A Comprehensive Review of Sound-Based Modalities for Automatic COVID-19 Detection using Deep Learning-Based Techniques

The World Health Organization has labeled the novel coronavirus illness (COVID-19) a pandemic since March 2020. It's a new viral infection with a respiratory tropism that could lead to atypical pneumonia. Thus, according to experts, early detection of the positive cases with people infected by the COVID-19 virus is highly needed. In this manner, patients will be segregated from other individuals, and the infection will not spread. As a result, developing early detection and diagnosis procedures to enable a speedy treatment process and stop the transmission of the virus has become a focus of research. Alternative early-screening approaches have become necessary due to the time-consuming nature of the current testing methodology such as Reverse transcription polymerase chain reaction (RT-PCR) test. The methods for detecting COVID-19 using deep learning (DL) algorithms using sound modality, which have become an active research area in recent years, have been thoroughly reviewed in this work. Although the majority of the newly proposed methods are based on medical images (i.e. X-ray and CT scans), we show in this comprehensive survey that the sound modality can be a good alternative to these methods, providing faster and easiest way to create a database with a high performance. We also present the most popular sound databases proposed for COVID-19 detection. © 2022 IEEE.

Residual Feature Attentional Fusion Network for Lightweight Chest CT Image Super-Resolution

The diagnosis of COVID-19 requires chest computed tomography (CT). High-resolution CT images can provide more diagnostic information to help doctors better diagnose the disease, so it is of clinical importance to study super-resolution (SR) algorithms applied to CT images to improve the reso-lution of CT images. However, most of the existing SR algorithms are studied based on natural images, which are not suitable for medical images;and most of these algorithms improve the reconstruction quality by increasing the network depth, which is not suitable for machines with limited resources. To alleviate these issues, we propose a residual feature attentional fusion network for lightweight chest CT image super-resolution (RFAFN). Specifically, we design a contextual feature extraction block (CFEB) that can extract CT image features more efficiently and accurately than ordinary residual blocks. In addition, we propose a feature-weighted cascading strategy (FWCS) based on attentional feature fusion blocks (AFFB) to utilize the high-frequency detail information extracted by CFEB as much as possible via selectively fusing adjacent level feature information. Finally, we suggest a global hierarchical feature fusion strategy (GHFFS), which can utilize the hierarchical features more effectively than dense concatenation by progressively aggregating the feature information at various levels. Numerous experiments show that our method performs better than most of the state-of-the-art (SOTA) methods on the COVID-19 chest CT dataset. In detail, the peak signal-to-noise ratio (PSNR) is 0.11 dB and 0.47 dB higher on CTtest1 and CTtest2 at x3 SR compared to the suboptimal method, but the number of parameters and multi-adds are reduced by 22K and 0.43G, respectively. Our method can better recover chest CT image quality with fewer computational resources and effectively assist in COVID-19.

Food Processing Waste in Vietnam: Utilization and Prospects in Food Industry for Sustainability Development

Vietnam is a country that produces a variety of agricultural products, including vegetables, tubers, fruits, and processed products. Along with the increase in population, the demand for consumers also increases, and the by-products of farming are increasing and being discharged into the environment. This is one of the critical research issues that need to be solved to ensure sustainability in agriculture. This review summarized recent studies on familiar sources of by-products in Vietnam, such as banana peels, citrus peels, dragon fruit skins, rice bran, and rice husks, and their potential in the food industry. Some solutions are also proposed to solve and turn this low-value raw material into a high-value product and serve a variety of products and consumers in the food industry. Especially after the COVID19 pandemic, the by-products contain valuable and reusable biological resources. These compounds could be future applications to support improving the consumer's immune system and various health benefits. Processed and utilized by-products from food production could not only help increase incomes for farmers, especially in developing countries like Vietnam but also could aid in ensuring food security and sustainability in agricultural production.

Embedded system vehicle based on multi-sensor fusion

As intelligent driving vehicles came out of concept into people’s life, the combination of safe driving and artificial intelligence becomes the new direction of future transportation development. Autonomous driving technology is developing based on control algorithms and model recognitions. In this paper, a cloud-based interconnected multi-sensor fusion autonomous vehicle system is proposed that uses deep learning (YOLOv4) and improved ORB algorithms to identify pedestrians, vehicles, and various traffic signs. A cloud-based interactive system is built to enable vehicle owners to master the situation of their vehicles at any time. In order to meet multiple application of automatic driving vehicles, the environment perception technology of multi-sensor fusion processing has broadened the uses of automatic driving vehicles by being equipped with automatic speech recognition (ASR), vehicle following mode and road patrol mode. These functions enable automatic driving to be used in applications such as agricultural irrigation, road firefighting and contactless delivery under new coronavirus outbreaks. Finally, using the embedded system equipment, an intelligent car was built for experimental verification, and the overall recognition accuracy of the system was over 96%. Author

Preoperative COVID-19 RT-PCR Testing of Asymptomatic Patients for Ophthalmic Surgery: Experience from Two Tertiary Referral Centers

Purpose: The aim of this study was to evaluate how prevalent asymptomatic SARS-CoV-2 virus infection (COVID-19) is among patients undergoing ophthalmic surgery at two tertiary referral hospitals. Material(s) and Method(s): This retrospective study included patients without COVID-19 symptoms who underwent preoperative screening using reverse transcription-polymerase chain reaction (RT-PCR) before ophthalmic surgery at the Kocaeli University and Gaziantep University departments of ophthalmology [between September 1, 2020, and December 15, 2020 (group 1);between March 1, 2021, and May 30, 2021 (group 2)]. Patients scheduled for surgery and followed up in the retina, glaucoma, pediatric ophthalmology and strabismus, cataract and refractive surgery, and cornea departments were examined. Result(s): RT-PCR was positive for SARS-CoV-2 in 12 (1.4%) of 840 patients in group 1 and 7 (1.1%) out of 600 patients in group 2. None of the patients were symptomatic of COVID-19. The majority of the patients were scheduled for retina or cataract and refractive surgery in both groups (group 1;retina: 29.2%, cataract and refractive: 57.0%, group-2;retina: 31.3%, cataract and refractive: 54.5%). SARS-CoV-2 RT-PCR testing was positive for seven patients in group 1 (7/245, 2.9%) and five patients in group 2 (5/188, 2.6%) who were scheduled for retinal surgery. Conclusion(s): The necessity, availability, and practicality of COVID-19 RT-PCR testing prior to ophthalmic surgeries varies depending on the protocols of each institution. COVID-19 RT-PCR testing is suggested especially before vitreoretinal surgeries and general anesthesia procedures, because of the difficulty in managing postoperative complications.Copyright © 2023 Gazi Eye Foundation. All rights reserved.

Improved method for COVID-19 Classification of Complex-Architecture CNN from Chest CT volumes using Orthogonal Ensemble Networks

This paper introduces the improved method for the COVID-19 classification based on computed tomography (CT) volumes using a combination of a complex-architecture convolutional neural network (CNN) and orthogonal ensemble networks (OEN). The novel coronavirus disease reported in 2019 (COVID-19) is still spreading worldwide. Early and accurate diagnosis of COVID-19 is required in such a situation, and the CT scan is an essential examination. Various computer-aided diagnosis (CAD) methods have been developed to assist and accelerate doctors' diagnoses. Although one of the effective methods is ensemble learning, existing methods combine some major models which do not specialize in COVID-19. In this study, we attempted to improve the performance of a CNN for the COVID-19 classification based on chest CT volumes. The CNN model specializes in feature extraction from anisotropic chest CT volumes. We adopt the OEN, an ensemble learning method considering inter-model diversity, to boost its feature extraction ability. For the experiment, We used chest CT volumes of 1283 cases acquired in multiple medical institutions in Japan. The classification result on 257 test cases indicated that the combination could improve the classification performance. © 2023 SPIE.

Qualitative methods in a learning healthcare system: Balancing rigour and the need for rapid analysis and dissemination

Qualitative methods are used to capture stakeholder perspectives within learning healthcare systems (LHS), but there is a need to specify methods that balance rigour and pragmatic approaches to inform quality improvement (QI). Utilizing examples from two QI projects within the OTNY LHS, we illustrate methods and strategies that build team capacity and flexibility to respond to an evolving LHS. Method(s): Qualitative methods were tailored to fit each project's timelines and goals, to inform both practice and research. Tools to facilitate rapid cycle feedback included interview/focus group summary templates, aggregate summaries that synthesize findings by stakeholder group, case matrix templates for rapid extraction and systematic categorization of data along topic areas, and dissemination materials adapted for stakeholder audience and project phases. Strategies to maintain rigour included processes for data reduction and interpretation, a multi-disciplinary approach for analysis, frequent consensus-based meetings, data triangulation, and member checks. Result(s): Rapid cycle approaches yielded interim results that reshaped research questions or identified critical gaps. Case summary analysis exploring the impact of COVID-19 revealed limited information on telehealth challenges amongst OTNY participants, necessitating a shift in recruitment and interview focus. For another project, analytic methods were sequenced to rapidly inventory suggestions from interview summaries on how to enhance OTNY practice to better address racism, while subsequent thematic analysis of transcripts captured participants' experiences of racism for context. Challenges included concurrent alignment of data collection and analysis, tailoring summary templates to maximize utility for rapid analysis, and maintaining flexibility to respond to evolving findings and LHS stakeholder input. Conclusion(s): The diverse methods and strategies illustrated by these projects offer guidance for balancing.

COVID-19 Detection by Using Handcrafted Features Extracted From Chest CT-Scan Images

Coronavirus illness, which was initially diagnosed in 2019 but has propagated rapidly across the globe, has led to increased fatalities. According to professional physicians who examined chest CT scans, COVID-19 behaves differently than various viral cases of pneumonia. Even though the illness only recently emerged, a number of research investigations have been performed wherein the progression of the disease impacts mostly on the lungs are identified using thoracic CT scans. In this work, automated identification of COVID-19 is used by using machine learning classifier trained on more than 1000+ lung CT Scan images. As a result, immediate diagnosis of COVID-19, which is very much necessary in the opinion of healthcare specialists, is feasible. To improve detection accuracy, the feature extraction method are applied on regions of interests. Feature extraction approaches, including Discrete Wavelet Transform (DWT), Grey Level Cooccurrence Matrix (GLCM), Grey Level Run Length Matrix (GLRLM), and Grey-Level Size Zone Matrix (GLSZM) algorithms are used. Then the classification by using Support Vector Machines (SVM) is used. The classification accuracy is assessed by using precision, specificity, accuracy, sensitivity and F-score measures. Among all feature extraction methods, the GLCM approach has given the optimum classification accuracy of 95.6%. . © 2023 IEEE.

HarDNet and Dual-Code Attention Mechanism based model for medical images segmentation

During the formation of medical images, they are easily disturbed by factors such as acquisition devices and tissue backgrounds, causing problems such as blurred image backgrounds and difficulty in differentiation. In this paper, we combine the HarDNet module and the multi-coding attention mechanism module to optimize the two stages of encoding and decoding to improve the model segmentation performance. In the encoding stage, the HarDNet module extracts medical image feature information to improve the segmentation network operation speed. In the decoding stage, the multi-coding attention module is used to extract both the position feature information and channel feature information of the image to improve the model segmentation effect. Finally, to improve the segmentation accuracy of small targets, the use of Cross Entropy and Dice combination function is proposed as the loss function of this algorithm. The algorithm has experimented on three different types of medical datasets, Kvasir-SEG, ISIC2018, and COVID-19CT. The values of JS were 0.7189, 0.7702, 0.9895, ACC were 0.8964, 0.9491, 0.9965, SENS were 0.7634, 0.8204, 0.9976, PRE were 0.9214, 0.9504, 0.9931. The experimental results showed that the model proposed in this paper achieved excellent segmentation results in all the above evaluation indexes, which can effectively assist doctors to diagnose related diseases quickly and improve the speed of diagnosis and patients’quality of life. Author

Claim Extraction and Dynamic Stance Detection in COVID-19 Tweets

The information ecosystem today is noisy, and rife with messages that contain a mix of objective claims and subjective remarks or reactions. Any automated system that intends to capture the social, cultural, or political zeitgeist, must be able to analyze the claims as well as the remarks. Due to the deluge of such messages on social media, and their tremendous power to shape our perceptions, there has never been a greater need to automate these analyses, which play a pivotal role in fact-checking, opinion mining, understanding opinion trends, and other such downstream tasks of social consequence. In this noisy ecosystem, not all claims are worth checking for veracity. Such a check-worthy claim, moreover, must be accurately distilled from subjective remarks surrounding it. Finally, and especially for understanding opinion trends, it is important to understand the stance of the remarks or reactions towards that specific claim. To this end, we introduce a COVID-19 Twitter dataset, and present a three-stage process to (i) determine whether a given Tweet is indeed check-worthy, and if so, (ii) which portion of the Tweet ought to be checked for veracity, and finally, (iii) determine the author's stance towards the claim in that Tweet, thus introducing the novel task of topic-agnostic stance detection. © 2023 ACM.

You & type 2 @ home: Testing a fully remote model for diabetes annual reviews using the you & type 2 personalised care and support planning pathway in combination with remote monitoring technologies at two primary care practices in South London

Aims: A proof-of- concept pilot investigating the profile of person who engaged with remote testing for their annual diabetes review, and service user (SU) and primary care practice acceptability for completing annual diabetes review tests remotely (blood pressure, finger-stick blood test and urine test). Method(s): A mixed methods evaluation based on SU surveys sent to all 144 pilot participants, semi-structured SU and staff interviews, and demographic and clinical data extraction from primary care electronic patient record system. Result(s): Profile: The pathway was considered suitable for people who were working, digitally capable, younger, had household support to complete the tests, had non-complex diabetes, or a combination of these attributes. It was deemed less suitable for the very elderly, the less digitally capable, those with complex health needs or socially isolated. SU Acceptability: Interviewees and survey respondents overall deemed the remote tests acceptable for use. Convenience and reduced exposure to Covid-19 were motivating factors for participation. Preference for face-to- face care or concerns around using digital technologies were key reasons for decline. Staff Acceptability: The pathway was deemed acceptable and was successfully implemented at both practices. Support from a designated pathway co-ordinator and project manager were key factors linked to acceptability and success. The remote pathway was seen as an opportunity to reduce primary care pressures on in-person care. Conclusion(s): It is possible to successfully conduct annual diabetes reviews remotely. Although not appropriate nor desirable to everyone, remote testing provides a viable alternative to in-person testing for certain individuals.

Prevalence of Glaucoma in Cataract Patients - a Comparison between Pre-Covid and Post-Covid Periods

Objectives: Globally, cataract and glaucoma are the predominant causes of blindness. Screening glaucoma in patients referred for cataract surgery is a convenient tool for detecting glaucoma cases in rural population. The COVID period has adversely affected eye care as the routine screening and follow-ups at hospital were substantially reduced owing to pandemic restrictions. We aim to study the impact of COVID on detection of glaucoma in patients with cataract. Method(s): It was a retrospective study conducted to compare the prevalence of glaucoma in rural patients presenting with cataract pre- and post-COVID. Details of 975 consecutive patients each were taken prior to March 2020 (pre-COVID) and after October 2021 (post-COVID) from hospital database and patient case files. Result(s): The prevalence of glaucoma was higher during the pre-COVID time (3.8%) as compared to pre-COVID (3.8%), but the result was not statistically significant. In both the groups, primary open-angle glaucoma was the pre-dominant form of glaucoma, with prevalence being 1.5% and 2.2% in the pre-COVID and post-COVID groups, respectively. The mean intraocular pressure and mean VCDR values were higher in the post-COVID group as compared to the pre-COVID group, and the result was statistically significant. Conclusion(s): This was the first study to compare the prevalence of glaucoma in patients with cataract in rural population in the pre-COVID and post-COVID periods. In the aftermath of the pandemic, the present study emphasizes the role of screening and follow-ups in glaucoma management to prevent irreversible loss of vision.Copyright © 2023 The Authors.

Digital interventions for substance use disorders in young people: Rapid review

This rapid review provides an overview of recent literature on the nature of digital interventions for young people in terms of technologies used, substances and populations targeted, and theoretical or therapeutic models employed. A keyword search was conducted using MEDLINE and other databases for 2015-2021. Following a title/ and full-text screening of articles and consensus decision on study inclusion, data extraction proceeded using an extraction grid. Data synthesis relied on an adapted conceptual framework (Stockings et al., 2016) that involved a three-level treatment spectrum for youth substance use (prevention, early intervention, and treatment). The review identified 43 articles describing 39 digital interventions. Most were early interventions (n = 28), followed by prevention (n = 6) and treatment (n = 5). Of the five technologies identified, web-based interventions (n = 14) were most common. Digital interventions have mainly focused on alcohol use (n = 20), reflecting limited concern for other substance use and co-occurring use. Yet the rise in substance use and related harms during the Covid-19 pandemic highlights a critical need for more innovative substance use interventions. Technologies with more immersive and interactive features, such as VR and game-based interventions, call for further exploration. Only one intervention was culturally tailored and purposefully designed for gender minority youth, and another was geared to young men. As well, most interventions used a personalized or normative feedback approach, while a harm reduction approach guided only one intervention. The incorporation of culturally tailored interventions and harm reduction approaches may promote uptake and stronger engagement with digital interventions amongst youth.

LLDSNet: Limited labeled data segmentation network for COVID-19 CT lesion segmentation

With the COVID-19 outbreak in 2019, the world is facing a major crisis and people's health is at serious risk. Accurate segmentation of lesions in CT images can help doctors understand disease infections, prescribe the right medicine and control patients' conditions. Fast and accurate diagnosis not only can make the limited medical resources get reasonable allocation, but also can control the spread of disease, and computer-aided diagnosis can achieve this purpose, so this paper proposes a deep learning segmentation network LLDSNet based on a small amount of data, which is divided into two modules: contextual feature-aware module (CFAM) and shape edge detection module (SEDM). Due to the different morphology of lesions in different CT, lesions with dispersion, small lesion area and background area imbalance, lesion area and normal area boundary blurred, etc. The problem of lesion segmentation in COVID-19 poses a major challenge. The CFAM can effectively extract the overall and local features, and the SEDM can accurately find the edges of the lesion area to segment the lesions in this area. The hybrid loss function is used to avoid the class imbalance problem and improve the overall network performance. It is demonstrated that LLDSNet dice achieves 0.696 for a small number of data sets, and the best performance compared to five currently popular segmentation networks. © 2023 SPIE.

COVID-VTS: Fact Extraction and Verification on Short Video Platforms

We introduce a new benchmark, COVID-VTS, for fact-checking multi-modal information involving short-duration videos with COVID19-focused information from both the real world and machine generation. We propose, TwtrDetective, an effective model incorporating cross-media consistency checking to detect token-level malicious tampering in different modalities, and generate explanations. Due to the scarcity of training data, we also develop an efficient and scalable approach to automatically generate misleading video posts by event manipulation or adversarial matching. We investigate several state-of-the-art models and demonstrate the superiority of TwtrDetective. © 2023 Association for Computational Linguistics.

Medical image mis-segmentation region refinement framework based on dynamic graph convolution

In medical image segmentation tasks, it is hard for traditional Convolutional Neural Network (CNN) to capture essential information such as spatial structure and global contextual semantic features since it suffers from a limited receptive field. The deficiency weakens the CNN segmentation performance in the lesion boundary regions. To handle the aforementioned problems, a medical image mis-segmentation region refinement framework based on dynamic graph convolution is proposed to refine the boundary and under-segmentation regions. The proposed framework first employs a lightweight dual-path network to detect the boundaries and nearby regions, which can further obtain potentially misclassified pixels from the coarse segmentation results of the CNN. Then, we construct the pixels into the appropriate graphs by CNN-extracted features. Finally, we design a dynamic residual graph convolutional network to reclassify the graph nodes and generate the final refinement results. We chose UNet and its eight representative improved networks as the basic networks and tested them on the COVID, DSB, and BUSI datasets. Experiments demonstrated that the average Dice of our framework is improved by 1.79%, 2.29%, and 2.24%, the average IoU is improved by 2.30%, 3.53%, and 2.39%, and the Se is improved by 5.08%, 4.78%, and 5.31% respectively. The experimental results prove that the proposed framework has the refinement capability to remarkably strengthen the segmentation result of the basic network. Furthermore, the framework has the advantage of high portability and usability, which can be inserted into the end of mainstream medical image segmentation networks as a plug-and-play enhancement block.

Thermal Shock Viral RNA Release in COVID-19 Testing

Introduction/Objective The public health emergency of the COVID-19 pandemic emphasized the crucial role of medical laboratory professionals and scientists in molecular diagnostics laboratories to ensure success in infection control strategies. The demand for laboratory testing using nucleic acid amplification tests to detect SARS-CoV-2 RNA imposed strains in laboratory supplies. Here, we explored an alternative cost-effective solution that will simplify the pre-PCR steps by using a simple heating method to release viral RNA. Methods/Case Report Samples tested using the reference automated extraction method were used:100 samples identified as positive for SARS-CoV-2 RNA and 500 samples tested negative for SARS-CoV-2 RNA were used for the study and sorted with equal distribution according to Ct values of (a) <20, (b) 20–30, and (c) >30.100 ul from swab preserved in Universal Transport Medium was treated with 30 μg of proteinase K, and another set was tested without proteinase K pre-treatment. All samples with or without proteinase K were diluted to minimize PCR inhibitors. The thermal shock protocol was set at (98°C, 5 minutes;4°C, 2 minutes) and screened for purity. Performance and method verification studies were performed. Internal extraction, positive template, and no template controls were markers used for testing quality. The experimental study was performed by qualified testing personnel and all under the same experimental conditions. Results (if a Case Study enter NA) The Ct values from the thermal shock RNA release were compared to the automated extraction method and statistically analyzed.The criteria for acceptability for validation of this new RNA extraction proceeding were set to 100% concordance compared to the commercial kit using an automated extraction. PCR efficiency was at 98% and a slope of -3.3. Within run precision of 2% and limits of detection from 200 to 20,000 copies/uL The method 100% (50/50) concordance on samples previously identified as negative by automated methods and identified 86% (86/100) with a mean difference of 3 Ct. Conclusion Our findings suggest that the thermal shock treatment of nasopharyngeal swabs in viral transport media can successfully extract viral nucleic acid for nucleic acid amplification and is a reasonable alternative for chemical extraction methods when molecular diagnostic laboratories persistently encounter supply chain issues.

An approach for Face Detection and Face Recognition using OpenCV and Face Recognition Libraries in Python

With the extraordinary growth in images and video data sets, there is a mind-boggling want for programmed understanding and evaluation of data with the assistance of smart frameworks, since physically it is a long way off. Individuals, unlike robots, have a limited capacity to distinguish unexpected expressions. As a result, the programmed face proximity frame- work is important in face identification, appearance recognition, head-present evaluation, human-PC cooperation, and other applications. Software that uses facial recognition for face detection and identification is regarded as biometric. This study converts the mathematical aspects of a person's face into a face print, which is then stored in a database to verify an individual's identification. A deep learning system compares a digital image or an image taken quickly to a previously stored image(which is saved in the database). The face has a significant function in interpersonal communication for identifying oneself. Face recognition technology determines the size and placement of a human face in a digital picture. Facial recognition software has a wide range of uses in the consumer market and in the security and surveillance sectors. The COVID pandemic has brought facial recognition into greater focus lately than ever before. Face detection and recognition play a vital part in security systems that people need to interact with without making physical contact. The pattern of online exam proctoring is employing face detection and recognition. Facial recognition is used in the airline sector to enable rapid, accurate identification and verification at every stage of the passenger trip. In this research, we focused on image quality because it is the major drawback in existing algorithms and used OPEN CV, Face Recognition, and designed algorithms using libraries in python. This study discusses a method for facial recognition along with its implementation and applications. © 2023 IEEE.

ID: 219407 The Cardiac Patient and Neuromodulation, When Anticoagulation Dictates the Intervention

Introduction: Patients with cardiac comorbidities present unique challenges for undergoing interventional pain procedures. Consensus guidelines on safe anticoagulation management are categorized by procedure, patient specific bleeding risk factors, and class of anticoagulation (Table 1, Table 2).1 Specifically, some procedures occur in close proximity to the spinal cord, require large gauge needles and styletted leads, while others are in compressible locations with minimal tissue disruption. Further, pain-induced hypercoagulation increases the risk of thrombo-vascular events.1 This accentuates the importance of interdisciplinary perioperative coordination with the prescribing cardiologist. Case: A 71-year-old male with past-medical-history of CABG, bilateral femoral-popliteal bypass, atrial fibrillation on apixaban and ticagrelor, and multiple cardiac stents presented with intermittent shooting axial back pain radiating to right buttock, lateral thigh, and calf, worsened with activity. MRI demonstrated thoracic myelomalacia, multi-level lumbar disc herniation, and moderate central canal stenosis. An initial multi-model treatment approach utilizing pharmacologic agents, physical therapy, ESI's, and RFA failed to alleviate symptoms. After extensive discussion with his cardiologist, he was scheduled for a three-day SCS trial. Ticagrelor and apixaban were held throughout the 3-day trial and for 5 and 3 days prior, respectively, while ASA was maintained. Successful trial with tip placement at T6 significantly improved function and pain scores (Figure 1). Upon planned percutaneous implant, the cardiologist recommended against surgical implantation and holding anticoagulation. Alternatively, the patient underwent bilateral lumbar medial branch PNS implant with sustained improvement in lower back symptoms. However, he contracted COVID, resulting in delayed lead explanation (>60 days) without complication. Conclusion(s): Interventional pain practice advisories are well established for anticoagulation use in the perioperative period.1,2 However, there is limited high-quality research on the appropriate length to hold anticoagulation prior to surgery for high thrombotic risk patients. Collegial decision making with the cardiologist was required to avoid deleterious procedural complications. However, they may be unfamiliar with the nuances between interventions or between trial and implant. Prospective studies have shown that low risk procedures, such as the PNS, may not require holding anticoagulants.3 Other case data has demonstrated post-SCS epidural hematoma with ASA use after being held for 1-week prior to surgery. Our patient was unable to undergo SCS implant and instead elected for a lower risk procedure with excellent efficacy. 4 However, delayed PNS lead extraction due to COVID19 hospitalization presented further risk of infection and lead fracture.5 PNS may prove to be an appropriate treatment option for patients who are anticoagulated and are not SCS candidates. Disclosure: Elliot Klein, MD,MPH: None, Clarence Kong, MD: None, Shawn Sidharthan, MD: None, Peter Lascarides, DO: None, Yili Huang, DO: NoneCopyright © 2023

TwIdw—A Novel Method for Feature Extraction from Unstructured Texts

Featured ApplicationThe research has a potential application in the field of fake news detection. By using the feature extraction technique, TwIdw, proposed in this paper, more relevant and informative features can be extracted from the text data, which can lead to an enhancement in the accuracy of the classification models employed in these tasks.This research proposes a novel technique for fake news classification using natural language processing (NLP) methods. The proposed technique, TwIdw (Term weight–inverse document weight), is used for feature extraction and is based on TfIdf, with the term frequencies replaced by the depth of the words in documents. The effectiveness of the TwIdw technique is compared to another feature extraction method—basic TfIdf. Classification models were created using the random forest and feedforward neural networks, and within those, three different datasets were used. The feedforward neural network method with the KaiDMML dataset showed an increase in accuracy of up to 3.9%. The random forest method with TwIdw was not as successful as the neural network method and only showed an increase in accuracy with the KaiDMML dataset (1%). The feedforward neural network, on the other hand, showed an increase in accuracy with the TwIdw technique for all datasets. Precision and recall measures also confirmed good results, particularly for the neural network method. The TwIdw technique has the potential to be used in various NLP applications, including fake news classification and other NLP classification problems.