Algorithmic Classification of Five Characteristic Types of Paraphasias

2016 ◽  
Vol 25 (4S) ◽  
Author(s):  
Gerasimos Fergadiotis ◽  
Kyle Gorman ◽  
Steven Bedrick
Keyword(s):  
Language Processing ◽  
Classification Accuracy ◽  
Measurement Data ◽  
Quality Measurement ◽  
Similarity Criterion ◽  
Cost Effective ◽  
Phonological Similarity ◽  
Diagnostic Tools ◽  
Real Word

Purpose This study was intended to evaluate a series of algorithms developed to perform automatic classification of paraphasic errors (formal, semantic, mixed, neologistic, and unrelated errors). Method We analyzed 7,111 paraphasias from the Moss Aphasia Psycholinguistics Project Database (Mirman et al., 2010) and evaluated the classification accuracy of 3 automated tools. First, we used frequency norms from the SUBTLEXus database (Brysbaert & New, 2009) to differentiate nonword errors and real-word productions. Then we implemented a phonological-similarity algorithm to identify phonologically related real-word errors. Last, we assessed the performance of a semantic-similarity criterion that was based on word2vec (Mikolov, Yih, & Zweig, 2013). Results Overall, the algorithmic classification replicated human scoring for the major categories of paraphasias studied with high accuracy. The tool that was based on the SUBTLEXus frequency norms was more than 97% accurate in making lexicality judgments. The phonological-similarity criterion was approximately 91% accurate, and the overall classification accuracy of the semantic classifier ranged from 86% to 90%. Conclusion Overall, the results highlight the potential of tools from the field of natural language processing for the development of highly reliable, cost-effective diagnostic tools suitable for collecting high-quality measurement data for research and clinical purposes.

MULTI-TONE WAVEFORM MEASUREMENT SYSTEM ENABLING CHARACTERIZATION OF MICROWAVE DEVICES

2014 ◽  
Vol 23 (10) ◽  
pp. 1450141
Author(s):  
MUHAMMAD AKMAL CHAUDHARY ◽  
JONATHAN LEES ◽  
JOHANNES BENEDIKT ◽  
PAUL TASKER
Keyword(s):  
Measurement System ◽  
Measurement Data ◽  
Quality Measurement ◽  
Cost Effective ◽  
Operating Conditions ◽  
Efficient Manner ◽  
Vector Error Correction ◽  
Sampling Oscilloscope ◽  
Rf Components ◽  
Vector Error

This paper presents a fully automated time domain, waveform measurement system, capable of measuring multi-tone waveforms up to a frequency of 14 GHz. Multi-tone waveform measurement capabilities will prove useful in enhancing the understanding of the response of devices under realistic operating conditions, and allow for detailed investigation into device problems leading to memory effects. The system, which is based around a standard sampling oscilloscope, is capable of measuring all four traveling waves simultaneously. It is a cost effective solution, capable of capturing high quality measurement data, it consists of two test sets one to measure RF components of the signal and one to measure IF components, which are then recombined before being measured by the sampling oscilloscope. Vector error correction is applied to the measured data to fully calibrate the system to the device plane, ensuring any dispersion in the connecting hardware is removed. A multi-tone waveform sampling method is employed, ensuring the waveforms are captured in the most efficient manner. Device results are presented showing the multi-tone voltage and current waveforms at the device plane. Some useful applications of the system are demonstrated and explained.

scholarly journals EEG-based classification of natural sounds reveals specialized responses to speech and music

2019 ◽  
Author(s):  
Nathaniel J Zuk ◽  
Emily S Teoh ◽  
Edmund C Lalor
Keyword(s):  
Classification Accuracy ◽  
Average Power ◽  
Cost Effective ◽  
List Type ◽  
Natural Sounds ◽  
Brain Responses ◽  
Species Specific ◽  
The Brain ◽  
Better Than

AbstractHumans can easily distinguish many sounds in the environment, but speech and music are uniquely important. Previous studies, mostly using fMRI, have identified separate regions of the brain that respond selectively for speech and music. Yet there is little evidence that brain responses are larger and more temporally precise for human-specific sounds like speech and music, as has been found for responses to species-specific sounds in other animals. We recorded EEG as healthy, adult subjects listened to various types of two-second-long natural sounds. By classifying each sound based on the EEG response, we found that speech, music, and impact sounds were classified better than other natural sounds. But unlike impact sounds, the classification accuracy for speech and music dropped for synthesized sounds that have identical “low-level” acoustic statistics based on a subcortical model, indicating a selectivity for higher-order features in these sounds. Lastly, the trends in average power and phase consistency of the two-second EEG responses to each sound replicated the patterns of speech and music selectivity observed with classification accuracy. Together with the classification results, this suggests that the brain produces temporally individualized responses to speech and music sounds that are stronger than the responses to other natural sounds. In addition to highlighting the importance of speech and music for the human brain, the techniques used here could be a cost-effective and efficient way to study the human brain’s selectivity for speech and music in other populations.HighlightsEEG responses are stronger to speech and music than to other natural soundsThis selectivity was not replicated using stimuli with the same acoustic statisticsThese techniques can be a cost-effective way to study speech and music selectivity

scholarly journals Artificial Intelligence in Fractured Dental Implant Detection and Classification: Evaluation Using Dataset from Two Dental Hospitals

Diagnostics ◽  
2021 ◽  
Vol 11 (2) ◽  
pp. 233
Author(s):  
Dong-Woon Lee ◽  
Sung-Yong Kim ◽  
Seong-Nyum Jeong ◽  
Jae-Hong Lee
Keyword(s):  
Dental Implant ◽  
Classification Accuracy ◽  
Reliability And Validity ◽  
Mechanical Complication ◽  
Radiographic Images ◽  
Acceptable Accuracy ◽  
Reliable Detection ◽  
Classification Evaluation ◽  
Accuracy Performance

Fracture of a dental implant (DI) is a rare mechanical complication that is a critical cause of DI failure and explantation. The purpose of this study was to evaluate the reliability and validity of a three different deep convolutional neural network (DCNN) architectures (VGGNet-19, GoogLeNet Inception-v3, and automated DCNN) for the detection and classification of fractured DI using panoramic and periapical radiographic images. A total of 21,398 DIs were reviewed at two dental hospitals, and 251 intact and 194 fractured DI radiographic images were identified and included as the dataset in this study. All three DCNN architectures achieved a fractured DI detection and classification accuracy of over 0.80 AUC. In particular, automated DCNN architecture using periapical images showed the highest and most reliable detection (AUC = 0.984, 95% CI = 0.900–1.000) and classification (AUC = 0.869, 95% CI = 0.778–0.929) accuracy performance compared to fine-tuned and pre-trained VGGNet-19 and GoogLeNet Inception-v3 architectures. The three DCNN architectures showed acceptable accuracy in the detection and classification of fractured DIs, with the best accuracy performance achieved by the automated DCNN architecture using only periapical images.

scholarly journals Measles Virus IgG Avidity Assay for Use in Classification of Measles Vaccine Failure in Measles Elimination Settings

2012 ◽  
Vol 19 (11) ◽  
pp. 1810-1817 ◽  
Author(s):  
Sara Mercader ◽  
Philip Garcia ◽  
William J. Bellini
Keyword(s):  
Measles Virus ◽  
Area Under The Curve ◽  
Epidemiological Data ◽  
Diagnostic Tools ◽  
High Avidity ◽  
Characteristic Analysis ◽  
Vaccine Failure ◽  
Avidity Assay ◽  
Igg Avidity

ABSTRACTIn regions where endemic measles virus has been eliminated, diagnostic assays are needed to assist in correctly classifying measles cases irrespective of vaccination status. A measles IgG avidity assay was configured using a commercially available measles-specific IgG enzyme immunoassay by modifying the protocol to include three 5-min washes with diethylamine (60 mM; pH 10.25) following serum incubation; serum was serially diluted, and the results were expressed as the end titer avidity index. Receiver operating characteristic analysis was used for evaluation and validation and to establish low (≤30%) and high (≥70%) end titer avidity thresholds. Analysis of 319 serum specimens expected to contain either high- or low-avidity antibodies according to clinical and epidemiological data indicated that the assay is highly accurate, with an area under the curve of 0.998 (95% confidence interval [CI], 0.978 to 1.000), sensitivity of 91.9% (95% CI, 83.2% to 97.0%), and specificity of 98.4% (95% CI, 91.6% to 100%). The assay is rapid (<2 h) and precise (standard deviation [SD], 4% to 7%). In 18 samples from an elimination setting outbreak, the assay identified 2 acute measles cases with low-avidity results; both were IgM-positive samples. Additionally, 11 patients (15 samples) with modified measles who were found to have high-avidity IgG results were classified as secondary vaccine failures; one sample with an intermediate-avidity result was not interpretable. In elimination settings, measles IgG avidity assays can complement existing diagnostic tools in confirming unvaccinated acute cases and, in conjunction with adequate clinical and epidemiologic investigation, aid in the classification of vaccine failure cases.

scholarly journals Morphological Characteristics, Classifications and Difficulties in the Use of Diagnostic Criteria for Serrated Lesions of the Large Intestine

2021 ◽  
Author(s):  
Cesar de Souza Bastos Junior ◽  
Vera Lucia Nunes Pannain ◽  
Adriana Caroli-Bottino
Keyword(s):  
Diagnostic Criteria ◽  
Morphological Characteristics ◽  
World Health ◽  
Diagnostic Tools ◽  
Serrated Adenoma ◽  
The Past ◽  
The World ◽  
Serrated Lesions ◽  
Health Organization

Abstract Introduction Colorectal carcinoma (CRC) is the most common gastrointestinal neoplasm in the world, accounting for 15% of cancer-related deaths. This condition is related to different molecular pathways, among them the recently described serrated pathway, whose characteristic entities, serrated lesions, have undergone important changes in their names and diagnostic criteria in the past thirty years. The multiplicity of denominations and criteria over the last years may be responsible for the low interobserver concordance (IOC) described in the literature. Objectives The present study aims to describe the evolution in classification of serrated lesions, based on the last three publications of the World Health Organization (WHO) and the reproducibility of these criteria by pathologists, based on the evaluation of the IOC. Methods A search was conducted in the PubMed, ResearchGate and Portal Capes databases, with the following terms: sessile serrated lesion; serrated lesions; serrated adenoma; interobserver concordance; and reproducibility. Articles published since 1990 were researched. Results and Discussion The classification of serrated lesions in the past thirty years showed different denominations and diagnostic criteria. The reproducibility and IOC of these criteria in the literature, based on the kappa coefficient, varied in most studies, from very poor to moderate. Conclusions Interobserver concordance and the reproducibility of microscopic criteria may represent a limitation for the diagnosis and appropriate management of these lesions. It is necessary to investigate diagnostic tools to improve the performance of the pathologist's evaluation, for better concordance, and, consequently, adequate diagnosis and treatment.

scholarly journals Quantitative Ship Collision Frequency Estimation Models: A Review

2021 ◽  
Vol 9 (5) ◽  
pp. 533
Author(s):  
Mirko Čorić ◽  
Sadko Mandžuka ◽  
Anita Gudelj ◽  
Zvonimir Lušić
Keyword(s):  
Risk Assessment ◽  
Collision Frequency ◽  
Frequency Estimation ◽  
Cost Effective ◽  
Future Improvement ◽  
Ship Collision ◽  
Estimation Models

Ship collisions are one of the most common types of maritime accidents. Assessing the frequency and probability of ship collisions is of great importance as it provides a cost-effective and practical way to mitigate risk. In this paper, we present a review of quantitative ship collision frequency estimation models for waterway risk assessment, accompanied by a classification of the models and a description of their main modelling characteristics. Models addressing the macroscopic perspective in the estimation of ship collision frequency on waterways are reviewed in this paper with a total of 29 models. We extend the existing classification methodology and group the collected models accordingly. Special attention is given to the criteria used to detect potential ship collision candidates, as well as to causation probability and the correlation of models with real ship collision statistics. Limitations of the existing models and future improvement possibilities are discussed. The paper can be used as a guide to understanding current achievements in this field.

scholarly journals An Imbalanced Image Classification Method for the Cell Cycle Phase

Information ◽  
2021 ◽  
Vol 12 (6) ◽  
pp. 249
Author(s):  
Xin Jin ◽  
Yuanwen Zou ◽  
Zhongbing Huang
Keyword(s):  
Cell Cycle ◽  
Deep Learning ◽  
Image Classification ◽  
Classification Accuracy ◽  
Data Augmentation ◽  
Cycle Phase ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Cellular Life

The cell cycle is an important process in cellular life. In recent years, some image processing methods have been developed to determine the cell cycle stages of individual cells. However, in most of these methods, cells have to be segmented, and their features need to be extracted. During feature extraction, some important information may be lost, resulting in lower classification accuracy. Thus, we used a deep learning method to retain all cell features. In order to solve the problems surrounding insufficient numbers of original images and the imbalanced distribution of original images, we used the Wasserstein generative adversarial network-gradient penalty (WGAN-GP) for data augmentation. At the same time, a residual network (ResNet) was used for image classification. ResNet is one of the most used deep learning classification networks. The classification accuracy of cell cycle images was achieved more effectively with our method, reaching 83.88%. Compared with an accuracy of 79.40% in previous experiments, our accuracy increased by 4.48%. Another dataset was used to verify the effect of our model and, compared with the accuracy from previous results, our accuracy increased by 12.52%. The results showed that our new cell cycle image classification system based on WGAN-GP and ResNet is useful for the classification of imbalanced images. Moreover, our method could potentially solve the low classification accuracy in biomedical images caused by insufficient numbers of original images and the imbalanced distribution of original images.

scholarly journals Inhibitory and facilitatory effects of phonological and orthographic similarity on L2 word recognition across modalities in bilinguals

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Candice Frances ◽  
Eugenia Navarra-Barindelli ◽  
Clara D. Martin
Keyword(s):  
Lexical Decision ◽  
Language Processing ◽  
Theoretical Implication ◽  
Phonological Similarity ◽  
Target Language ◽  
Visual Modality ◽  
Decision Task ◽  
Auditory Modality ◽  
Orthographic Similarity ◽  
Special Category

AbstractLanguage perception studies on bilinguals often show that words that share form and meaning across languages (cognates) are easier to process than words that share only meaning. This facilitatory phenomenon is known as the cognate effect. Most previous studies have shown this effect visually, whereas the auditory modality as well as the interplay between type of similarity and modality remain largely unexplored. In this study, highly proficient late Spanish–English bilinguals carried out a lexical decision task in their second language, both visually and auditorily. Words had high or low phonological and orthographic similarity, fully crossed. We also included orthographically identical words (perfect cognates). Our results suggest that similarity in the same modality (i.e., orthographic similarity in the visual modality and phonological similarity in the auditory modality) leads to improved signal detection, whereas similarity across modalities hinders it. We provide support for the idea that perfect cognates are a special category within cognates. Results suggest a need for a conceptual and practical separation between types of similarity in cognate studies. The theoretical implication is that the representations of items are active in both modalities of the non-target language during language processing, which needs to be incorporated to our current processing models.

scholarly journals Semi-automated regional classification of the style of activity of slow rock-slope deformations using PS InSAR and SqueeSAR velocity data

Landslides ◽  
2021 ◽  
Author(s):  
Chiara Crippa ◽  
Elena Valbuzzi ◽  
Paolo Frattini ◽  
Giovanni B. Crosta ◽  
Margherita C. Spreafico ◽  
...  
Keyword(s):  
Rock Slope ◽  
Progressive Failure ◽  
Principal Component ◽  
Cost Effective ◽  
Displacement Rate ◽  
Machine Learning Classification ◽  
Management Perspective ◽  
Alpine Environments

AbstractLarge slow rock-slope deformations, including deep-seated gravitational slope deformations and large landslides, are widespread in alpine environments. They develop over thousands of years by progressive failure, resulting in slow movements that impact infrastructures and can eventually evolve into catastrophic rockslides. A robust characterization of their style of activity is thus required in a risk management perspective. We combine an original inventory of slow rock-slope deformations with different PS-InSAR and SqueeSAR datasets to develop a novel, semi-automated approach to characterize and classify 208 slow rock-slope deformations in Lombardia (Italian Central Alps) based on their displacement rate, kinematics, heterogeneity and morphometric expression. Through a peak analysis of displacement rate distributions, we characterize the segmentation of mapped landslides and highlight the occurrence of nested sectors with differential activity and displacement rates. Combining 2D decomposition of InSAR velocity vectors and machine learning classification, we develop an automatic approach to characterize the kinematics of each landslide. Then, we sequentially combine principal component and K-medoids cluster analyses to identify groups of slow rock-slope deformations with consistent styles of activity. Our methodology is readily applicable to different landslide datasets and provides an objective and cost-effective support to land planning and the prioritization of local-scale studies aimed at granting safety and infrastructure integrity.

scholarly journals BioVerbNet: a large semantic-syntactic classification of verbs in biomedicine

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Olga Majewska ◽  
Charlotte Collins ◽  
Simon Baker ◽  
Jari Björne ◽  
Susan Windisch Brown ◽  
...  
Keyword(s):  
Natural Language ◽  
Language Processing ◽  
Model Performance ◽  
Representation Learning ◽  
Verb Classes ◽  
Expert Annotation ◽  
Biomedical Texts ◽  
Time Required ◽  
Verb Meaning

Abstract Background Recent advances in representation learning have enabled large strides in natural language understanding; However, verbal reasoning remains a challenge for state-of-the-art systems. External sources of structured, expert-curated verb-related knowledge have been shown to boost model performance in different Natural Language Processing (NLP) tasks where accurate handling of verb meaning and behaviour is critical. The costliness and time required for manual lexicon construction has been a major obstacle to porting the benefits of such resources to NLP in specialised domains, such as biomedicine. To address this issue, we combine a neural classification method with expert annotation to create BioVerbNet. This new resource comprises 693 verbs assigned to 22 top-level and 117 fine-grained semantic-syntactic verb classes. We make this resource available complete with semantic roles and VerbNet-style syntactic frames. Results We demonstrate the utility of the new resource in boosting model performance in document- and sentence-level classification in biomedicine. We apply an established retrofitting method to harness the verb class membership knowledge from BioVerbNet and transform a pretrained word embedding space by pulling together verbs belonging to the same semantic-syntactic class. The BioVerbNet knowledge-aware embeddings surpass the non-specialised baseline by a significant margin on both tasks. Conclusion This work introduces the first large, annotated semantic-syntactic classification of biomedical verbs, providing a detailed account of the annotation process, the key differences in verb behaviour between the general and biomedical domain, and the design choices made to accurately capture the meaning and properties of verbs used in biomedical texts. The demonstrated benefits of leveraging BioVerbNet in text classification suggest the resource could help systems better tackle challenging NLP tasks in biomedicine.

Sign in / Sign up

Export Citation Format

Share Document