Test Accuracy Research Articles

A novel and efficient risk minimization-based missing value imputation algorithm

Missing value imputation (MVI) is a key task in data science, in which learning models are built from incomplete data. In contrast to externally driven MVI algorithms, this study proposes a novel risk-minimisation-based MVI algorithm (RM-MVI) that considers both the internal characteristics of missing data and the external performance for specific classification applications. RM-MVI is technically designed for labelled data and is applied in two stages: filling with structural risk minimization (SRM) and refining with empirical risk minimization (ERM). In the filling stage, an autoencoder with a single hidden layer is trained on the original dataset without missing values. Missing values are first initialised with random numbers, and the imputation values are then preliminarily optimised based on the derived updating rule to minimise the structural risk-oriented objective function. After the imputation values have been preliminarily optimised in the filling stage, a neural-network-based classifier is trained in the refining stage to optimise the imputation values sophisticatedly by reducing the empirical risk. Experiments were conducted on several benchmark datasets to validate the feasibility, rationality, and effectiveness of the proposed RM-MVI algorithm. The results show that (1) the optimization processes of the imputation values corresponding to the SRM and ERM are convergent so that the optimized imputation values can be obtained; (2) SRM can ensure distribution consistency of the imputation values that are preliminarily optimised in the filling stage, while ERM can optimise the imputation values sophisticatedly in the refining stage, which is more helpful for classifier training; and (3) the RM-MVI algorithm can yield considerably better MVI performance on benchmark datasets than 11 well-known MVI algorithms, such as a 26% higher distribution consistency ratio and 2% to 5% higher testing accuracies for 6 classifiers on average. This demonstrates that RM-MVI is a viable approach for addressing MVI problems.

Outcomes of Intraoperative Frozen Section Examination of Surgical Resection Margins of the Penis in Penile Cancer

IntroductionFrozen section examination (FSE) of the tumor resection margins is important during penile-preserving surgery (PPS) in penile cancer. The margin status will impact on how much penile or urethral tissue is excised. We aim to evaluate the outcomes of intraoperative FSE of resection margins in PPS. Patients and MethodsA retrospective analysis of patients with penile squamous cell carcinoma (SCC) who underwent a FSE of resection margins between 2010 and 2022 was conducted. FSEs were compared with the final histopathological analysis and the Diagnostic Testing Accuracy (DTA): sensitivity, specificity, positive (PPV) and negative predictive values (NPV) were calculated. ResultsOverall, 137 FSE were performed. The median (IQR) age was 65 (53-75 years). 118 (86.1%) patients had negative FSE margins, 16 (11.7%) had positive FSE margins and 3 (2.2%) had equivocal (atypical cells) results. The sensitivity, specificity, PPV, NPV and diagnostic accuracy of penile FSE were 66.7%, 100%, 100%, 93.2% and 94% respectively. 18 patients underwent further resection in the same episode due to a positive or equivocal FSE and 12 (66.7%) achieved negative margins. Limitations include the retrospective nature of the study and lack of control arm to compare with. ConclusionsIntraoperative FSE performed at our center for the assessment of penile SCC margins is 66.7% sensitive and 100% specific. FSE should be considered in PPS, as it's an essential and a reliable diagnostic tool in minimizing over-treatment.

Diagnostic Accuracy of Contrast Enhanced Computed Tomography (CECT) in Cervical Lymph Node Metastasis of Oral Carcinoma: A Systematic Review and Meta-Analysis.

To compare the diagnostic accuracy of Contrast Enhanced Computed Tomography (CECT) compared to conventional imaging modalities and histopathological investigation in cervical lymph node metastasis in adults through a novel meta- analysis. The review protocol is registered under PROSPERO(CRD42021225704) and performed in accordance to Preferred Reporting Items for Systematic Reviews and Meta-Analysis - Diagnostic Test Accuracy (PRISMA- DTA) checklist. Databases like PubMed, Google Scholar, EBSCOhost were searched from 2000 to 2023 to identify the diagnostic potential of CECT in cervical lymph node metastasis of oral carcinoma. True-positive, false-positive, true-negative, false-negative, sensitivity, specificity values were extracted or calculated if not present for each study. Quality of selected studies was evaluated based on Quality assessment of diagnostic accuracy studies (QUADAS)- 2 tool. Meta-analysis was performed in Meta-Disc 1.4 software and Review Manager 5.3 using a bivariate model parameter for the pooled sensitivity and pooled specificity. Additional analysis was performed in terms of positive likelihood ratio (+LR), negative likelihood ratio (-LR), diagnostic odds ratio (DOR) and summary receiver operating characteristics (SROC) with Area Under Curve (AUC) and p<0.05 as statistically significant. Six studies were included for qualitative synthesis and as well as for meta-analysis. Two studies had high risk of bias while four studies had low risk of bias. 651 patients underwent CECT and were taken for meta-analysis. The meta-analysis revealed that CECT for diagnosing cervical lymph node metastasis had a pooled sensitivity of 71%, pooled specificity of 14% with 60% Area Under Curve (AUC). +LR of 0.84, -LR of 1.36 and DOR of 0.59. CECT has an overall fair diagnostic ability and is a valid and reliable tool in diagnosing the target condition overcoming high reliance on master specialized capacity for their execution and understanding like other conventional imaging techniques. CECT can be concluded for secondary level of prevention for cervical node metastasis of oral carcinoma under early diagnosis and prompt treatment. However, further standardized accuracy studies are indicated to improve the overall diagnostic accuracy of CECT.

Deep convolutional neural networks for filtering out normal frames in reviewing wireless capsule endoscopy videos

Wireless capsule endoscopy (WCE) has emerged as a valuable non-invasive technique for visualizing the entire gastrointestinal (GI) tract. However, manual evaluation of WCE videos is a time-consuming and costly process. In this study, we present a novel diagnostic assistant system that employs deep convolutional neural networks (DCNNs) to accelerate the evaluation process. Our primary objective is to achieve a high negative predictive value (NPV), which is essential for the efficient identification of normal frames. Six distinct DCNN models were developed and implemented with this objective in mind. The models were trained on a limited dataset encompassing common GI pathologies that reflect real clinical scenarios. Each DCNN architecture comprises a convolutional part derived from renowned pre-trained networks and a custom-designed classifier block optimized for high NPV and classification accuracy. Following a comprehensive assessment utilizing the 5-fold cross-validation approach, the VG_BFCG model was identified as the most effective, exhibiting an average test accuracy of 0.946 and an NPV of 0.983. Moreover, in the event of encountering novel pathologies not present in the training data, our models exhibited robustness in NPV, which is of great importance for practical applications. For example, the DN_BFCG model demonstrated consistent performance, with an NPV exceeding 0.99 across a range of new pathologies. This validates the reliability of our models in clinical settings. Our findings suggest that our developed DCNN architectures have the potential to enhance the efficiency and accuracy of WCE video analysis, which could transform the landscape of gastroenterological diagnostics.

Evaluating a web-based visual acuity and refractive error self-assessment tool in myopic children.

Demands for myopia management are rising. A web-based tool that allows home-performed self-assessments of visual acuity (VA) and refractive error may enable hybrid care pathways and aid in identifying those with deteriorating visual performance. The tool has been validated in adult populations, but has yet to be evaluated in children. This study compared home-performed VA and refraction self-assessments to conventional measurements obtained at the clinic in a population of myopic children. Myopic children aged ≥6 years old were invited to perform web-based eye tests at home, assisted by a parent. At two myopia control clinics, they also underwent measurements of VA using a Snellen chart and refractive error using cycloplegic autorefraction. Agreement between the tests, repeatability of the web-based test and associations between clinical characteristics and web-based test accuracy were evaluated. A total of 147 children were enrolled, of whom 116 (51% male; mean age 13 ± 3 years; mean spherical equivalent refraction (SEQ) -5.58 ± 3.05) performed the web-based tests at home. Overall, the home-performed VA self-assessment and the Snellen chart assessment at the clinic agreed well (mean difference 0.03 ± 0.11 logMAR). A significant proportional bias was identified (β 0.65, p < 0.001), indicating underestimated web-based VA scores when the child's vision declined. The sensitivity to detect VA poorer than 0.10 logMAR was 94%; the specificity was 71%. The web-based refractive error algorithm measured more myopia progression compared to clinic observations (mean difference SEQ 0.40 ± 0.51 dioptres). Age, sex or use of atropine drops were not significantly associated with test accuracy. The web-based test for self-assessing vision, performed at home by children with assistance from their parents, yielded VA scores with a precision similar to Snellen chart testing conducted in a clinical setting. However, the web-based refractive error algorithm overestimated myopia progression and requires recalibration for this specific age group.

Interpretation of results of PCR and B-D-glucan for the diagnosis of Pneumocystis Jirovecii Pneumonia in immunocompromised adults with acute respiratory failure

BackgroundThe accuracy of a diagnostic test depends on its intrinsic characteristics and the disease incidence. This study aims to depict post-test probability of Pneumocystis pneumonia (PJP), according to results of PCR and Beta-D-Glucan (BDG) tests in patients with acute respiratory failure (ARF).Materials and methodsDiagnostic performance of PCR and BDG was extracted from literature. Incidence of Pneumocystis pneumonia was assessed in a dataset of 2243 non-HIV immunocompromised patients with ARF. Incidence of Pneumocystis pneumonia was simulated assuming a normal distribution in 5000 random incidence samples. Post-test probability was assessed using Bayes theorem.ResultsIncidence of PJP in non-HIV ARF patients was 4.1% (95%CI 3.3-5). Supervised classification identified 4 subgroups of interest with incidence ranging from 2.0% (No ground glass opacities; 95%CI 1.4–2.8) to 20.2% (hematopoietic cell transplantation, ground glass opacities and no PJP prophylaxis; 95%CI 14.1–27.7). In the overall population, positive post-test probability was 32.9% (95%CI 31.1–34.8) and 22.8% (95%CI 21.5–24.3) for PCR and BDG, respectively. Negative post-test probability of being infected was 0.10% (95%CI 0.09–0.11) and 0.23% (95%CI 0.21–0.25) for PCR and BDG, respectively. In the highest risk subgroup, positive predictive value was 74.5% (95%CI 72.0-76.7) and 63.8% (95%CI 60.8–65.8) for PCR and BDG, respectively.ConclusionAlthough both tests yield a high intrinsic performance, the low incidence of PJP in this cohort resulted in a low positive post-test probability. We propose a method to illustrate pre and post-test probability relationship that may improve clinician perception of diagnostic test performance according to disease incidence in predefined clinical settings.

Utility of Point-of-Care Diagnosis of Chronic Heart Failure Using an Express Test for Semi-Quantitative Determination of NT-proBNP Levels.

To evaluate the accuracy of a rapid test for semi-quantitative determination of NT-proBNP levels in the diagnosis of CHF in comparison with quantitative assessment; to study the strength of the association of the results of this NT-proBNP test with indicators of the CHF severity. The concentration of NT-proBNP was determined in 44 patients at bedside both semi-quantitatively using an express test (BioTest, Novosibirsk) and quantitatively in a laboratory. In 11 patients, the severity of CHF was assessed with the CHF Clinical Status Scale (CSS). Echocardiography was performed in all patients. The sensitivity of the quantitative and semi-quantitative tests coincided and was 95%. The specificity of the quantitative test was 100% in our study, whereas the semi-quantitative test showed a specificity of 92%. The negative predictive value of either test was 96%. The diagnostic accuracy was 98% and 93%, respectively. In patients with significantly high NT-proBNP concentrations, the semi-quantitative test demonstrated a reduced ability to verify values above 1800 pg/ml; in patients with threshold concentrations, the semi-quantitative test showed an increased subthreshold sensitivity. Increases in the NT-proBNP concentration correlated with the severity of CHF according to the stage of the disease. Due to the sufficiently high sensitivity, specificity, ease of use, and speed of obtaining results, the rapid test for semi-quantitative measuring NT-proBNP is promising for outpatient screening bedside diagnosis of CHF and in the emergency room to confirm or exclude CHF. When determining the dynamics of NT-proBNP during the treatment of CHF, the use of the semi-quantitative rapid test with visual assessment of the results may produce an error compared to the quantitative assessment, which will probably not allow tracking the effect of therapy or predicting exacerbation of the disease.

Rancang Bangun SPK Kualitas Air Sungai Metode Fuzzy Tsukamoto (Studi Kasus: 4 Kecamatan Karawang)

Research reveals that the Citarum River in West Java is seriously polluted, causing problems for humans and the environment, so monitoring of river water quality is necessary. This research proposes the development of a web-based Decision Support System with the SDLC Waterfall model, using the Fuzzy Tsukamoto method based on seven water parameters (EC, TDS, Salinity, pH, ORP, SG, Temperature) to simplify the determination of water quality. This research uses the Research and Development methodology with the SDLC Waterfall model software development approach, which includes the stages of analyzing data requirements and fuzzy logic systems, designing system architecture and features, implementing technology and process results, functional manual testing and model accuracy, and maintenance in the form of documentation and storage. The results showed that the developed DSS can classify water quality based on 7 parameters with a value scale divided into five categories, namely Good, Poor, Medium, Good, and Very Good. Testing was conducted using data from 8 river points in 4 sub-districts in Karawang Regency on April 1, 2024. The accuracy of the DSS model reached 80%. The development of this DSS is expected to provide an initial overview of water quality in a location without requiring an in-depth technical understanding of the water parameters used. This data can also be an initial indicator to determine whether further action or further investigation is required. Suggestions for future research are to integrate the system with IoT to improve its performance and benefits.

A New Hybrid Model for Artificial Intelligence Assisted Tire Defect Detection: CTLDF+EnC

This paper focuses on an artificial intelligence based worn tire detection system proposed to detect cracks in the tires of vehicle drivers. Although drivers are generally aware of the importance of tire tread depth and air pressure, they are not aware of the risks associated with tire oxidation. However, tire oxidation and cracks can cause significant problems affecting driving safety. In this paper, we propose a new hybrid architecture for tire crack detection, CTLDF+EnC (Cascaded Transfer Learning Deep Features + Ensemble Classifiers), which uses deep features from pre-trained transfer learning methods in combination with ensemble learning methods. The proposed hybrid model utilizes features from nine transfer learning methods and classifiers including Stacking, Soft and Hard voting ensemble learning methods. Unlike X-Ray image-based applications for industrial use, the model proposed in this study can work with images obtained from any digital imaging device. Among the models proposed in the study, the highest test accuracy value was obtained as 76.92% with the CTLDF+EnC (Stacking) hybrid model. With CTLDF+EnC (Soft) and CTLDF+EnC (Solid) models, 74.15% and 72.92% accuracy values were obtained respectively. The results of the study show that the proposed hybrid models are effective in detecting tire problems. In addition, a low-cost and feasible structure is presented.

Diagnostic accuracy and confounders of vagus nerve ultrasound in amyotrophic lateral sclerosis-a single-center case series and pooled individual patient data meta-analysis.

Several single-center studies proposed utility of vagus nerve (VN) ultrasound for detecting disease severity, autonomic dysfunction, and bulbar phenotype in amyotrophic lateral sclerosis (ALS). However, the resulting body of literature shows opposing results, leaving considerable uncertainty on the clinical benefits of VN ultrasound in ALS. Relevant studies were identified up to 04/2024 and individual patient data (IPD) obtained from the respective authors were pooled with a so far unpublished cohort (from Munich). An IPD meta-analysis of 109 patients with probable or definite ALS (El Escorial criteria) and available VN cross-sectional area (CSA) was performed, with age, sex, ALS Functional Rating Scale-revised (ALSFRS-R), disease duration, and bulbar phenotype as independent variables. Mean age was 65years (± 12) and 47% of patients (± 12) had bulbar ALS. Mean ALSFRS-R was 38 (± 7), and mean duration was 18months (± 18). VN atrophy was highly prevalent [left: 67% (± 5), mean CSA 1.6mm2 (± 0.6); right: 78% (± 21), mean CSA 1.8 mm2 (± 0.7)]. VN CSA correlated with disease duration (mean slope: left -0.01; right -0.01), but not with ALSFRS-R (mean slope: left 0.004; mean slope: right -0.002). Test accuracy for phenotyping bulbar vs. non-bulbar ALS was poor (summary receiver operating characteristic area under the curve: left 0.496; right 0.572). VN atrophy in ALS is highly prevalent and correlates with disease duration, but not with ALSFRS-R. VN CSA is insufficient to differentiate bulbar from non-bulbar ALS phenotypes. Further studies are warranted to analyze the link between VN atrophy, autonomic impairment, and survival in ALS.

A new method to orient a 3-dimensional facial model to natural head position: A preliminary report on accuracy and reproducibility

ObjectivesTo establish and validate a novel method to orient a 3-dimensional (3D) facial model to natural head position (NHP) in a stereophotogrammetric system using a 2-dimensional frontal full-face photograph of NHP. Material and MethodsSpecific technique procedure was reported for our method, and in vitro model experiment was performed for accuracy test. A preliminary volunteer study was then planned for reproducibility test. ResultsThe accuracy on a 3D-printed test model was within 0.15°. Within an observational cohort of 22 dental students, the angular deviations of different maxillofacial regions (e.g., central forehead, left and right zygomatic regions, apex of nose and mental region) were no more than 2° between the 3D NHP models acquired with a shorter time-interval (1 h from baseline) or a longer time-interval (7 days from baseline), which were all considered clinically insignificant. In addition, the angular deviations were significantly larger with a 7d-interval than with a 1h-interval, indicting a decline in 3D NHP reproducibility over short time duration. ConclusionThe current method may represent a clinically useful protocol for recording and transferring 3D NHP in stereophotogrammetry. Clinical RelevanceIt may provide reliable and meaningful reference information for evaluating craniofacial morphology, and be of clinical use in the diagnosis, treatment and follow-ups of patients with aesthetic or deformed craniofacial problems.

Electronic Load Device for Testing DC Power Supply

The project deals with a programmable electronic load device capable of dynamic variations of its load parameters to simulate a myriad of electrical conditions. Hence, one of the objectives of this paper is to implement, in a designed setup, variously heated loads, such as constant current, voltage, resistance, and power modes. That said, the designed system has hardware and software aspects that integrate advanced microcontroller capabilities so that high accuracy is assured through a series of test procedures on the system. The hardware design of the project encompasses choosing various resistors and relays modules (ESP32), selection of the proper power supplies, and microcontrollers. A program is being developed in the Arduino IDE that will program the microcontroller to do dynamic load control.It all began with the circuit design and resistor value choice, taking into consideration their power rating about effective heat dissipation. Dynamic loading control by the ESP32 microcontroller allows variable changes which are automatic in nature. On top of this, relay modules were used for effective switching of load and its adjustments, hence making the system flexible and responsive. The firmware developed in Arduino IDE automates control processes, which enables the smooth running of operations. In this regard, our experiments output results that show the device—a programmable electronic load mimicking load situations and providing performance information. Notably, the automated control system coupled with adjustable load capabilities increased the testing process efficiency and accuracy to a great deal. The overall effect of the programmable electronic load device is that it makes testing more professional by way of automation and provision of precise management for the test settings of the load. Its versatility also provides simulation of a wide scope of scenarios accommodating test conditions. It is designed to cater to heat regulation and power dispersion for the reliability of performance whiles in operation

Forecasting alzheimer's disease in its early stages with the use of machine learning algorithms

When it comes to dementia in the elderly, Alzheimer's disease (AD) takes the cake. Metabolic disorders impact a huge portion of the global population; as a result, there is a growing interest in using machine learning to better understand these conditions. The annual increase in their incidence rates is quite concerning. Neurodegenerative alterations impact the brain in Alzheimer's disease. Diseases impacting memory and functioning will affect healthcare providers, patients, and families at an increasing rate as the population ages. The monetary, social, and economic spheres will all feel these impacts to a significant degree. It is difficult to foretell the onset of Alzheimer's disease in its early stages. Compared to treatments administered later in the disease's progression, those administered early in the disease's progression are more effective and have less mild side effects. Finding the optimal parameters for Alzheimer's disease prediction has involved the use of several techniques, including Decision Tree, Random Forest, Support Vector Machine, Gradient Boosting, and Voting classifiers. Parameters such as F1-score for ML models, Precision, Recall, and Accuracy are used to evaluate the performance of models that use the Open Access Series of Imaging Studies (OASIS) data to make predictions about Alzheimer's disease. Medical professionals can use the suggested categorization system to identify these illnesses. Reducing yearly mortality rates of Alzheimer's disease through early diagnosis using these ML algorithms is tremendously advantageous. When tested using AD test data, the suggested approach outperforms the competition with an average validation accuracy of more than 90%. In contrast to previous studies, this one achieved a far higher test accuracy score.

DeepDate: A deep fusion model based on whale optimization and artificial neural network for Arabian date classification.

As agricultural technology continues to develop, the scale of planting and production of date fruit is increasing, which brings higher yields. However, the increasing yields also put a lot of pressure on the classification step afterward. Image recognition based on deep learning algorithms can help to identify and classify the date fruit species, even in natural light. In this paper, a deep fusion model based on whale optimization and an artificial neural network for Arabian date classification is proposed. The dataset used in this study includes five classes of date fruit images (Barhi, Khalas, Meneifi, Naboot Saif, Sullaj). The process of designing each model can be divided into three phases. The first phase is feature extraction. The second phase is feature selection. The third phase is the training and testing phase. Finally, the best-performing model was selected and compared with the currently established models (Alexnet, Squeezenet, Googlenet, Resnet50). The experimental results show that, after trying different combinations of optimization algorithms and classifiers, the highest test accuracy achieved by DeepDate was 95.9%. It takes less time to achieve a balance between classification accuracy and time consumption. In addition, the performance of DeepDate is better than that of many deep transfer learning models such as Alexnet, Squeezenet, Googlenet, VGG-19, NasNet, and Inception-V3. The proposed DeepDate improves the accuracy and efficiency of classifying date fruits and achieves better results in classification metrics such as accuracy and F1. DeepDate provides a promising classification solution for date fruit classification with higher accuracy. To further advance the industry, it is recommended that stakeholders invest in technology transfer programs to bring advanced image recognition and AI tools to smaller producers, enhancing sustainability and productivity across the sector. Collaborations between agricultural technologists and growers could also foster more tailored solutions that address specific regional challenges in date fruit production.

Multi-Dimensional Fragmentomics Enables Early and Accurate Detection of Colorectal Cancer.

Colorectal cancer (CRC) is frequently diagnosed in advanced stages, highlighting the need for developing approaches for early detection. Liquid biopsy using cell-free DNA (cfDNA) fragmentomics is a promising approach, but the clinical application is hindered by complexity and cost. This study aimed to develop an integrated model using cfDNA fragmentomics for accurate, cost-effective early-stage CRC detection. Plasma cfDNA was extracted and sequenced from a training cohort of 360 participants, including 176 CRC patients and 184 healthy controls. An ensemble stacked model comprising five machine learning models was employed to distinguish CRC patients from healthy controls using five cfDNA fragmentomic features. The model was validated in an independent cohort of 236 participants (117 CRC patients and 119 controls) and a prospective cohort of 242 participants (129 CRC patients and 113 controls). The ensemble stacked model showed remarkable discriminatory power between CRC patients and controls, outperforming all base models and achieving a high area under the ROC curve (AUC) of 0.986 in the validation cohort. It reached 94.88% sensitivity and 98% specificity for detecting CRC in the validation cohort, with sensitivity increasing as cancer progressed. The model also demonstrated consistently high accuracy in within-run and between-run tests and across various conditions in healthy individuals. In the prospective cohort, it achieved 91.47% sensitivity and 95.58% specificity. This integrated model capitalizes on the multiplex nature of cfDNA fragmentomics to achieve high sensitivity and robustness, offering significant promise for early CRC detection and broad patient benefit.

Early detection of uterine corpus endometrial carcinoma utilizing plasma cfDNA fragmentomics

BackgroundUterine corpus endometrial carcinoma (UCEC) is a prevalent gynecologic malignancy with a favorable prognosis if detected early. However, there is a lack of accurate and reliable early detection tests for UCEC. This study aims to develop a precise and non-invasive diagnostic method for UCEC using circulating cell-free DNA (cfDNA) fragmentomics.MethodsPeripheral blood samples were collected from all participants, and cfDNA was extracted for analysis. Low-coverage whole-genome sequencing was performed to obtain cfDNA fragmentomics data. A robust machine learning model was developed using these features to differentiate between UCEC and healthy conditions.ResultsThe cfDNA fragmentomics-based model showed high predictive power for UCEC detection in training (n = 133; AUC 0.991) and validation cohorts (n = 89; AUC 0.994). The model manifested a specificity of 95.5% and a sensitivity of 98.5% in the training cohort, and a specificity of 95.5% and a sensitivity of 97.8% in the validation cohort. Physiological variables and preanalytical procedures had no significant impact on the classifier’s outcomes. In terms of clinical benefit, our model would identify 99% of Chinese UCEC patients at stage I, compared to 21% under standard care, potentially raising the 5-year survival rate from 84 to 95%.ConclusionThis study presents a novel approach for the early detection of UCEC using cfDNA fragmentomics and machine learning showing promising sensitivity and specificity. Using this model in clinical practice could significantly improve UCEC management and control, enabling early intervention and better patient outcomes. Further optimization and validation of this approach are warranted to establish its clinical utility.

An attention-based deep learning for acute lymphoblastic leukemia classification

The bone marrow overproduces immature cells in the malignancy known as Acute Lymphoblastic Leukemia (ALL). In the United States, about 6500 occurrences of ALL are diagnosed each year in both children and adults, comprising nearly 25% of pediatric cancer cases. Recently, many computer-assisted diagnosis (CAD) systems have been proposed to aid hematologists in reducing workload, providing correct results, and managing enormous volumes of data. Traditional CAD systems rely on hematologists’ expertise, specialized features, and subject knowledge. Utilizing early detection of ALL can aid radiologists and doctors in making medical decisions. In this study, Deep Dilated Residual Convolutional Neural Network (DDRNet) is presented for the classification of blood cell images, focusing on eosinophils, lymphocytes, monocytes, and neutrophils. To tackle challenges like vanishing gradients and enhance feature extraction, the model incorporates Deep Residual Dilated Blocks (DRDB) for faster convergence. Conventional residual blocks are strategically placed between layers to preserve original information and extract general feature maps. Global and Local Feature Enhancement Blocks (GLFEB) balance weak contributions from shallow layers for improved feature normalization. The global feature from the initial convolution layer, when combined with GLFEB-processed features, reinforces classification representations. The Tanh function introduces non-linearity. A Channel and Spatial Attention Block (CSAB) is integrated into the neural network to emphasize or minimize specific feature channels, while fully connected layers transform the data. The use of a sigmoid activation function concentrates on relevant features for multiclass lymphoblastic leukemia classification The model was analyzed with Kaggle dataset (16,249 images) categorized into four classes, with a training and testing ratio of 80:20. Experimental results showed that DRDB, GLFEB and CSAB blocks’ feature discrimination ability boosted the DDRNet model F1 score to 0.96 with minimal computational complexity and optimum classification accuracy of 99.86% and 91.98% for training and testing data. The DDRNet model stands out from existing methods due to its high testing accuracy of 91.98%, F1 score of 0.96, minimal computational complexity, and enhanced feature discrimination ability. The strategic combination of these blocks (DRDB, GLFEB, and CSAB) are designed to address specific challenges in the classification process, leading to improved discrimination of features crucial for accurate multi-class blood cell image identification. Their effective integration within the model contributes to the superior performance of DDRNet.

Blood Biomarkers to Detect Alzheimer Disease in Primary Care and Secondary Care

An accurate blood test for Alzheimer disease (AD) could streamline the diagnostic workup and treatment of AD. To prospectively evaluate a clinically available AD blood test in primary care and secondary care using predefined biomarker cutoff values. There were 1213 patients undergoing clinical evaluation due to cognitive symptoms who were examined between February 2020 and January 2024 in Sweden. The biomarker cutoff values had been established in an independent cohort and were applied to a primary care cohort (n = 307) and a secondary care cohort (n = 300); 1 plasma sample per patient was analyzed as part of a single batch for each cohort. The blood test was then evaluated prospectively in the primary care cohort (n = 208) and in the secondary care cohort (n = 398); 1 plasma sample per patient was sent for analysis within 2 weeks of collection. Blood tests based on plasma analyses by mass spectrometry to determine the ratio of plasma phosphorylated tau 217 (p-tau217) to non-p-tau217 (expressed as percentage of p-tau217) alone and when combined with the amyloid-β 42 and amyloid-β 40 (Aβ42:Aβ40) plasma ratio (the amyloid probability score 2 [APS2]). The primary outcome was AD pathology (determined by abnormal cerebrospinal fluid Aβ42:Aβ40 ratio and p-tau217). The secondary outcome was clinical AD. The positive predictive value (PPV), negative predictive value (NPV), diagnostic accuracy, and area under the curve (AUC) values were calculated. The mean age was 74.2 years (SD, 8.3 years), 48% were women, 23% had subjective cognitive decline, 44% had mild cognitive impairment, and 33% had dementia. In both the primary care and secondary care assessments, 50% of patients had AD pathology. When the plasma samples were analyzed in a single batch in the primary care cohort, the AUC was 0.97 (95% CI, 0.95-0.99) when the APS2 was used, the PPV was 91% (95% CI, 87%-96%), and the NPV was 92% (95% CI, 87%-96%); in the secondary care cohort, the AUC was 0.96 (95% CI, 0.94-0.98) when the APS2 was used, the PPV was 88% (95% CI, 83%-93%), and the NPV was 87% (95% CI, 82%-93%). When the plasma samples were analyzed prospectively (biweekly) in the primary care cohort, the AUC was 0.96 (95% CI, 0.94-0.98) when the APS2 was used, the PPV was 88% (95% CI, 81%-94%), and the NPV was 90% (95% CI, 84%-96%); in the secondary care cohort, the AUC was 0.97 (95% CI, 0.95-0.98) when the APS2 was used, the PPV was 91% (95% CI, 87%-95%), and the NPV was 91% (95% CI, 87%-95%). The diagnostic accuracy was high in the 4 cohorts (range, 88%-92%). Primary care physicians had a diagnostic accuracy of 61% (95% CI, 53%-69%) for identifying clinical AD after clinical examination, cognitive testing, and a computed tomographic scan vs 91% (95% CI, 86%-96%) using the APS2. Dementia specialists had a diagnostic accuracy of 73% (95% CI, 68%-79%) vs 91% (95% CI, 88%-95%) using the APS2. In the overall population, the diagnostic accuracy using the APS2 (90% [95% CI, 88%-92%]) was not different from the diagnostic accuracy using the percentage of p-tau217 alone (90% [95% CI, 88%-91%]). The APS2 and percentage of p-tau217 alone had high diagnostic accuracy for identifying AD among individuals with cognitive symptoms in primary and secondary care using predefined cutoff values. Future studies should evaluate how the use of blood tests for these biomarkers influences clinical care.

Revisiting serum storage protocols: assessing the impact of temperature variations on indirect immunofluorescence testing for pemphigus and bullous pemphigoid.

Indirect immunofluorescence (IIF) plays a crucial role in the diagnosis of pemphigus and bullous pemphigoid (BP) by detecting the presence of circulating autoantibodies in the serum of patients. The standard serum transportation method requires delivery to laboratories at 2-8 °C within a day and storage at -20 to -80 °C. However, this protocol poses logistical challenges. We conducted a study to assess how temperature variations affect the effectiveness of IIF tests. This case-control study analysed 203 serum specimens: 102 from patients with pemphigus and 101 from patients with BP. Specimens were stored at -80 °C (control), 24 °C, and 40 °C for seven days before analysis to investigate variations in IIF titres compared to the control conditions. In pemphigus serum, 95% at 24 °C and 76% at 40 °C showed no titre difference compared to controls. Similarly, 89% of BP serum at 24 °C and 82% at 40 °C matched the control titres. While 57 specimens across both groups experienced reduced titres, the decrease was primarily marginal (one-step reduction in 54 cases, two-step in 3), with no transition from positive to negative results. Storing serum at 24-40 °C for up to seven days before testing slightly influences IIF outcomes for pemphigus and BP. These findings could prompt a significant revision in the existing strict transport guidelines, ensuring efficient use of resources without sacrificing the accuracy of diagnostic tests.

Clinical and biochemical data for the diagnosis of endogenous hypercortisolism: the "Cushingomic" approach.

The clinical presentation of Cushing's syndrome (CS) overlaps with common conditions. Recommended screening tests are serum cortisol after 1-mg overnight dexamethasone suppression test (DST), urinary free cortisol (UFC), and late-night salivary cortisol (LNSC). We analyzed the diagnostic accuracy of screening tests in 615 patients without CS (263 suspected CS, 319 adrenal and 33 pituitary incidentaloma) and 40 with CS. Principal component analysis, K-means clustering, and neural network were used to compute an integrated analysis among tests, comorbidities, and signs/symptoms of hypercortisolism. The diagnostic accuracy of screening tests for CS was high, DST and UFC were slightly superior to LNSC. The threshold of DST should be adapted to the population considered, especially in adrenal incidentaloma with mild autonomous cortisol secretion: the cutoff to differentiate CS should be increased to 196 nmol/L. Diabetes, hypertension, and obesity were more common in patients without CS: the direction of their vectors was not aligned and their correlation with screening tests was poor. Clustering allowed us to differentiate those patients without CS in cluster one (aged osteoporotic patients with impaired screening tests), cluster two (hypertensive and metabolic phenotype), and cluster three (young subjects with a low likelihood of overt CS). A neural network model that combined screening tests and clinical presentation was able to predict the CS diagnosis in the validation cohort with 99% precision and 86% accuracy. Despite the high diagnostic accuracy of screening tests to detect CS, cortisol-related comorbidities or adrenal incidentaloma should be considered when interpreting a positive test.