Reduce Background Noise Research Articles

Decision fusion based on vibrometric data for increase of reliability of damage detection and indentification for plate-like structures

In the realm of structural health monitoring (SHM) for plate-like structures, the need for accurate and reliable damage detection and identification techniques is paramount. This paper explores the application of decision fusion algorithms to vibrometric data, with a focus on increasing the reliability of damage assessment for such structures. The methodology involves leveraging diverse input data sources, including variations in data processing algorithms, excitation parameters, and multiple measurements. By amalgamating these sources of diversity, we aim to mitigate measurement noise and amplify damage indications, ultimately enhancing the precision of the final diagnostic outcome. The study presents results obtained from laser vibrometry measurements on a 10 mm thick aluminum sample containing flat-bottom holes of varying diameters. A network of 12 PZT transducers was used to excite elastic waves, with impulse and chirp excitation techniques employed. Convolution techniques using OpenCV libraries were applied for data processing and fusion. The outcomes reveal a significant reduction in background noise, demonstrating the effectiveness of the fusion approach. Furthermore, the presentation outlines a roadmap for the future development of this methodology, which includes automating data processing, evaluating the impact of damage indicators, processing parameters and usage of different excitation transducers.

Development of a diamagnetic loop in KAIMIR

We developed a diamagnetic loop for the estimation of plasma stored energy in the KAIST Magnetic Mirror magnetic mirror device [Oh et al., J. Plasma Phys. 90, 975900202 (2024)]. Diamagnetic loops are used to estimate the plasma stored energy from measurements of the diamagnetic flux in plasma with an applied external magnetic field. However, diamagnetic flux measurements are accompanied by the vacuum flux, which generally exceeds the diamagnetic flux by over 10 000 times. Therefore, it is critical to attain a high signal-to-noise ratio with minimized noise in diamagnetic flux measurements. In this study, we employed a novel method to reduce background noise and improve the signal-to-noise ratio. Using two identical loops with opposite polarities, we successfully removed parasitic capacitive noise from the external insulation while amplifying the inductive signal two times. To eliminate the vacuum flux, we utilized two coaxial loops with different radii positioned at the same axial location. Results obtained from six paired loops confirmed the successful removal of the vacuum flux. The plasma stored energy was also found to agree well with Langmuir probe measurements, which verifies the diamagnetic flux measurements using the developed loop.

Automated object detection for muon tomography data analysis

In recent years, there have been ongoing efforts to improve screening technologies to improve security and prevent terrorist threats. The most widely used technologies for scanning shipping containers are gamma and X-ray radiography, which can be harmful to operators and the environment. Muon tomography screening systems are considered as a potential tool to enhance border security and prevent terrorist threats or smuggling, especially in the context of shipping container inspections. Muon tomography uses naturally occurring cosmic ray muons to create detailed images of the inside of objects, such as shipping containers, without the need for physical intervention. Various realistic smuggling scenarios were simulated using the GEANT4 toolkit. The implemented filtering algorithms successfully reduce background noise from the surrounding cargo, enabling the detection of concealed threats and contraband. With the tools provided by the ROOT data analysis package, prohibited items can be automatically detected and localized in a cargo container.

AGD-GAN: Adaptive Gradient-Guided and Depth-supervised generative adversarial networks for ancient mural sketch extraction

To address the overlooked issues of multi-scale detail feature extraction and disease noise suppression in mural sketch extraction, we proposed a novel generative adversarial network with Adaptive Gradient-guided and Depth-supervised (AGD-GAN), which can generate high-quality mural sketches in an unsupervised manner. AGD-GAN first enriches mural feature details at various scales by introducing a cross-channel residual attention module, significantly improving the detail extraction effects. The Adaptive Gradient-guided strategy is based on the gradient attention maps, which can adaptively adjust the weights between gradient information and detail feature preservation according to the degree of damage in the mural, further balancing the preservation of mural detail features and the suppression of disease noise. Finally, the Depth-supervised reinforces constraints on the position and shape of the sketches by introducing a depth-predicted loss function, thereby reducing background noise interference and controlling the shape of the generated sketches. We compared with eight state-of-the-art algorithms quantitatively and qualitatively, experimental results demonstrate the promising capability of the proposed AGD-GAN to extract clear, coherent, and comprehensive mural sketches. We release the source code at https://github.com/Alice77bai/AGD-GAN.

Advanced optimization-based weighted features for ensemble deep learning smart occupancy detection network for road traffic parking

In day-to-day activities, the advanced technology like Internet of Things (IoT) emerges to improve the lifestyle of people. In metropolitan cities, real-time parking is the long-lasting problem that we face in our daily life activities. Urban parking regulation gained more attention because of its capability to diminish energy consumption, congested traffic, and manifestation. The parking space detection for vehicles in real-time is a crucial role for on-street parking control models where the data is delivered to the drivers through internet with the help of Global Positioning System (GPS). Hence, the network congestion gets increased, where the existing model does not performed well in the parking space effectively. This research paper recommends a novel strategy for detecting occupancy in road traffic parking by IoT and deep networks to overcome these difficulties. At first, the required information is gathered from the standard datasets as CNRPark + EXT for further processing. After the data collection process, the preprocessing phase is executed, where the data cleaning and data transformation approaches are done with a multi-scale retinex network. The pre-processing technique is effectively performed to reduce background noise. Further, the feature extraction is progressed through the Residual Attention Network, where the relevant features are extracted from the pre-processed data. Subsequently, the optimal features and weights are selected by the novel Advanced Pelican Optimization Algorithm (APOA) to get the optimal weighted feature selection. This optimal weighted feature is further given to the Ensemble Deep Networks (EDN) by integrating the Deep Conditional Random Field (DCRF), and Extreme Deep Learning (EDL) model. In final process, the detection of occupancy in road traffic parking is determined by averaging both classifiers outcomes. Finally, the results are compared with numerous optimization algorithms and classifiers to ensure the developed system's efficacy. Throughout the validation, the developed model outperforms with an accuracy as 96.6, precision as 92.9 and Mathew's Correlation Coefficient (MCC) as 92.3. Thus, the developed model shows better performance than the existing traditional approaches.

Effect of manual OCTA segmentation correction to improve image quality and visibility of choroidal neovascularization in AMD

In this retrospective case series on neovascular age-related macular degeneration (nAMD), we aimed to improve Choroidal Neovascularization (CNV) visualization in Optical Coherence Tomography Angiography (OCTA) scans by addressing segmentation errors. Out of 198 eyes, 73 OCTA scans required manual segmentation correction. We compared uncorrected scans to those with minimal (2 corrections), moderate (10 corrections), and detailed (50 corrections) efforts targeting falsely segmented Bruch’s Membrane (BM). Results showed that 55% of corrected OCTAs exhibited improved quality after manual correction. Notably, minimal correction (2 scans) already led to significant improvements, with additional corrections (10 or 50) not further enhancing expert grading. Reduced background noise and improved CNV identification were observed, with the most substantial improvement after two corrections compared to baseline uncorrected images. In conclusion, our approach of correcting segmentation errors effectively enhances image quality in OCTA scans of nAMD. This study demonstrates the efficacy of the method, with 55% of resegmented OCTA images exhibiting enhanced quality, leading to a notable increase in the proportion of high-quality images from 63 to 83%.

AeroDetectNet: a lightweight, high-precision network for enhanced detection of small objects in aerial remote sensing imagery

Object detection in remote sensing imagery exhibits difficulties due to complex backgrounds, diverse object scales, and intricate spatial context relationships. Motivated by the problems mentioned above, this paper introduces AeroDetectNet, a novel lightweight and high-precision object detection network custom-designed for aerial remote sensing scenarios, building upon the YOLOv7-tiny algorithm. It enhances performance through four key improvements: the normalized Wasserstein distance for consistent object size sensitivity, the Involution module for reduced background noise, a self-designed RCS-biformer module for better spatial context interpretation, and a self-designed WF-CoT SPPCSP feature pyramid for improved feature map weighting and context capture. Ablation studies conducted on a hybrid dataset composed of three open-source remote sensing datasets (including NWPU VHR-10 remote sensing images, RSOD remote sensing images, and VisDrone UAV images) have demonstrated the effectiveness of four improvements specifically for small-size object detection. Visualizations through Grad-CAM further demonstrate AeroDetectNet’s capacity to extract and focus on key object features. Upon individual testing across three open-source datasets, AeroDetectNet has successfully demonstrated its ability to identify objects in images with a smaller pixel area. Through experimental comparisons with other related studies, the AeroDetectNet achieved a competitive mAP while maintaining fewer model parameters, highlighting its highly accurate and lightweight properties.

A Simple but Powerful Method to Reduce Background Noise in Celestial Gamma-Ray Images

Reducing background noise is essential for clear data analysis and signal detection. In this study, we introduce a novel denoising algorithm tailored for gamma-ray astronomical data, utilizing local density sampling techniques. The proposed algorithm preferentially amplifies regions of high local density, which are indicative of potential sources, while diminishing the influence of low-density areas. This selective enhancement significantly bolsters the performance of clustering algorithms in pinpointing point sources, effectively minimizing the interference of high-density noise that could be mistaken for genuine sources. We have implemented this algorithm on Fermi Large Area Telescope data, and our results showcase a marked advancement in the clustering algorithm's capacity to discard false sources.

LightCF-Net: A Lightweight Long-Range Context Fusion Network for Real-Time Polyp Segmentation.

Automatically segmenting polyps from colonoscopy videos is crucial for developing computer-assisted diagnostic systems for colorectal cancer. Existing automatic polyp segmentation methods often struggle to fulfill the real-time demands of clinical applications due to their substantial parameter count and computational load, especially those based on Transformer architectures. To tackle these challenges, a novel lightweight long-range context fusion network, named LightCF-Net, is proposed in this paper. This network attempts to model long-range spatial dependencies while maintaining real-time performance, to better distinguish polyps from background noise and thus improve segmentation accuracy. A novel Fusion Attention Encoder (FAEncoder) is designed in the proposed network, which integrates Large Kernel Attention (LKA) and channel attention mechanisms to extract deep representational features of polyps and unearth long-range dependencies. Furthermore, a newly designed Visual Attention Mamba module (VAM) is added to the skip connections, modeling long-range context dependencies in the encoder-extracted features and reducing background noise interference through the attention mechanism. Finally, a Pyramid Split Attention module (PSA) is used in the bottleneck layer to extract richer multi-scale contextual features. The proposed method was thoroughly evaluated on four renowned polyp segmentation datasets: Kvasir-SEG, CVC-ClinicDB, BKAI-IGH, and ETIS. Experimental findings demonstrate that the proposed method delivers higher segmentation accuracy in less time, consistently outperforming the most advanced lightweight polyp segmentation networks.

Enabling Low-Dose In Vivo Benchtop X-ray Fluorescence Computed Tomography through Deep-Learning-Based Denoising.

X-ray Fluorescence Computed Tomography (XFCT) is an emerging non-invasive imaging technique providing high-resolution molecular-level data. However, increased sensitivity with current benchtop X-ray sources comes at the cost of high radiation exposure. Artificial Intelligence (AI), particularly deep learning (DL), has revolutionized medical imaging by delivering high-quality images in the presence of noise. In XFCT, traditional methods rely on complex algorithms for background noise reduction, but AI holds promise in addressing high-dose concerns. We present an optimized Swin-Conv-UNet (SCUNet) model for background noise reduction in X-ray fluorescence (XRF) images at low tracer concentrations. Our method's effectiveness is evaluated against higher-dose images, while various denoising techniques exist for X-ray and computed tomography (CT) techniques, only a few address XFCT. The DL model is trained and assessed using augmented data, focusing on background noise reduction. Image quality is measured using peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM), comparing outcomes with 100% X-ray-dose images. Results demonstrate that the proposed algorithm yields high-quality images from low-dose inputs, with maximum PSNR of 39.05 and SSIM of 0.86. The model outperforms block-matching and 3D filtering (BM3D), block-matching and 4D filtering (BM4D), non-local means (NLM), denoising convolutional neural network (DnCNN), and SCUNet in both visual inspection and quantitative analysis, particularly in high-noise scenarios. This indicates the potential of AI, specifically the SCUNet model, in significantly improving XFCT imaging by mitigating the trade-off between sensitivity and radiation exposure.

“A lonely time for deaf and hard of hearing kids”: A qualitative study of the impact of pandemic precautions on classroom communication for adolescents with hearing loss

BackgroundPandemic public health measures, such as masks and social distancing, present unique challenges for people who are hard-of-hearing. This study sought to understand how adolescents with varying levels of hearing loss would describe their experiences communicating in a classroom environment during a pandemic and its associated public health measures. MethodsQualitative study utilizing one-on-one semi-structured interviews conducted from July 2021 to April 2022. Interviews were transcribed and analyzed using a reflexive thematic analysis to conceptualize the main themes from the data. Adolescents ages 12–17 who attended school in person during the COVID-19 pandemic with either normal hearing or; bilateral cochlear implants or; bone-anchored hearing aids or; unilateral moderate to severe (40–70 dB) conductive hearing loss secondary to mastoidectomy, were interviewed. Participants were recruited from BC Children's Hospital Otolaryngology clinic via a convenience sample. ResultsFourteen adolescents were interviewed with a median age of 15, 9 with hearing loss and 5 with normal hearing. Pandemic associated challenges such as masks muffling speech, protocol fatigue, and missing pre-pandemic life were present in both the hearing loss and normal hearing groups. Classroom communication for adolescents with hearing loss was disproportionately affected by pandemic measures, leading to challenges making friends, feeling behind their peers in learning, and listening fatigue. Resilience was noted among adolescents with hearing loss in their ability to adapt to pandemic measures and changing classroom dynamics. For adolescents with unilateral hearing loss, the pandemic provided an improved listening environment via a reduction in background noise. ConclusionPre-existing classroom communication challenges for adolescents with hearing loss were amplified under pandemic public health measures and shared, in part, by those with normal hearing. These findings can be used to further inform classroom design to the optimize learning environment for deaf and hard of hearing students.

Noise Robust VAD Technique and its Performance Evaluation

In this real world of audio and speech signals, the ability to detect the voice activity and remove the background voice is important. This project presents that to remove noise robust Voice Activity Detection (VAD) technique that combines voice source and vocal tract system information using a zero-frequency filtering (ZFF) approach. As mentioned earlier ZFF technique is employed to combine and to compute a composite signal that encapsulate essential parameters such as fundamental frequency and formats, enabling robust VAD in time domain.This methodology offers a significant advantage in terms of computational efficiency compared to other methods, making it an attractive choice for real-time applications. By applying dynamic thresholding after spectral entropy-based weighting, this approach exhibits resilience or minimum across a range of Signal-to-Noise Ratios (SNRs), further enhancing its ability.This project provides comprehensive utility for audio segmentation, which offers the capability to process individual audio files or entire directories. It also includes the extraction of a composite signal, which is a critical component in enhancing voice quality and reducing background noise. With this, the project is open to use build further and use it for the big industries purpose and for communication purpose. Keywords— Voice Activity Detection (VAD), Zero-Frequency Filtering (ZFF), Signal Noise Ratio (SNR).

Remote sensing image information extraction based on Compensated Fuzzy Neural Network and big data analytics

Medical imaging AI systems and big data analytics have attracted much attention from researchers of industry and academia. The application of medical imaging AI systems and big data analytics play an important role in the technology of content based remote sensing (CBRS) development. Environmental data, information, and analysis have been produced promptly using remote sensing (RS). The method for creating a useful digital map from an image data set is called image information extraction. Image information extraction depends on target recognition (shape and color). For low-level image attributes like texture, Classifier-based Retrieval(CR) techniques are ineffective since they categorize the input images and only return images from the determined classes of RS. The issues mentioned earlier cannot be handled by the existing expertise based on a keyword/metadata remote sensing data service model. To get over these restrictions, Fuzzy Class Membership-based Image Extraction (FCMIE), a technology developed for Content-Based Remote Sensing (CBRS), is suggested. The compensation fuzzy neural network (CFNN) is used to calculate the category label and fuzzy category membership of the query image. Use a basic and balanced weighted distance metric. Feature information extraction (FIE) enhances remote sensing image processing and autonomous information retrieval of visual content based on time-frequency meaning, such as color, texture and shape attributes of images. Hierarchical nested structure and cyclic similarity measure produce faster queries when searching. The experiment’s findings indicate that applying the proposed model can have favorable outcomes for assessment measures, including Ratio of Coverage, average means precision, recall, and efficiency retrieval that are attained more effectively than the existing CR model. In the areas of feature tracking, climate forecasting, background noise reduction, and simulating nonlinear functional behaviors, CFNN has a wide range of RS applications. The proposed method CFNN-FCMIE achieves a minimum range of 4–5% for all three feature vectors, sample mean and comparison precision-recall ratio, which gives better results than the existing classifier-based retrieval model. This work provides an important reference for medical imaging artificial intelligence system and big data analysis.

An enhanced real-time human pose estimation method based on modified YOLOv8 framework

The objective of human pose estimation (HPE) derived from deep learning aims to accurately estimate and predict the human body posture in images or videos via the utilization of deep neural networks. However, the accuracy of real-time HPE tasks is still to be improved due to factors such as partial occlusion of body parts and limited receptive field of the model. To alleviate the accuracy loss caused by these issues, this paper proposes a real-time HPE model called CCAM-Person\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ extbf {CCAM-Person}}$$\\end{document} based on the YOLOv8 framework. Specifically, we have improved the backbone and neck of the YOLOv8x-pose real-time HPE model to alleviate the feature loss and receptive field constraints. Secondly, we introduce the context coordinate attention module (CCAM) to augment the model’s focus on salient features, reduce background noise interference, alleviate key point regression failure caused by limb occlusion, and improve the accuracy of pose estimation. Our approach attains competitive results on multiple metrics of two open-source datasets, MS COCO 2017 and CrowdPose. Compared with the baseline model YOLOv8x-pose, CCAM-Person improves the average precision by 2.8% and 3.5% on the two datasets, respectively.

Computational prediction of protein interactions in single cells by proximity sequencing.

Proximity sequencing (Prox-seq) simultaneously measures gene expression, protein expression and protein complexes on single cells. Using information from dual-antibody binding events, Prox-seq infers surface protein dimers at the single-cell level. Prox-seq provides multi-dimensional phenotyping of single cells in high throughput, and was recently used to track the formation of receptor complexes during cell signaling and discovered a novel interaction between CD9 and CD8 in naïve T cells. The distribution of protein abundance can affect identification of protein complexes in a complicated manner in dual-binding assays like Prox-seq. These effects are difficult to explore with experiments, yet important for accurate quantification of protein complexes. Here, we introduce a physical model of Prox-seq and computationally evaluate several different methods for reducing background noise when quantifying protein complexes. Furthermore, we developed an improved method for analysis of Prox-seq data, which resulted in more accurate and robust quantification of protein complexes. Finally, our Prox-seq model offers a simple way to investigate the behavior of Prox-seq data under various biological conditions and guide users toward selecting the best analysis method for their data.

Adaptive feature enhancement of modulation spectrum of ship radiation noise

The Detection of Envelope Modulation on Noise (DEMON) has been widely used in passive sonar target identification for determining the operational status and propeller structure characteristics of ships. However, the traditional DEMON spectrum is susceptible to environmental noise, which increases the difficulty of feature extraction. To address this issue, this paper investigates an adaptive feature enhancement method that combines Variational Mode Decomposition (VMD) and Singular Value Decomposition (SVD). The VMD algorithm utilizes variational equations to achieve adaptive modal decomposition of multiple feature components, while SVD utilizes the uncorrelated nature of signal and noise to achieve background noise reduction. Simulation results demonstrate that the feature strength and output signal-to-noise ratio of the VMD-SVD algorithm are higher than those of the traditional DEMON algorithm under both uniform and non-uniform modulation modes. Furthermore, processing of actual data validates the superiority of the VMD-SVD algorithm over the traditional DEMON algorithm, with the feature strength and output signal-to-noise ratio being maximally increased by 3 dB and 3.8 dB, respectively, compared to the traditional algorithm.

Exploiting the entire near-infrared spectral range to improve the detection of methane plumes with high-resolution imaging spectrometers

Abstract. Remote sensing emerges as an important tool for the detection of methane plumes emitted by so-called point sources, which are common in the energy sector (e.g., oil and gas extraction and coal mining activities). In particular, satellite imaging spectroscopy missions covering the shortwave infrared part of the solar spectrum are very effective for this application. These instruments sample the methane absorption features at the spectral regions around 1700 and 2300 nm, which enables the retrieval of methane concentration enhancements per pixel. Data-driven retrieval methods, in particular those based on the matched filter concept, are widely used to produce maps of methane concentration enhancements from imaging spectroscopy data. Using these maps enables the detection of plumes and the subsequent identification of active sources. However, retrieval artifacts caused by particular surface components may sometimes appear as false plumes or disturbing elements in the methane maps, which complicates the identification of real plumes. In this work, we use a matched filter that exploits a wide spectral window (1000–2500 nm) instead of the usual 2100–2450 nm window with the aim of reducing the occurrence of retrieval artifacts and background noise. This enables a greater ability to discriminate between surface elements and methane. The improvement in plume detection is evaluated through an analysis derived from both simulated data and real data from areas including active point sources, such as the oil and gas (O&G) industry from San Joaquin Valley (US) and the coal mines from the Shanxi region (China). We use datasets from the Precursore IperSpettrale della Missione Applicativa (PRISMA) and the Environmental Mapping and Analysis Program (EnMAP) satellite imaging spectrometer missions and from the Airborne Visible/Infrared Imaging Spectrometer – Next Generation (AVIRIS-NG) instrument. We find that the interference with atmospheric carbon dioxide and water vapor is generally almost negligible, while co-emission or overlapping of these trace gases with methane plumes leads to a reduction in the retrieved concentration values. Attenuation will also occur in the case of methane emissions situated above surface structures that are associated with retrieval artifacts. The results show that the new approach is an optimal trade-off between the reduction in background noise and retrieval artifacts. This is illustrated by a comprehensive analysis in a PRISMA dataset with 15 identified plumes, where the output mask from an automatic detection algorithm shows an important reduction in the number of clusters not related to CH4 emissions.

The reverse stethoscope technique: a rapid intervention for old age hearing loss

ABSTRACT Objectives Presbycusis is highly prevalent, affecting between a third and two-thirds of elderly populations. Effective communication between patient and physician has been shown to directly correlate with the quality of patient care. The Reverse Stethoscope Technique (RST) involves placing the earpieces into the patient’s ears and speaking into the diaphragm. Here, we aim to show the RST is a simple and effective method to communicate with patients suffering from presbycusis in inpatient settings. Methods Medical inpatients aged >60 years old without cognitive impairment were included in the study. A simple repetition exercise of basic English sentences was performed with and without the RST to assess hearing ability. Patients then undertook a short questionnaire. Results Our study included 109 patients with varied ‘self-reported’ hearing. 50.5% of our cohort reported past communication difficulties with health-care professionals due to hearing difficulties. We identified that the RST increased the mean number of sentences a patient could repeat from 2.6 to 3.5 representing a statistically significant increase (p < 0.001). The technique was most beneficial for patients with ‘okay’ and ‘poor’ hearing. The majority of patients (77/109) also stated the technique reduced background noise. The RST was equally effective with facemasks. Conclusion Current techniques to enhance communication with patients with hearing loss are often underutilized or poorly accessible. Our study identified that the RST is a quick and effective solution that can be easily implemented for patients struggling to communicate with health-care professionals in inpatient scenarios.

Sand Painting Generation Based on Convolutional Neural Networks.

Neural style transfer is an algorithm that transfers the style of one image to another image and converts the style of the second image while preserving its content. In this paper, we propose a style transfer approach for sand painting generation based on convolutional neural networks. The proposed approach aims to improve sand painting generation via neural style transfer, which can address the problem of blurred objects. Furthermore, it can reduce background noise caused by neural style transfers. First, we segment the main objects from the content image. Subsequently, we perform close-open filtering operations on the content image to obtain smooth images. Subsequently, we perform Sobel edge detection to process the images and obtain edge maps. Based on these edge maps and the input style image, we perform neural style transfer to generate sand painting images. Finally, we integrate the generated images to obtain the final stylized sand painting image. The results show that the proposed approach yields good visual effects from sand paintings. Moreover, the proposed approach achieves better visual effects for sand painting than the previous method.

Abstract A009: Targeting Structural Variants (DNA palindromes) in plasma cell-free DNA for Liquid Biopsy-based Cancer Detection

Abstract We aim to expand the utilities of our genome-wide detection method of chromosomal structural aberrations (structural variants, SVs) for blood-based cancer detection (liquid biopsy). SVs, such as genomic amplification, are prevalent DNA abnormalities and define tumor phenotypes. With cell-free DNA (cfDNA) in plasma, a genome-wide assay for SVs seems feasible for cancer detection. However, most current tests probe a part of the genome for cancer gene mutations and differential DNA methylation. Considering liquid biopsy's promise in clinics, approaches targeting SVs must be considered and developed. The low tumor fraction in plasma cfDNA is a major obstacle to developing liquid biopsy-based tests. Targeted assays overcome low signals by ultra-deep NGS, yet deep sequencing for the entire genome for SV detection would be costly and computationally challenging. In this regard, we have an established approach for enriching SVs (DNA palindromes or inverted repeats, also called fold-back inversions) from total genomic DNA to determine their frequency and genomic locations by sequencing (genome-wide analysis of palindrome formation, GAPF-seq). DNA palindromes arise from common events of genome instability, such as illegitimate repair of DNA double-strand breaks and telomere-telomere fusions, as an early step of genomic amplification. Therefore, a technique to identify DNA palindromes could serve as a tumor DNA detection test across tumor types, including breast cancer. The simple GAPF-seq procedure confers advantages over other methods because it simultaneously amplifies the target signal and reduces background noise. GAPF-seq exploits the inherent propensity of denatured DNA palindromes to form double-stranded DNA (dsDNA) by intra-molecular annealing. Genomic DNA will be denatured and quickly renatured to enrich DNA palindromes to favor intra-molecular annealing. Single-strand-specific nuclease S1 will eliminate the remaining single-stranded, non-palindromic DNA. The dsDNA from palindromes will be amplified by the process of preparing libraries for NGS, by which DNA palindromes appear as coverage peaks in the genome. As a proof-of-concept, we examined the read coverage of nearly 3 million 1-kb bins from low coverage NGS GAPF-seq data for 39 breast paired tumor/normal DNA. We found that the chromosomal distribution of the top 1000 high-coverage bins (GAPF profiles) could effectively differentiate breast tumor DNA, including all stage I tumor DNA, from paired normal DNA (AUROC=0.9885). Automated, three-fold cross-validation with machine learning (ML) algorithms from the STREAMLINE platform confirmed the highly accurate binary classification by GAPF profiles. The same approach was applied for 15 cfDNA from breast cancer patients and effectively differentiated cancer patients' cfDNA from healthy individuals' cfDNA (AUROC=0.900). These results support the potential of GAPF-seq as a unique, SV-targeting liquid biopsy approach for breast cancer. Citation Format: Fumie Igari, Hisashi Tanaka, Michael M Murata, Mitsue Saito, Armando E Giuliano. Targeting Structural Variants (DNA palindromes) in plasma cell-free DNA for Liquid Biopsy-based Cancer Detection [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr A009.