Accurate Categorization Research Articles

Validation of Military Occupational Specialty as a Proxy for Blast Exposure Using the Salisbury Blast Interview.

Evaluations of large datasets preclude the ability to directly measure individual experiences, instead relying on proxies to infer certain constructs. Blast exposure is a construct of study currently in its infancy, resulting in diverse definitions and measurement across studies. The purpose of the present study was to validate Military Occupational Specialty (MOS) as a proxy for blast exposure in combat veterans. 256 veterans (86.33% male) completed the Salisbury Blast Interview (SBI) and Mid-Atlantic MIRECC Assessment of Traumatic Brain Injury (MMA-TBI). MOS was collected through record review and categorized into low and high risk for blast exposure. Chi-square analyses and t-tests compared SBI metrics between MOS categories. ROC analyses evaluated diagnostic accuracy of MOS category in determining blast exposure severity. Veterans in high-risk MOS were more likely to have experienced blast and deployment TBI (p's<.001) compared to low-risk MOS. ROC analyses indicated good specificity (81.29-88.00) for blast and TBI outcomes, suggesting that low-risk MOS is generally associated with an absence of blast and TBI outcomes. Sensitivity was low (36.46-51.14), indicating that MOS risk level was not a good predictor of the presence of these outcomes. Results demonstrate that high-risk MOSs will identify individuals with blast exposure and TBI history whereas low-risk MOSs will capture a highly variable group. Accuracy of MOS categorization was not acceptable for diagnostic-level tests; however, results support its use as a screening measure for a history of exposure to blast, use in epidemiological studies, and considerations for military policy.

A Computational Physics-based Approach to Predict Unbound Brain-to-Plasma Partition Coefficient, Kp,uu.

The blood-brain barrier (BBB) plays a critical role in preventing harmful endogenous and exogenous substances from penetrating the brain. Optimal brain penetration of small-molecule central nervous system (CNS) drugs is characterized by a high unbound brain/plasma ratio (Kp,uu). While various medicinal chemistry strategies and in silico models have been reported to improve BBB penetration, they have limited application in predicting Kp,uu directly. We describe a physics-based computational approach, a quantum mechanics (QM)-based energy of solvation (E-sol), to predict Kp,uu. Prospective application of this method in internal CNS drug discovery programs highlights the utility and accuracy of this new method, which showed a categorical accuracy of 79% and an R2 of 0.61 from a linear regression model.

Intelligent classification of coal seams using spontaneous combustion susceptibility in IoT paradigm

ABSTRACT Coal is the cheapest source of energy. It is much cheaper than nuclear and petroleum. However, spontaneous combustion is inherent in coal seams that lead to fires at their place of storage. Therefore, early detection is required to ensure safety. This article aims to develop an intelligent system for the classification of coal seams based on their spontaneous heating on the Internet of Things (IoT) paradigm. It utilizes an artificial intelligence (AI) approach by utilizing a two-stage neural network architecture to render an accurate and robust categorization of coal samples. Three commonly available proximate analysis parameters of the coal samples have been used as input to this system. These inherent parameters are found to be sufficient to classify coal samples into different categories according to their combustion susceptibility. Three publicly available datasets are involved in the assessment purpose. Our proposed method has outperformed previous approaches. The coal seams classification results agree with the real field experiences.

Tinker Hunt

With tech taking over most of the life process in recent times, primary education still lacks interactive processes using technology in the process of “Learning by Doing”. Most of the Ed. techs offer visual learning with no or near-to-zero input by children. Tinker Hunt can be a possible and one of its kind of interactive learning application used to tackle the same. A rough sketching input by children can yield results over the internet in the form of pictures, detailed information, etc. Tinker Hunt is created using Google’s QuickDraw Dataset, using hundreds of sketch patterns for multiple categories of icon-level doodles, collected from hundreds of individuals that are recognized by CNN trained model. Also, the app is tested for high accuracy in sketch recognition and categorization. Key Word: Machine Learning; QuickDraw Dataset; Interactive Learning; Convolutional Neural Network; Artificial Neural Network

Partial urogenital mobilization in cloacal malformation: is it a viable option?

Total Urogenital Mobilization (TUM) has been the standard surgical approach for the urogenital complex in Cloacal Malformations (CM) since its inception in 1997. Partial Urogenital Mobilization (PUM) in CM remains an under-utilized or under-reported option. The main anatomical difference between TUM and PUM is the division of the pubo-urethral ligaments. We explored the feasibility of PUM in a select subset of our patients with CM and report early outcomes. We retrospectively reviewed prospectively collected data of all our CM patients who had primary reconstruction at our centre from 2012 to 2020. We included in our review the patients who underwent PUM. Mullerian abnormalities, spinal cord involvement, common channel length (CC), urethral length (UL), surgical reconstruction, and outcomes including urinary continence, recurrent UTI, ultrasound and preoperative DMSA/MAG3, cystovaginoscopy post-reconstruction, and post-void residuals were noted. Fifty-three patients had primary reconstruction, and of these, eleven had a common channel less than 3cm. Of the eleven, only one underwent TUM. In the PUM group, two underwent filum untethering (20%). Mullerian duplication was noted in 5 patients (50%). The median CC length=1.6cm (range=1.5cm-2.7cm), and median UL=1.5cm (range=1.5cm-2.5cm). Follow-up ranged from 9 to 134months (median=63months). Post-reconstruction all had a separate urethral and vaginal opening on examination and cysto-vaginoscopy. The continence outcomes are summarized in Fig.1. Although TUM is the most common solution for the urogenital complex in CM, a subset would be suitable for PUM, and this option is under-utilized or under-reported in literature. We presume that many who had TUM probably only needed a PUM, and therefore could report better outcomes from a bladder function aspect. It is important to differentiate the two, and outcomes should be appropriately categorized. Our default approach is a PUM in all CM with less than 3cm common channel. Only the lateral and posterior aspects of the urogenital complex are mobilized and if the urethra did not reach a satisfactory level for easy intermittent catheterization, then we proceed to a TUM dividing the pubo-urethral ligaments. PUM avoids the potential complications related to dividing the pubo-urethral ligament in TUM. It may also avoid the need for CIC which is encountered in patients who undergo TUM. PUM is a viable alternative in cloacal malformations with good outcomes in those with a common channel under 3cm. This of course requires appropriate patient selection and accurate categorization of interventions to understand the true outcomes.

PO-04-188 PREMATURE VENTRICULAR COMPLEXES: ASSESSING BURDEN DENSITY IN A LARGE NATIONAL COHORT POPULATION TO BETTER DEFINE OPTIMAL ECG MONITORING DURATION

PVC burden frequency has been shown to correlate with the incidence of PVC-mediated cardiomyopathy. A single center study of 107 patients wearing Zio monitors indicated that measurement of PVC burden for 24-48 hours may not reflect the total burden. This study aims to assess PVC burden density, and the significance of long-term monitoring in a large national cohort. The objective of this study is to establish the optimal duration of ECG monitoring required to accurately characterize PVC burden severity and other significant factors. Patients with a PVC burden ≥5%, age ≥18 and a ≥13-day wear period were analyzed between June 2011 to April 2022. Patient demographics were reported, and the mean absolute error in PVC burden was calculated at baseline (<10 sec), 24 hours, 48 hours, and daily to14 days. Burden at the end of the wear period was considered the true value for PVC burden. The measured error was described for each subinterval (mean, SD, median, IQR). The accuracy in PVC burden was evaluated across monitoring durations at 10% (low),15% (moderate) and 20% (high) thresholds. There were 106,705 patients analyzed, 33.6% of patients were females with an average age of 67.3±15.4 years old. PVC density was more frequent in males than females, 66:34 respectively and in those older than 65 years of age. The PVC burden mean absolute error was calculated at baseline (<10sec), 24 hours, 48 hours, 7 days, and 13 days (Figure1). When comparing PVC burden densities with lower and higher than 20% at 24 hours and 7 days, data showed a decreased in mean absolute error from 2.78% and 3.97% at 24 hours to 1.27% and 1.75% at 7 days. Furthermore, the accuracy in categorization of PVC burden at 15% threshold increased from 89.2% at 24 hours to 95.1% at 7 days. Most common concerning arrhythmias detected in patients with PVC burden do show important delineating trend to consider (Table1). In this large national cohort, the reduction in PVC burden mean absolute error from 24 hours through 14 days underscores the importance of long-term ambulatory monitoring and extends prior study results. Increased monitoring duration through at least 7-10 days is associated with improved accuracy in categorization of patients by PVC burden level. Patients with PVC burden <20% may benefit from longer monitoring compared to >20% burden population. Concomitant arrythmias including VT and Afib in patient with PVC burden further emphasized the significance of closer monitoring and early detection.Tabled 1Table1Total Device Count (%)Low (5%-10%)* n=59,702Moderate(10%-20%)* n=34,605High (>20%)* n=12,398Overall* n=106,705VT37,858 (63.4)23,799 (68.8)9,251 (74.6)70,908 (66.5)AFIB9,021 (15.1)4,123 (11.9)1,119 (9.0)14,263 (13.4)SVT46,068 (77.2)26,712 (77.2)9,495 (76.6)82,275 (77.1)AVB1,467 (2.5)671 (1.9)181 (1.5)2,319 (2.2)Pause (≥3 sec)3,479 (5.8)1,682 (4.9)466 (3.8)5,627 (5.3) Open table in a new tab

Resolving Cross-modal Semantic Interference among Object Concepts Requires Medial Temporal Lobe Cortex

The ability to flexibly categorize object concepts is essential to semantic cognition because the features that make two objects similar in one context may be irrelevant and even constitute interference in another. Thus, adaptive behavior in complex and dynamic environments requires the resolution of feature-based interference. In the current case study, we placed visual and functional semantic features in opposition across object concepts in two categorization tasks. Successful performance required the resolution of functional interference in a visual categorization task and the resolution of visual interference in a functional categorization task. In Experiment 1, we found that patient D. A., an individual with bilateral temporal lobe lesions, was unable to categorize object concepts in a context-dependent manner. His impairment was characterized by an increased tendency to incorrectly group objects that were similar on the task-irrelevant dimension, revealing an inability to resolve cross-modal semantic interference. In Experiment 2, D. A.'s categorization accuracy was comparable to controls when lures were removed, indicating that his impairment is unique to contexts that involve cross-modal interference. In Experiment 3, he again performed as well as controls when categorizing simple concepts, suggesting that his impairment is specific to categorization of complex object concepts. These results advance our understanding of the anterior temporal lobe as a system that represents object concepts in a manner that enables flexible semantic cognition. Specifically, they reveal a dissociation between semantic representations that contribute to the resolution of cross-modal interference and those that contribute to the resolution of interference within a given modality.

New bidirectional recurrent neural network optimized by improved Ebola search optimization algorithm for lung cancer diagnosis

The early detection of cancerous and malignant lung cancer by medical imaging techniques, CT-scan for example, which never needs to do sampling reduces the risk of cancer growth and spreading. Accordingly, computer image processing and diagnostic system development, followed by cancer's classification into malignant and benign, is of primary importance in the early discovery of lung cancer which plays a pivotal role in the treatment improvement and saving the patient's life. This work intended to improve malignant and benign gland categorization accuracy and, as a result, detection accuracy. Here, a new methodology has been proposed to get an accurate lung cancer diagnosis system using an improved Bidirectional Recurrent neural network. The improvement of the network has been done by designing an improved form of an Ebola optimization search algorithm. Before applying the major diagnosis system, some preprocessing techniques have been done. The model is then applied to IQ-OTH/NCCD lung cancer dataset and its results are compared with some published works to indicate the eminence of the suggested method toward the comparative ones.

Neurological disease prediction using impaired gait analysis for foot position in cerebellar ataxia by ensemble approach

In neurological field, Cerebellar Ataxia (CA) prediction is done with Gait values of human actions. The Analysis of Gait (AoG) may lead the good treatment. The goal of this work was to develop a machine-learning-based model for predicting AoG using the poor gait patterns that occur before AoG. While executing designed AoG-provoking walking tasks, an accelerometer was connected to the lower back of 21 subjects with 12 different walking positions to gather acceleration impulses. The exercise was walking for one minute at each of 12 varied walking speeds on a split-belt treadmill in the range [0.6, 1.7] m/s in 0.1 m/s increments. To reduce the effects of weariness, the speed sequence was randomized and kept a secret from the subjects. Machine-learning algorithms like support vector machine (SVM) and k-nearest neighbours (KNN) have been tested in existing research studies. These algorithms perform well when the amount of data is little and the classification is binary. SVM, KNN, decision trees, and XGBoost algorithms have all been used in the proposed study on the CA data set. We discovered that the AdaBoost algorithm provides a more accurate categorization of the severity of CA disease.

Accuracy and Consistency in Social Categorization Across Context, Motivation, and Time

Photos provide a literal snapshot of a person in a particular context at a specific moment in time. Previous studies have found that people can accurately categorize others from single photos of their faces along various social dimensions, yet this research typically assumes that one photo of an individual representatively samples other photos of the same individual. Across four studies, we investigated this assumption by testing the consistency of perceptions of social categories (viz. sexual orientation and political affiliation) based on multiple photos of the same individuals. We found that judgments of social categories exceeded chance and significantly correlated across different photo contexts, across variability in targets’ motivations, and across time. These data supplement earlier work showing similar consistency for other types of social judgments. Thus, single face photos can consistently convey some aspects of an individual's appearance.

Classification of Blazar Candidates of Unknown Type in Fermi 4LAC by Unanimous Voting from Multiple Machine-learning Algorithms

The Fermi fourth catalog of active galactic nuclei (AGNs) data release 3 (4LAC-DR3) contains 3407 AGNs, out of which 755 are flat spectrum radio quasars (FSRQs), 1379 are BL Lacertae objects (BL Lac objects), 1208 are blazars of unknown (BCUs) type, while 65 are non-AGNs. Accurate categorization of many unassociated blazars still remains a challenge due to the lack of sufficient optical spectral information. The aim of this work is to use high-precision, optimized machine-learning (ML) algorithms to classify BCUs into BL Lac objects and FSRQs. To address this, we selected the 4LAC-DR3 Clean sample (i.e., sources with no analysis flags) containing 1115 BCUs. We employ five different supervised ML algorithms, namely, random forest, logistic regression, XGBoost, CatBoost, and neural network with seven features: photon index, synchrotron-peak frequency, pivot energy, photon index at pivot energy, fractional variability, ν F ν, at synchrotron-peak frequency, and variability index. Combining results from all models leads to better accuracy and more robust predictions. These five methods together classified 610 BCUs as BL Lac objects and 333 BCUs as FSRQs with a classification metric area under the curve >0.96. Our results are significantly compatible with recent studies as well. The output from this study provides a larger blazar sample with many new targets that could be used for forthcoming multiwavelength surveys. This work can be further extended by adding features in X-rays, UV, visible, and radio wavelengths.

Multi-Scale Region-Of-Interest based Deep Learning for fruit disease identification and classification

Agricultural productivity is mainly affected by the different types of fruit diseases. To reduce such impacts, automated fruit disease identification and categorization model is vital. So, a Multi-scale Deep Learning (MDL) model was designed to identify and categorize the variety of strawberry fruit diseases. This model has an Adaptive Receptive Field Module (ARFM), a Bottleneck Block (BB), Depth-wise Residual Blocks (DRBs) and the Squeeze Excitation Residual Network (SE-ResNet) units to extract the Feature Maps (FMs) related to all disease classes. But, this model cannot partition individual fruit samples. Also, the number of fruits with multiple infections within a particular image was not easily captured. So, this article develops a Multi-scale Region-Of-Interest (ROI)-based DL model called (MRDL) to improve the accuracy of identification and categorization of a single fruit sample with a certain disease in a real-time scenario. This model is performed in 2 different phases. First, more robust features are extracted by the DRB, ARFM, and BB in the initial phase. Then, an ROI Fused (ROIF) unit, the SE-ResNet unit, a classification unit, and a mask unit are involved in the second phase. The ROIF unit obtains the FMs for all candidate ROIs using the candidate bounding boxes from the initial phase. The extracted FMs per sampled position are aggregated to get the fused FM, which is resized by integrating the attention strategy with the SE-ResNet unit. Those resized fused FMs are independently passed to the classification and mask units to classify the candidate bounding boxes and localize the infected pixels in the candidate ROIs, respectively. Finally, the test results exhibit that the MRDL model achieves 91.1%, 90.8%, and 91.6% accuracy on apple, citrus, and tomato disease databases, respectively compared to the classical models.

An improved approach to Arabic news classification based on hyperparameter tuning of machine learning algorithms

This modern generation uses machine learning algorithms to tackle many issues, one of which is text classification. To develop classifiers that can predict the category of texts, and to optimize the output of these classifiers, hyperparameter tuning is required. The automatic optimization of the parameters of a machine learning model is referred to as hyperparameter tuning. Using NLP techniques and machine learning algorithms, we offer many methods in this research to improve the suggested classifier's accuracy values and identify the best hyperparameter. Grid search and random search are two methods for tweaking the hyperparameter. In any case, a comparison with other research works was made in order to assess the effectiveness of the suggested model and compare it against other models. The proposed approach appears to provide a robust solution for accurate Arabic text representation, interpretation, and categorization. It achieves the best performance using the CNN Arabic dataset in terms of overall accuracy, recall, precision, and F1-score by 95.16 %, 94.64 %,94.04 %, and 94.31 %.

Identification of Ancient Glass Components Based on Grey Correlation Analysis

The article is based on mathematical modelling of the significant data of the two categorize glass specimen to examine the factors influencing the climate of the covering of the glass artefacts, to derive statistical patterns and correlations of the chemical composition content, on this basis, the subclasses are classified, and the accurate categorization manner applied to the analysis and discovery of glass composition is given. On the basis of data processing, the association between weathering of glass surface and glass kind, adornment and color is analyzed by using visual histogram. And it was found that the weathering of lead-barium glass was more serious than that of high-potassium glass, that B-grain decoration was easily weathered, and that black and light blue glass were the most severely weathered. The most significant effect on weathering is due to the content of silica in the lead-barium group, and the most significant effect on weathering is due to the content of sodium oxide in the high-potassium group. The correlations between the 14 chemical components in the two categories of high potassium glass and lead-barium glass were analysed by establishing a grey correlation analysis model, and the correlations between the different chemical components in the different categories were calculated and ranked. Similarly, the correlation between lead oxide and magnesium oxide was the highest in the high potassium group, while strontium oxide, tin oxide and sulphur dioxide were all less correlated, i.e. the correlation between this component and other chemical components was characterised by the ANOVA.

Ultrafast Image Categorization in Biology and Neural Models.

Humans are able to categorize images very efficiently, in particular to detect the presence of an animal very quickly. Recently, deep learning algorithms based on convolutional neural networks (CNNs) have achieved higher than human accuracy for a wide range of visual categorization tasks. However, the tasks on which these artificial networks are typically trained and evaluated tend to be highly specialized and do not generalize well, e.g., accuracy drops after image rotation. In this respect, biological visual systems are more flexible and efficient than artificial systems for more general tasks, such as recognizing an animal. To further the comparison between biological and artificial neural networks, we re-trained the standard VGG 16 CNN on two independent tasks that are ecologically relevant to humans: detecting the presence of an animal or an artifact. We show that re-training the network achieves a human-like level of performance, comparable to that reported in psychophysical tasks. In addition, we show that the categorization is better when the outputs of the models are combined. Indeed, animals (e.g., lions) tend to be less present in photographs that contain artifacts (e.g., buildings). Furthermore, these re-trained models were able to reproduce some unexpected behavioral observations from human psychophysics, such as robustness to rotation (e.g., an upside-down or tilted image) or to a grayscale transformation. Finally, we quantified the number of CNN layers required to achieve such performance and showed that good accuracy for ultrafast image categorization can be achieved with only a few layers, challenging the belief that image recognition requires deep sequential analysis of visual objects. We hope to extend this framework to biomimetic deep neural architectures designed for ecological tasks, but also to guide future model-based psychophysical experiments that would deepen our understanding of biological vision.

Introducing a Conceptual Model for Assessing the Present State of Preservation in Heritage Buildings: Utilizing Building Adaptation as an Approach

Building adaptation comprises a variety of construction actions that enhance current condition and extend the life span of buildings. Architectural adaptation involves refurbishing, retrofitting, restoration, renovation, rehabilitation, adaptive reuse, material reuse, conservation, and preservation, as well as other activities to improve building conditions. In this study, the most relevant definitions, implementations, and ranges of the specified terminologies are examined; on the basis of this classification, a conceptual model is constructed to facilitate accurate categorization of building adaptations and its application in various case studies. The current state of heritage buildings in Erbil (the Erbil Citadel as a World Heritage Site and the buffer zone as cultural heritage) is evaluated. We aimed to investigate their current conservation status and determine whether the Erbil Citadel’s heritage building adaptation has addressed the issue of energy retrofitting to realize green and passive construction. The results indicate that the framework model is an effective assessment tool in the field of conserving heritage buildings. The majority of Erbil’s heritage buildings have been restored and rehabilitated for public interaction; nevertheless, additional interventions are required to improve the buildings’ energy efficiency and conservation for this to be recognized as sustainable heritage development. In the future, academics and practitioners may use the suggested framework to clearly and consistently describe the scope of the work in their building adaptation, thereby eliminating the high costs of correcting code and specification requirements that were not met.

Waste classification using vision transformer based on multilayer hybrid convolution neural network

The rapid advancement of deep learning technology has led to the presentation of various network architectures for classification, making it easier to implement intelligent waste classification systems. However, existing waste classification models have problems such as low accuracy and slow processing. The current system does not utilize automatic classification. The proposed method uses Vision Transformer based on Multilayer Hybrid Convolution Neural Network for automatic waste classification (VT-MLH-CNN). The proposed method enhances the accuracy of waste classification and reduces the time taken for classification. Initially, it collects the data images, then the features are extracted, and next, it is processed into data normalization. The proposed model performs better by altering the number of network modules and connections. After this study determines the proper waste picture categorization variables, the best strategy is selected as the final model. The simulation results indicated that the suggested approach has a simplified network model and greater waste categorization accuracy compared to certain current efforts. Numerous tests on the TrashNet dataset demonstrate the usefulness of the recommended method, which achieves classification accuracy of up to 95.8%, which is 5.28% and 4.6% greater than those state-of-the-art techniques.

Detection and Classification of Dog Skin Disease using Deep Learning

Dogs are beloved pets and loyal companions to millions of people worldwide. Unfortunately, dogs can also suffer from a variety of skin diseases that can cause discomfort, pain, and even life-threatening complications. Some dog skin diseases can be transmitted to humans through direct contact. Early detection and treatment of these skin diseases are crucial for the health and well-being of dogs and humans. Our project aims to provide a quick and precise approach to identifying various types of skin diseases in dogs. To expedite the process of identifying and diagnosing infections related to canine interaction, we plan to utilize a machine-learning model. This approach aims to reduce the time and expertise required for accurate and consistent diagnosis, which can otherwise be challenging and time-consuming. Two models, InceptionV3 and MobileNetV2, were utilized and compared in our implementation. In the case of InceptionV3, a training accuracy of 0.99 and a validation accuracy of 0.98 were achieved. For MobileNetV2, we attained a validation accuracy of 96 and a categorical accuracy of 97. Key Words: Dermatophytosis, zoonosis, image classification, deep learning, Transfer learning, InceptionV3, MobileNetV2, CNN, DNN.

An Automated Method for Classifying Liver Lesions in Contrast-Enhanced Ultrasound Imaging Based on Deep Learning Algorithms.

Contrast-enhanced ultrasound (CEUS) is an important imaging modality in the diagnosis of liver tumors. By using contrast agent, a more detailed image is obtained. Time-intensity curves (TIC) can be extracted using a specialized software, and then the signal can be analyzed for further investigations. The purpose of the study was to build an automated method for extracting TICs and classifying liver lesions in CEUS liver investigations. The cohort contained 50 anonymized video investigations from 49 patients. Besides the CEUS investigations, clinical data from the patients were provided. A method comprising three modules was proposed. The first module, a lesion segmentation deep learning (DL) model, handled the prediction of masks frame-by-frame (region of interest). The second module performed dilation on the mask, and after applying colormap to the image, it extracted the TIC and the parameters from the TIC (area under the curve, time to peak, mean transit time, and maximum intensity). The third module, a feed-forward neural network, predicted the final diagnosis. It was trained on the TIC parameters extracted by the second model, together with other data: gender, age, hepatitis history, and cirrhosis history. For the feed-forward classifier, five classes were chosen: hepatocarcinoma, metastasis, other malignant lesions, hemangioma, and other benign lesions. Being a multiclass classifier, appropriate performance metrics were observed: categorical accuracy, F1 micro, F1 macro, and Matthews correlation coefficient. The results showed that due to class imbalance, in some cases, the classifier was not able to predict with high accuracy a specific lesion from the minority classes. However, on the majority classes, the classifier can predict the lesion type with high accuracy. The main goal of the study was to develop an automated method of classifying liver lesions in CEUS video investigations. Being modular, the system can be a useful tool for gastroenterologists or medical students: either as a second opinion system or a tool to automatically extract TICs.

Voice matters: Social categorization and stereotyping of speakers based on sexual orientation and nationality categories

This research examined how listeners categorize and stereotype speakers belonging to intersecting social categories (nationality; sexual orientation) based on voice alone. In Study 1, British heterosexuals categorized the nationality and sexual orientation of British and Italian speakers who self-identified as gay or heterosexual. Participants correctly categorized British speakers as co-nationals and Italian speakers as foreigners. Categorization accuracy of gay speakers’ sexual orientation was poor. Italian gay speakers were perceived as most likely to be gay and non-native speakers. Study 2 examined stereotyping of speakers who sounded either native or foreign, and sounded either gay or heterosexual. Foreign-accented (vs. native-accented) speakers were rated as less competent, and gay-sounding (vs. heterosexual-sounding) speakers as less gender typical. Foreign-accented gay speakers were perceived as the least competent and gender typical.