Feature-Guided CNN for Denoising Images From Portable Ultrasound Devices

Dong, Guanfang; Ma, Yingnan; Basu, Anup

doi:10.1109/access.2021.3059003

Cited by 29 publications

(6 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the major issues in prenatal anomaly recognition is obtaining high diagnostic accuracy to ensure reliable identification of probable anomalies. To overcome this issue, Explainable AI techniques have emerged as a possible option [132]. The creation and deployment of AI models and algorithms that provide not only accurate predictions but also transparency and comprehensibility in their decision-making process is referred to as explainable AI.…”

Section: Enhancing Diagnostic Accuracymentioning

confidence: 99%

Ultrasound imaging based recognition of prenatal anomalies: a systematic clinical engineering review

Sriraam,

Chinta,

Suresh

et al. 2024

Prog. Biomed. Eng.

View full text Add to dashboard Cite

For prenatal screening, ultrasound imaging allows for real-time observation of developing fetal anatomy. Understanding normal and aberrant forms through extensive fetal structural assessment enables for early detection and intervention. However, the reliability of anomaly diagnosis varies depending on operator expertise and device limits. First trimester scans in conjunction with circulating biochemical markers are critical in identifying high-risk pregnancies, but they also pose technical challenges. Recent engineering advancements in automated diagnosis, such as AI-based ultrasound image processing and multimodal data fusion, are developing to improve screening efficiency, accuracy, and consistency. Still, creating trust in these data-driven solutions is necessary for integration and acceptability in clinical settings. Transparency can be promoted by explainable AI (XAI) technologies that provide visual interpretations and illustrate the underlying diagnostic decision making process. An explanatory framework based on deeplearning is suggested to construct charts depicting anomaly screening results from ultrasound video feeds. AI modeling can then be applied to these charts to connect defects with probable deformations. Overall, engineering approaches that increaseimaging, automation, and interpretability hold enormous promise for altering traditional workflows and expanding diagnostic capabilities for better prenatal care.

show abstract

Section: Enhancing Diagnostic Accuracymentioning

confidence: 99%

Ultrasound imaging based recognition of prenatal anomalies: a systematic clinical engineering review

Sriraam,

Chinta,

Suresh

et al. 2024

Prog. Biomed. Eng.

View full text Add to dashboard Cite

show abstract

“…In 2021 Dong et al [17] a Feature-guided CNN for image denoising using portable ultrasonography equipment was developed. To achieve high-quality denoising outcomes for clinical images, a feature masking layer was utilized to power a tiered denoising system.…”

Section: Literature Surveymentioning

confidence: 99%

FOE NET: Segmentation of Fetal in Ultrasound Images Using V-NET

Pregitha R.,

Kumar R. S.,

Selvakumar C.

2023

Int. j. electr. comput. eng. syst. (Online)

View full text Add to dashboard Cite

Ultrasound is a non-invasive method to diagnose and treat medical conditions. It is becoming increasingly popular to use portable ultrasound scanning devices to reduce patient wait times and make healthcare more convenient for patients. By using ultrasound imaging, you will be able to obtain images with better quality and also gain information about soft tissues. The interference caused by tissues reflected in ultrasound waves resulted in intensified speckle sound, complicating imaging. In this paper, a novel Foe-Net has been proposed for segmenting the fetal in ultrasound images. Initially, the input US images are noise removal phase using two different filters Adaptive Gaussian Filter (AGF) and Adaptive Bilateral Filter (ABF) used to reduce the noise artifacts. Then, the US images are enhanced using CLAHE and MSR for smoothing to enhance the image quality. Finally, the denoised images are input to the V-net is used to segment the fetal in the US images. The experimental outcomes of the proposed Multi-Scale Retinex (MSR) is an image enhancement technique that improves image quality by adjusting its illumination and enhancing details. Foe-Net was measured by specific parameters such as specificity, precision, and accuracy. The proposed Foe-Net achieves an overall accuracy of 99.48%, specificity of 98.56 %, and precision of 96.82 % for segmented fetal in ultrasound images. The proposed Foe-Net attains better pre-processing outcomes at low error rates and, high SNR, high PSNR, and high SSIM values.

show abstract

“…Most of carried works are based on convolutional neural networks (CNNs) mainly for their ability to extract both global and local features out of the data [15,16]. Recently, the denoising CNN called DnCNN [17] based on residual learning has shown good results in various tasks such as image denoising [18] and physical disturbances removing [19]. Applied to spectroscopy, the DnCNN could successfully improve the line spectral estimation [7] and allow a better atomic element identification out of spectra [8].…”

Section: Introductionmentioning

confidence: 99%

Spectroscopic analysis improvement using convolutional neural networks

Saura

Gonzalez

Benkadda

et al. 2023

J. Phys. D: Appl. Phys.

View full text Add to dashboard Cite

Removing noisy components of signals coming from edge tokamak plasmas, astrochemical organic matter or astronomical objects is one of the promising path to improve the underlying elements identification. Methods such as the penalized semi-supervised non negative matrix factorization (PSNMF) used to extract such prominent elements perform well on complex signals. However, it's results' confidence decreases as the noise increases. In this context, we have tried to address this limitation by removing part of the undesired noise in atomic spectra using artificial intelligence (AI) method based on convolutional neural networks (CNN). More specifically, we have tested different architectures of CNN classically used in denoising task: residual CNNs and auto-encoders, to benchmark their respective denoising capacity. The dataset used is made of atom and ion spectra extracted from the NIST Atomic Spectra Database. In the case of ions, we have considered several degrees of ionization. The synthetic added noise is generated from a typical experimental noise profile randomly modified for each signal. Performance of each AI-based noise remover is measured by analysing the increase of the element identification precision obtained by the PSNMF. Auto-encoders and residual CNN are both suitable to remove the undesired noise and increase the PSNMF efficiency even for noise-prevailing signals. In this particular case, the auto-encoder architecture seems to be globally more accurate and should be selected when considering noisy multi-element atomic spectra.

show abstract

Feature-Guided CNN for Denoising Images From Portable Ultrasound Devices

Cited by 29 publications

References 24 publications

Ultrasound imaging based recognition of prenatal anomalies: a systematic clinical engineering review

Ultrasound imaging based recognition of prenatal anomalies: a systematic clinical engineering review

FOE NET: Segmentation of Fetal in Ultrasound Images Using V-NET

Spectroscopic analysis improvement using convolutional neural networks

Contact Info

Product

Resources

About