0.71, which was better than the AUC of 0.63 obtained with just PSA. When PSA and PVT1 exon 9 were combined, the median AUC increased to 0.75. Combining all three biomarkers with PSA increased the median AUC to 0.813. For the Black and Hispanic subpopulation, PVT1 exon 9 achieved the highest AUC of 0.92 which was significantly higher than the median AUC of 0.63 achieved with PSA. No significant change in AUC was observed when the PVT1 biomarkers were combined with PSA. Conclusion Our results show that PVT1 exon 9 is more accurate than PSA in predicting PCa, achieving near perfect accuracy within the Black and Hispanic subpopulation. Whereas combining PSA with PVT1 biomarkers improved the performance of PSA for all races, no significant improvement in PSA’s performance was observed in Black and Hispanic subpopulations. Thus, PVT1 exon 9 may hold the key to reducing disparity in PCa detection among moAA while improving the predictive accuracy of PSA for the general population Citation Format: Emmanuel Owusu Asante-Asamani, Dinuka Sewandi De Silva, Gargi Pal, Michael Liss, Robin Leach, Olorunseun Ogunwobi. Prediction of positive prostate biopsy is significantly improved in Black and Hispanic men when serum PVT1 exon 9 copy number is combined with serum prostate specific antigen [abstract]. In: Proceedings of the AACR Special Conference: Advances in Prostate Cancer Research; 2023 Mar 15-18; Denver, Colorado. Philadelphia (PA): AACR; Cancer Res 2023;83(11 Suppl):Abstract nr A064.

Spectral-domain modelling of a non-linear wave energy converter: Analytical derivation and computational experiments

Harvesting energy from waves is gaining popularity in the last decades, due to growing technological maturity and reduction of manufacturing costs of Wave Energy Converter (WEC) devices. However, the optimal design of WECs represents a challenge to its large-scale commercialisation. Behind this multi-objective optimisation, a significant number of simulations are necessary to identify the optimal configuration of the device that minimise certain techno-economic parameter. Frequency-domain models are suggested in an early stage since they provide a simple and fast simulation tool. On the other hand, non linear time-domain numerical models are more suitable during an advanced design phase where it is essential to assess performance and acting loads with greater reliability. The aim of this work is the development of a spectral-domain model of a WEC device that combines the computational speed of frequency-domain models with the accuracy of non-linear models to improve the design in early stage. Three different models of the ISWEC, a WEC developed by the MOREnergy Lab of Politecnico di Torino (Italy), are derived and compared: a fully non-linear time-domain model, a frequency-domain model, and a spectral-domain model. The spectral-domain modelling of the ISWEC system represents an important novelty for the technological development of the system and an example framework for similar technologies. Moreover, in the energy harvesting field, no previous work addressed the analytical representation of a realistic electro-mechanical converter with both torque and power saturation, and the representation of a gyro-pendulum system in spectral-domain remains untreated. The numerical experiments demonstrate that the spectral-domain model shows an improved accuracy compared with the frequency-domain model, with a reduction of percentage error of the computed gross power in the order of two times. Despite a not perfect accuracy of the spectral-domain model against the time-domain one, the spectral-domain framework is considered as a valid solution due to its computational efficiency, that is demonstrated to be, at least, three order of magnitude higher than the time-domain one.

DILI: A Distribution-Driven Learned Index

Targeting in-memory one-dimensional search keys, we propose a novel DIstribution-driven Learned Index tree ( DILI ), where a concise and computation-efficient linear regression model is used for each node. An internal node's key range is equally divided by its child nodes such that a key search enjoys perfect model prediction accuracy to find the relevant leaf node. A leaf node uses machine learning models to generate searchable data layout and thus accurately predicts the data record position for a key. To construct DILI, we first build a bottom-up tree with linear regression models according to global and local key distributions. Using the bottom-up tree, we build DILI in a top-down manner, individualizing the fanouts for internal nodes according to local distributions. DILI strikes a good balance between the number of leaf nodes and the height of the tree, two critical factors of key search time. Moreover, we design flexible algorithms for DILI to efficiently insert and delete keys and automatically adjust the tree structure when necessary. Extensive experimental results show that DILI outperforms the state-of-the-art alternatives on different kinds of workloads.

Continuous Face Authentication in Real Time using Web-Cam

Abstract: Face authentication is becoming increasingly important for security and identification purposes, but existing systems are often limited by factors such as lighting, facial expressions, and changes in appearance over time. This paper proposes a solution that combines deep learning techniques with adaptive face representation, enabling the system to continuously authenticate the user even in challenging conditions. this research paper also presents a newly developed model for face recognition. The model is based on a deep learning approach that combines convolutional neural networks (CNN) with a recurrent neural network (RNN) architecture. This approach allows the model to not only extract high-level features from facial images, but also to capture temporal dependencies between frames, resulting in a more accurate and robust recognition system. The model was trained and tested on several benchmark datasets, and achieved state-of-the-art performance in terms of accuracy and efficiency. Overall, The proposed model combines convolutional neural networks with a recurrent neural network architecture, allowing it to extract high-level features and capture temporal dependencies between frames for improved accuracy and robustness. The model was trained and tested on a benchmark dataset, achieving an impressive test accuracy of 100%, as well as a perfect precision, recall, and F1-score. Additionally, the paper compares the results of the proposed model to those of a traditional SVM model, which also achieved a perfect accuracy, precision, recall, and F1-score. The comparison highlights the effectiveness of the proposed approach and demonstrates its potential as a highly reliable and efficient authentication solution.

Three-Biomarker Joint Strategy for Early and Accurate Diagnosis of Acute Myocardial Infarction via a Multiplex Electrochemiluminescence Immunoarray Coupled with Robust Machine Learning.

Acute myocardial infarction (AMI) represents a leading cause of death globally. Key to AMI recovery is timely diagnosis and initiation of treatment, ideally within 3 h of symptom onset. Cardiac troponin T (cTnT) is the gold standard yet a low cTnT result cannot rule out AMI at early times. Here, we develop a three-biomarker joint strategy for early and accurate diagnosis of AMI via an electrochemiluminescence (ECL) immunoarray coupled with robust machine learning. The ECL immunoarray is based on an array microchip with a single-electrode and chemiluminescent immuno-Gold (ciGold) nanoassemblies. The ciGold immunoarray was obtained by successively assembling nanocomposites of Cu2+/cysteine complexes and N-(aminobutyl)-N-(ethylisoluminol) bifunctionalized gold nanoparticles combined with chitosan and antibody conjugated gold nanoparticles on the surface of a microchip. Three biomarkers, including cardiac troponin I, heart type fatty acid binding protein, and copeptin, were simultaneously detected in 260 serum samples from patients presenting with chest pain by an innovative multiplexed ECL immunoarray, and classified via the three-biomarker joint assessment model using support vector machines. The model achieved perfect discrimination (100% sensitivity and specificity) for AMI vs non-AMI patients, substantially higher than cTnT alone. Within 12 h of symptom onset, high-sensitivity cardiac troponin T (hs-cTnT) misclassified >20% of patients, while the joint biomarker assessment model retained perfect accuracy. As the time between symptom onset and testing became shorter, the degree to which the joint assessment model outperformed hs-cTnT increased. The proposed three-biomarker joint strategy is obviously superior to hs-cTnT for early and accurate diagnosis of AMI, hopefully reducing AMI mortality and saving limited medical resources.

Estimations of REDD+ opportunity costs: Aligning methods with objectives

Since the early 2000′s when discussions started around the concept of REDD+ , numerous studies estimated the opportunity costs often implicitly viewed as proxies for the total costs of avoided deforestation. We reviewed their methods and the stated objectives to assess their alignment, suitability and robustness. We found a lack of correspondence between methods and objectives, which include setting compensation levels, identifying cost-efficient activities, or assessing financial viability with respect to carbon markets. While the search for perfect accuracy is a fallacy, some major mistakes must be avoided, e.g. equating the costs of avoided deforestation to opportunity costs or including an inappropriate group of economic agents in the calculations. Best practises must be followed to ensure that estimations properly inform intervention design, decisions and policies; they refer to methods applied and their suitability for related decisions. They include: the need to first define objectives and then adapt methods because estimations can be done variously with contrasted results; to cautiously determine the scope - which stakeholders and stages of value chain are considered - while prioritising the broad picture and avoiding getting lost in estimation details; to discuss sensitivity of results and recognise uncertainties; and to consider the implications of scale in the study design. Given the urgency to preserve natural forest and mitigate climate change in the most cost-effective way, we believe that our analysis and its recommendations are a step in this direction with expected increased emissions reductions due to improved interventions.

Statistical power, accuracy, reproducibility and robustness of a graph clusterability test

Not all graphs are clusterable. Not all graphs have a clustered structure and can be meaningfully summarized through vertex clustering. Clusterable graphs are characterized by pockets of densely connected vertices that are only sparsely connected to the remaining graph. In this article, we re-introduce a very simple and intuitive, yet highly informative, statistical hypothesis test for graph clusterability that is based on vertex and neighborhood samples. The goal of this test is to determine if a graph meets the necessary structural conditions to be summarized meaningfully through vertex clusters. Our test is based on the hypothesis that a clusterable graph will display, on average, a local neighborhood induced subgraph density that is greater than the graph’s overall density. The test is also applied to graph comparisons, to test whether one graph has a stronger clustered structure than another. Significance is assessed using the t-statistic. Since it is based on sampling, we provide a focused examination of our test’s sensitivity to sample size. The main contribution of this article is a detailed examination of our test’s accuracy, sensitivity to sample size, conclusion reproducibility and robustness. Our empirical results remain consistent with our earlier conclusions and demonstrate the almost perfect accuracy of our test, even with very small samples of the graph. They also reveal that our test remains robust even under severe departures from the null hypothesis.

Identifying responders to elamipretide in Barth syndrome: Hierarchical clustering for time series data

BackgroundBarth syndrome (BTHS) is a rare genetic disease that is characterized by cardiomyopathy, skeletal myopathy, neutropenia, and growth abnormalities and often leads to death in childhood. Recently, elamipretide has been tested as a potential first disease-modifying drug. This study aimed to identify patients with BTHS who may respond to elamipretide, based on continuous physiological measurements acquired through wearable devices.ResultsData from a randomized, double-blind, placebo-controlled crossover trial of 12 patients with BTHS were used, including physiological time series data measured using a wearable device (heart rate, respiratory rate, activity, and posture) and functional scores. The latter included the 6-minute walk test (6MWT), Patient-Reported Outcomes Measurement Information System (PROMIS) fatigue score, SWAY Balance Mobile Application score (SWAY balance score), BTHS Symptom Assessment (BTHS-SA) Total Fatigue score, muscle strength by handheld dynamometry, 5 times sit-and-stand test (5XSST), and monolysocardiolipin to cardiolipin ratio (MLCL:CL). Groups were created through median split of the functional scores into “highest score” and “lowest score”, and “best response to elamipretide” and “worst response to elamipretide”. Agglomerative hierarchical clustering (AHC) models were implemented to assess whether physiological data could classify patients according to functional status and distinguish non-responders from responders to elamipretide. AHC models clustered patients according to their functional status with accuracies of 60–93%, with the greatest accuracies for 6MWT (93%), PROMIS (87%), and SWAY balance score (80%). Another set of AHC models clustered patients with respect to their response to treatment with elamipretide with perfect accuracy (all 100%).ConclusionsIn this proof-of-concept study, we demonstrated that continuously acquired physiological measurements from wearable devices can be used to predict functional status and response to treatment among patients with BTHS.

Abstract 4172: Serum PVT1 exon 9 copy number is a better predictor of positive prostate biopsy in Black and Hispanic men than serum prostate specific antigen

Abstract Background: Men of African ancestry (moAA) experience the highest incidence and mortality rates of prostate Cancer (PCa) in the United States, where 1 in 8 men are expected to be diagnosed with the disease in their lifetime. Yet, most biomarkers used in PCa screening today, which are based on prostate specific antigen (PSA), are developed and validated in predominantly White populations with weak or nonexistent validation in moAA, making moAA more likely to be subjected to unnecessary prostate biopsies. To reduce this disparity in PCa detection and management, there is a critical need for new biomarkers that target the unique genetic make-up of moAA. We recently showed that plasmacytoma variant translocation 1 (PVT1), exons 4A, 4B, and 9 are overexpressed in the prostate tissues and detected in higher copy numbers in serum of moAA with PCa. In this study, we investigated the predictive power of these potential biomarkers in detecting PCa across different racial populations. Methodology: Forty serum samples from White, Hispanic, and Black men, 50 percent of whom had PCa, were analyzed to obtain copy numbers for PVT1 exons 4A, 4B, and 9. Seven logistic regression models were developed to predict PCa using single biomarkers, all pairwise combinations of biomarkers and all three biomarkers. The median area under the receiver operator characteristic curve (AUC) was used to measure the predictive accuracy of each model after a repeated, stratified k-fold cross validation. Model accuracy was first evaluated using data from all races and then for just the Black or Hispanic subpopulation. Our best performing models were compared with PSA as a predictor of PCa. We also evaluated the predictive accuracy of composite models that combined our biomarkers with PSA. Results: In predicting PCa for all races, our model with only PVT1 exon 9 achieved the highest median AUC of 0.71, which was better than the AUC of 0.63 obtained with just PSA. When PSA and PVT1 exon 9 were combined, the median AUC increased to 0.75. Combining all three biomarkers with PSA increased the median AUC to 0.813. For the Black or Hispanic subpopulation, PVT1 exon 9 achieved the highest AUC of 0.92 which was significantly higher than the median AUC of 0.63 achieved with PSA. No significant change in AUC was observed when the PVT1 biomarkers were combined with PSA. Conclusion: Our results show that PVT1 exon 9 is more accurate than PSA in predicting PCa, achieving near perfect accuracy within the Black or Hispanic subpopulation. Whereas combining PSA with PVT1 biomarkers improved the performance of PSA for all races, no significant improvement in PSA’s performance was observed in Black or Hispanic subpopulation. Thus, PVT1 exon 9 may hold the key to reducing disparity in PCa detection among moAA while improving the predictive accuracy of PSA for the general population. Citation Format: Emmanuel Owusu Asante-Asamani, Sobana Handi D. De Silva, Gargi Pal, Micheal Liss, Robin J. Leach, Olorunseun O. Ogunwobi. Serum PVT1 exon 9 copy number is a better predictor of positive prostate biopsy in Black and Hispanic men than serum prostate specific antigen. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4172.

Outlier galaxy images in the Dark Energy Survey and their identification with unsupervised machine learning

The Dark Energy Survey is able to collect image data of an extremely large number of extragalactic objects, and it can be reasonably assumed that many unusual objects of high scientific interest are hidden inside these data. Due to the extreme size of DES data, identifying these objects among many millions of other celestial objects is a challenging task. The problem of outlier detection is further magnified by the presence of noisy or saturated images. When the number of tested objects is extremely high, even a small rate of noise or false positives leads to a very large number of false detections, making an automatic system impractical. This study applies an automatic method for automatic detection of outlier objects in the first data release of the Dark Energy Survey. By using machine learning-based outlier detection, the algorithm is able to identify objects that are visually different from the majority of the other objects in the database. An important feature of the algorithm is that it allows to control the false-positive rate, and therefore can be used for practical outlier detection. The algorithm does not provide perfect accuracy in the detection of outlier objects, but it reduces the data substantially to allow practical outlier detection. For instance, the selection of the top 250 objects after applying the algorithm to more than 2⋅106 DES images provides a collection of uncommon galaxies. Such collection would have been extremely time-consuming to compile by using manual inspection of the data.

ResNet101 Model Performance Enhancement in Classifying Rice Diseases with Leaf Images

Indonesia is the fourth biggest rice producer in Asia with its production accounting for 35.4 million metric tons yearly. This figure can increase unless rice crop failure is resolved. Identifying rice diseases, however, may serve as an approach to minimizing the risk of crop failure. The classification to detect rice diseases was previously researched using ResNet101 method with 100% accuracy. Despite this perfect accuracy, this approach does not come without an issue, where the prediction is not yet optimal for each label and loss results which are regarded as too high due to overfitting. Departing from this issue, this research aims to improve the model by reducing the layer complexity of the model and comparing two layers structures of the model, two different data, and the ResNet101 model. The performance resulting from the model could be enhanced with the structuring of simple architectural layers. Despite the small quantity of dataset, the model performance can yield 100% accuracy in the classification of rice diseases with a loss value of 2.91%. The model performance in this research experienced a 2.7% increase at the loss value and it could accurately classify the type of rice diseases according to leaf images on each label. The problem solved by this research is that ResNet101 is able to classify rice disease accurately even with a small amount of data by utilizing the appropriate layer arrangement with data requirements. In addition, the overfitting that occurred in previous research can also be resolved properly. This matter proves that the correlation between the layers of the model with the amount of data is very influential.

Specifications and Accuracy of Rainfall Forecast Required for Pre-Release at Multi-Purpose Reservoirs in Japan

Pre-release is the discharge from a reservoir before a flood to enhance flood control capability. Its success depends on the performance of rainfall forecasting. However, there is little information regarding the causal relationship between its performance and the success of pre-release. Therefore, the rainfall forecast required for pre-release at 326 multi-purpose reservoirs in Japan is shown quantitatively in this paper. In our analysis, pre-release was simulated based on tentative rainfall forecasts made using observed rainfall data from a period of 17 years (2006 to 2022) with some processing. Then, outputs were evaluated in terms of two risks: not avoiding emergency spillway gate operation and no recovery of water use capacity. The results of five elements were reached: (1) the characteristics of situations requiring pre-release, the required (2) forecast length and (3) spatial resolution, the required accuracy of (4) the rainfall amount, and (5) the position of rainfall zone. For (1), pre-release is required nationwide in typhoons or stationary fronts at a frequency of four instances per year. For (2) and (3), assuming perfect accuracy, the current specifications of rainfall forecast in Japan: forecast length of 84 h or more and the combined use of 5 km and 20 km spatial resolution are generally effective in themselves. For (4) and (5), possible uncertainties in the rainfall amount and the position of rainfall zone needs to be decreased by one digit for avoiding emergency spillway gate operations, while excessive pre-release tends not to result in no recovery of water use capacity.

Estimating sensitivity and specificity of diagnostic tests using latent class models that account for conditional dependence between tests: a simulation study

BackgroundLatent class models are increasingly used to estimate the sensitivity and specificity of diagnostic tests in the absence of a gold standard, and are commonly fitted using Bayesian methods. These models allow us to account for ‘conditional dependence’ between two or more diagnostic tests, meaning that the results from tests are correlated even after conditioning on the person’s true disease status. The challenge is that it is not always clear to researchers whether conditional dependence exists between tests and whether it exists in all or just some latent classes. Despite the increasingly widespread use of latent class models to estimate diagnostic test accuracy, the impact of the conditional dependence structure chosen on the estimates of sensitivity and specificity remains poorly investigated.MethodsA simulation study and a reanalysis of a published case study are used to highlight the impact of the conditional dependence structure chosen on estimates of sensitivity and specificity. We describe and implement three latent class random-effect models with differing conditional dependence structures, as well as a conditional independence model and a model that assumes perfect test accuracy. We assess the bias and coverage of each model in estimating sensitivity and specificity across different data generating mechanisms.ResultsThe findings highlight that assuming conditional independence between tests within a latent class, where conditional dependence exists, results in biased estimates of sensitivity and specificity and poor coverage. The simulations also reiterate the substantial bias in estimates of sensitivity and specificity when incorrectly assuming a reference test is perfect. The motivating example of tests for Melioidosis highlights these biases in practice with important differences found in estimated test accuracy under different model choices.ConclusionsWe have illustrated that misspecification of the conditional dependence structure leads to biased estimates of sensitivity and specificity when there is a correlation between tests. Due to the minimal loss in precision seen by using a more general model, we recommend accounting for conditional dependence even if researchers are unsure of its presence or it is only expected at minimal levels.

External validation of the Chicago Chiari Outcome Scale in adults with Chiari malformation type I.

The Chicago Chiari Outcome Scale (CCOS) serves as a standardized clinical outcome evaluation tool among patients with Chiari malformation type I (CM-I). While the reliability of this scale has been proven for pediatric patients, the literature lacks CCOS validation when used solely in adults. Therefore, this study aimed to determine the validity of the CCOS in an external cohort of adult patients. The authors retrospectively analyzed the medical records of symptomatic patients with CM-I who underwent posterior fossa decompression between 2010 and 2018 in six neurosurgical departments. Each patient was clinically assessed at the latest available follow-up. Gestalt outcome was determined as improved, unchanged, or worsened compared with the preoperative clinical state. Additionally, the CCOS score was calculated for each patient based on the detailed clinical data. To verify the ability of the CCOS to determine clinical improvement, the area under the receiver operating characteristic (AUROC) curve was evaluated. A logistic regression analysis using all four components of the CCOS (pain symptoms, nonpain symptoms, functionality, and complications) was performed to establish predictors of the improved outcome. Seventy-five individuals with a mean age of 42 ± 15.32 years were included in the study. The mean follow-up duration was 52 ± 33.83 months. Considering gestalt outcome evaluation, 41 patients (54.7%) were classified as improved, 24 (32%) as unchanged, and 10 (13.3%) as worsened. All patients with a CCOS score of 14 or higher improved, while all those with a CCOS score of 8 or lower worsened. The AUROC was 0.986, suggesting almost perfect accuracy of the CCOS in delineating clinical improvement. A CCOS score of 13 showed high sensitivity (0.93) and specificity (0.97) for identifying patients with clinical improvement. Additionally, a meaningful correlation was found between higher CCOS scores in each component and better outcomes. Patient stratification by total CCOS score showed that those categorized as improved, unchanged, and worsened scored prevalently between 13 and 16 points, 10 and 12 points, and 4 and 9 points, respectively. In this adult cohort, the CCOS was found to be almost perfectly accurate in reflecting postoperative clinical improvement. Moreover, all four CCOS components (pain symptoms, nonpain symptoms, functionality, and complications) significantly correlated with patient clinical outcomes.

Correlation Between Gene Expression of Interferon Regulatory Factor-5 and Disease Activity Index in Systemic Lupus Erythematosus Iraqi Patients

Systemic lupus erythematosus (SLE) is a heterogeneous autoimmune disease characterized by elevated levels of circulating anti-nuclear autoantibodies and interferon-alpha (INFs-α). Interferon regulatory factor-5 (IRF5) plays an important role in the induction of type I interferon and pro-inflammatory cytokines, and participates in the SLE pathogenesis. This study aimed to investigate the role of IRF5 gene expression levels in a sample of SLE Iraqi patients and its correlation with disease activity, and to identify its diagnostic ability as a biomarker reflecting disease activity. Blood samples were taken from 45 participants diagnosed with SLE cases classified according to the American College of Rheumatology (ACR) criteria. They were scored via the SLE disease activity index 2000 (SLEDAI-2K) to assess the disease activity and according to it, they were subdivided into “SLE (I) group” (SLEDAI-2k ≤5), and “SLE (II) group” (SLEDAI-2k >5), as well as age and gender matched healthy control group. RNA was isolated from whole blood samples and gene expression levels of IRF5 were determined using real-time polymerase chain reaction (PCR). Our results revealed that the expression levels of the IRF5 gene were significantly increased in SLE (I) and SLE (II) patient groups compared with the control group (p<0.05, and p<0.01) respectively, as well as higher in SLE (II) group than the SLE (I) group (p<0.01). Moreover, the expression levels of IRF5 were found to be related positively and significantly to the disease activity index in both SLE patient groups. The analysis of receiver operator curves (ROC) for gene expression levels of IRF5 in SLE (II) group showed a perfect accuracy to distinguish between SLE patients and healthy individuals (AUC=0.989, sensitivity= 95.5%, and specificity= 88.0%). However, in SLE (I) group showed a good accuracy to discriminate between SLE patients and healthy individuals. (AUC=0.769, sensitivity= 69.6%, and specificity= 80.0%). The correlation between gene expression levels of IRF5 with other parameters revealed that a significant positive correlation was found with uric acid and ALP in SLE (I) group, while in SLE (II) group with urea, creatinine, and uric acid. Our conclusion suggests that the up-regulation of IRF5 gene expression levels correlates positively with disease activity in SLE patients reflecting the possibility of using it as an immunological biomarker for diagnosis, and monitoring the disease flare.

A New Respiratory Diseases Detection Model in Chest X-Ray Images Using CNN

Convolutional Neural Network (CNN)-based deep learning techniques have recently demonstrated increased potential and effectiveness in image recognition applications, such as those involving medical images. Deep-learning models can recognize targets with performance comparable to radiologists when used with CXR. The primary goal of this research is to examine a deep learning technique used on the radiography dataset to detect COVID-19 in X-ray medical images. The proposed system consists of several stages, from pre-processing, passing through the feature reduction using more than one technique, to the classification stage based on a proposed model. The test was applied to the COVID-19 Radiography dataset of normal and three lung infections (COVID-19, Viral Pneumonia, and Lung Opacity). The proposed CNN model has shown its ability to classify COVID, normal, and other lung infections with perfect accuracy of 99.94%. Consequently, the AI-based early-stage detection algorithms will be enhanced, increasing the accuracy of the X-ray-based modality for the screening of various lung diseases.

From Capillary Electrophoresis to Deep Sequencing: An Improved HIV-1 Drug Resistance Assessment Solution Using In Vitro Diagnostic (IVD) Assays and Software.

Drug-resistance mutations were mostly detected using capillary electrophoresis sequencing, which does not detect minor variants with a frequency below 20%. Next-Generation Sequencing (NGS) can now detect additional mutations which can be useful for HIV-1 drug resistance interpretation. The objective of this study was to evaluate the performances of CE-IVD assays for HIV-1 drug-resistance assessment both for target-specific and whole-genome sequencing, using standardized end-to-end solution platforms. A total of 301 clinical samples were prepared, extracted, and amplified for the three HIV-1 genomic targets, Protease (PR), Reverse Transcriptase (RT), and Integrase (INT), using the CE-IVD DeepChek® Assays; and then 19 clinical samples, using the CE-IVD DeepChek® HIV Whole Genome Assay, were sequenced on the NGS iSeq100 and MiSeq (Illumina, San Diego, CA, USA). Sequences were compared to those obtained by capillary electrophoresis. Quality control for Molecular Diagnostics (QCMD) samples was added to validate the clinical accuracy of these in vitro diagnostics (IVDs). Nineteen clinical samples were then tested with the same sample collection, handling, and measurement procedure for evaluating the use of NGS for whole-genome HIV-1. Sequencing analyzer outputs were submitted to a downstream CE-IVD standalone software tailored for HIV-1 analysis and interpretation. The limits of range detection were 1000 to 106 cp/mL for the HIV-1 target-specific sequencing. The median coverage per sample for the three amplicons (PR/RT and INT) was 13,237 reads. High analytical reproducibility and repeatability were evidenced by a positive percent agreement of 100%. Duplicated samples in two distinct NGS runs were 100% homologous. NGS detected all the mutations found by capillary electrophoresis and identified additional resistance variants. A perfect accuracy score with the QCMD panel detection of drug-resistance mutations was obtained. This study is the first evaluation of the DeepChek® Assays for targets specific (PR/RT and INT) and whole genome. A cutoff of 3% allowed for a better characterization of the viral population by identifying additional resistance mutations and improving the HIV-1 drug-resistance interpretation. The use of whole-genome sequencing is an additional and complementary tool to detect mutations in newly infected untreated patients and heavily experienced patients, both with higher HIV-1 viral-load profiles, to offer new insight and treatment strategies, especially using the new HIV-1 capsid/maturation inhibitors and to assess the potential clinical impact of mutations in the HIV-1 genome outside of the usual HIV-1 targets (RT/PR and INT).

Single-molecule identification of folded proteins from nanopore ionic current signatures.

Single-molecule protein sequencing would provide unprecedented insights into cellular processes and the diseases that arise from protein misfunction. Recently, nanopores have emerged as a platform for fast label-free analyte detection guided by the ionic current blockades generated from single molecules translocating under an applied electric field. The major bottleneck to nanopore protein sequencing is the need to distinguish twenty proteinogenic amino acids accurately at every single residue. Here, we demonstrate a proof-of-principle proteome characterization system that relies on ionic current signatures produced by complete translocation of natively folded proteins through an array of nanopores. Using the steric exclusion model, we computed the ionic current spectra for all structurally known proteins, initially excluding all membrane proteins, proteins containing partially-disordered domains, and multimeric protein assemblies - about 4000 proteins in total. For each protein, we computationally simulated a fluid driven tumbling through a nanopore by calculating current for 5000 randomly chosen orientations that changes the effective cross-section and thus modulates current. Ionic current signatures of proteins with varying molecular weight clearly differed by the magnitude of blockades confirming the steric exclusion origin of the current. Interestingly, proteins with similar molecular weight also showed a varied range of mean blockade currents due to changing current signatures originating from the protein shape. Using a support vector machine learning algorithm trained on our data set, we were able to theoretically distinguish the weight and shape of the 4000 chosen proteins to near perfect accuracy, in the absence of instrumentation noise. We then theoretically introduced a thermal noise source into our simulated ionic current recordings to demonstrate the minimum requirements of an experimental system that would permit accurate protein identification from a single molecule translocation.

Head-to-Head Diagnostic Test Accuracy Meta-analysis of Colonoscopy and Fecal Immunochemical Test in Detecting Advanced Colon Neoplasia.

Background: Studies on the use of fecal immunochemical test (FIT) in colorectal screening have long assumed perfect accuracy for colonoscopy. No study to date has directly compared the diagnostic accuracy of colonoscopy and FIT to detect advanced neoplasia (AN) in a head-to-head diagnostic accuracy meta-analysis. Methods: A comprehensive electronic search was performed for a head-to-head comparison of FIT and colonoscopy using a third acceptable reference standard in asymptomatic adults. Cochrane methodology was used to perform a head-to-head diagnostic test accuracy (DTA) meta-analysis. Quality assessment tool for diagnostic accuracy studies-2 (QUADAS-2) was used to assess the risk of bias in included studies. Results: Two studies met the eligibility criteria. Overall sensitivity and specificity were 98.5 (95% CI 96.3-100%) and 100% (99.9-100%) for colonoscopy and 16.4% (10.3-22.6%) and 95.4% (94.3-96.4%) for FIT. Colonoscopy was significantly better than FIT (P < 0.0001). The positive and negative likelihood ratios (LRs) were 1.75 (1.57-1.96) and 0.03 (0.01-0.08) for colonoscopy and 3.02 (2.01-4.55) and 0.88 (0.82-0.95) for FIT, respectively. Conclusion: Colonoscopy provides significantly better diagnostic accuracy to detect AN compared with FIT (GRADE: ⨁⨁◯◯). Our study provided precise sensitivity and specificity of both colonoscopy and FIT and a revision in screening policies based on an updated cost-effectiveness analysis considering the results of the head-to-head analysis.

A comparative assessment of artificial intelligence models used for early prediction and evaluation of chronic kidney disease

Chronic Kidney Disease (CKD) is one of the most prevalent and fatal diseases influencing people on a larger that remains dormant until irreversible damage has been done to an individual’s kidney. Progression of CKD is related to a variety of great complications, including increased incidence of various disorders, anemia, hyperlipidemia, nerve damage, pregnancy complication, and even complete kidney failure. Millions of people die from this disease every year. Diagnosing CKD is a cumbersome task as no major symptoms can be used as a benchmark to detect the disease. In cases when diagnosis persists, some results may be interpreted incorrectly. This study proposes using a deep neural network-based Multi-Layer Perceptron Classifier to diagnose CKD in patients. The model was trained using data from 400 people and considered various symptoms and signs, including age, blood sugar, red blood cell count, etc. Experiments reveal that the proposed model achieves perfect testing accuracy in classification tasks. Our goal is to facilitate introducing Deep Learning approaches to learning from the dataset attribute reports and accurately detecting CKD. The paper’s primary contribution is a Deep Neural Network model for chronic kidney disease diagnosis that achieves 100% accuracy, outperforming standard machine learning models such as support vector machines and naive Bayes classifiers. This paper provides a detailed explanation of the multi-layer perceptron classifier, which uses the deep neural network provided by the PyTorch library as its basis. Neural models can be a better alternative for adaption techniques for classifying chronic kidney disease Because they can handle non-linearity in the data, compute complex data heaps fetched from datasets, and adapt and learn on their own about the key information using the layers of neurons present in the structure.