Current status of deep learning applications in abdominal ultrasonography

Song, Kyoung Doo

doi:10.14366/usg.20085

Cited by 14 publications

(13 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, DL has recently emerged as the leading machine learning tool in various research fields. It is one of the most popular artificial intelligence techniques used in the medical field, especially for image and video processing, thus numerous studies applying deep learning to ultrasound/CEUS imaging have been actively conducted [ 16 , 17 ]. The current section reviews CEUS based DL approaches for medical investigation, in general, and for FLL diagnosis, in particular.…”

Section: State Of the Artmentioning

confidence: 99%

Deep Neural Architectures for Contrast Enhanced Ultrasound (CEUS) Focal Liver Lesions Automated Diagnosis

Căleanu

Sîrbu

Simion

2021

Sensors

View full text Add to dashboard Cite

Computer vision, biomedical image processing and deep learning are related fields with a tremendous impact on the interpretation of medical images today. Among biomedical image sensing modalities, ultrasound (US) is one of the most widely used in practice, since it is noninvasive, accessible, and cheap. Its main drawback, compared to other imaging modalities, like computed tomography (CT) or magnetic resonance imaging (MRI), consists of the increased dependence on the human operator. One important step toward reducing this dependence is the implementation of a computer-aided diagnosis (CAD) system for US imaging. The aim of the paper is to examine the application of contrast enhanced ultrasound imaging (CEUS) to the problem of automated focal liver lesion (FLL) diagnosis using deep neural networks (DNN). Custom DNN designs are compared with state-of-the-art architectures, either pre-trained or trained from scratch. Our work improves on and broadens previous work in the field in several aspects, e.g., a novel leave-one-patient-out evaluation procedure, which further enabled us to formulate a hard-voting classification scheme. We show the effectiveness of our models, i.e., 88% accuracy reported against a higher number of liver lesion types: hepatocellular carcinomas (HCC), hypervascular metastases (HYPERM), hypovascular metastases (HYPOM), hemangiomas (HEM), and focal nodular hyperplasia (FNH).

show abstract

Section: State Of the Artmentioning

confidence: 99%

Deep Neural Architectures for Contrast Enhanced Ultrasound (CEUS) Focal Liver Lesions Automated Diagnosis

Căleanu

Sîrbu

Simion

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…At present, diagnostic imaging has considerable potential in the diagnosis of and intervention into abdominal pathologies, and was recently also supported by integration with artificial intelligence methods [36][37][38][39][40][41][42][43][44][45][46][47][48][49][50]. The suggestion of ischemic colitis based on imaging findings can permit early diagnosis and prompt therapy, which is essential for a successful outcome [51].…”

Section: Imagingmentioning

confidence: 99%

Large Bowel Ischemia/Infarction: How to Recognize It and Make Differential Diagnosis? A Review

Iacobellis

Narese

Berritto³

et al. 2021

Diagnostics

View full text Add to dashboard Cite

Ischemic colitis represents the most frequent form of intestinal ischemia occurring when there is an acute impairment or chronic reduction in the colonic blood supply, resulting in mucosal ulceration, inflammation, hemorrhage and ischemic necrosis of variable severity. The clinical presentation is variable and nonspecific, so it is often misdiagnosed. The most common etiology is hypoperfusion, almost always associated with generalized atherosclerotic disease. The severity ranges from localized and transient ischemia to transmural necrosis of the bowel wall, becoming a surgical emergency, with significant associated morbidity and mortality. The diagnosis is based on clinical, laboratory suspicion and radiological, endoscopic and histopathological findings. Among the radiological tests, enhanced-CT is the diagnostic investigation of choice. It allows us to make the diagnosis in an appropriate clinical setting, and to define the entity of the ischemia. MR may be adopted in the follow-up in patients with iodine allergy or renal dysfunctions, or younger patients who should avoid radiological exposure. In the majority of cases, supportive therapy is the only required treatment. In this article we review the pathophysiology and the imaging findings of ischemic colitis.

show abstract

“…Artificial intelligence, such as deep learning algorithms, is gaining extensive attention in medical imaging and studies applying deep learning to US are being actively conducted, mainly in the field of diffuse liver disease, especially hepatic fibrosis and steatosis evaluation[ 70 , 71 ]. In a recent study of 3446 patients a deep convolutional neural network showed high accuracy in the assessment of METAVIR score using US images and achieved better performance than that of radiologists in the diagnosis of cirrhosis[ 72 ].…”

Section: Artificial Intelligence and Radiomic Ultrasound Analysismentioning

confidence: 99%

New frontiers in liver ultrasound: From mono to multi parametricity

Bartolotta¹,

Taibbi²,

Randazzo³

et al. 2021

WJGO

View full text Add to dashboard Cite

Modern liver ultrasonography (US) has become a “one-stop shop” able to provide not only anatomic and morphologic but also functional information about vascularity, stiffness and other various liver tissue properties. Modern US techniques allow a quantitative assessment of various liver diseases. US scanning is no more limited to the visualized plane, but three-dimensional, volumetric acquisition and consequent post-processing are also possible. Further, US scan can be consistently merged and visualized in real time with Computed Tomography and Magnetic Resonance Imaging examinations. Effective and safe microbubble-based contrast agents allow a real time, dynamic study of contrast kinetic for the detection and characterization of focal liver lesions. Ultrasound can be used to guide loco-regional treatment of liver malignancies and to assess tumoral response either to interventional procedures or medical therapies. Microbubbles may also carry and deliver drugs under ultrasound exposure. US plays a crucial role in diagnosing, treating and monitoring focal and diffuse liver disease. On the basis of personal experience and literature data, this paper is aimed to review the main topics involving recent advances in the field of liver ultrasound.

show abstract

Current status of deep learning applications in abdominal ultrasonography

Cited by 14 publications

References 35 publications

Deep Neural Architectures for Contrast Enhanced Ultrasound (CEUS) Focal Liver Lesions Automated Diagnosis

Deep Neural Architectures for Contrast Enhanced Ultrasound (CEUS) Focal Liver Lesions Automated Diagnosis

Large Bowel Ischemia/Infarction: How to Recognize It and Make Differential Diagnosis? A Review

New frontiers in liver ultrasound: From mono to multi parametricity

Contact Info

Product

Resources

About