Gastrointestinal tract disease detection via deep learning based Duo-Feature Optimized Hexa-Classification model

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 100, P. 106994 - 106994

Published: Oct. 12, 2024

Language: Английский

---

Endoscopic diagnosis and treatment for portal hypertension: not yet ready for clinical practice!

Journal of Hepatology, Journal Year: 2024, Volume and Issue: 81(4), P. e181 - e182

Published: Feb. 7, 2024

Language: Английский

---

Gastrointestinal tract disease detection via deep learning based structural and statistical features optimized hexa-classification model

Technology and Health Care, Journal Year: 2024, Volume and Issue: 32(6), P. 4453 - 4473

Published: July 19, 2024

BACKGROUND: Gastrointestinal tract (GIT) diseases impact the entire digestive system, spanning from mouth to anus. Wireless Capsule Endoscopy (WCE) stands out as an effective analytic instrument for diseases. Nevertheless, accurately identifying various lesion features, such irregular sizes, shapes, colors, and textures, remains challenging in this field. OBJECTIVE: Several computer vision algorithms have been introduced tackle these challenges, but many relied on handcrafted resulting inaccuracies instances. METHODS: In work, a novel Deep SS-Hexa model is proposed which combination two different deep learning structures extracting features WCE images detect GIT ailment. The gathered are denoised by weighted median filter remove noisy distortions augment enhancing training data. structural statistical (SS) feature extraction process sectioned into phases analysis of distinct regions gastrointestinal. first stage, image retrieved using MobileNet with support SiLU activation function retrieve relevant features. second phase, segmented intestine transformed learn local information. These SS parallelly fused selecting best walrus optimization algorithm. Finally, belief network (DBN) used classified hexa classes namely normal, ulcer, pylorus, cecum, esophagitis polyps basis selected RESULTS: attains overall average accuracy 99.16% disease detection based KVASIR KID datasets. achieves high level minimal computational cost recognition illness. CONCLUSIONS: Model progresses range 0.04%, 0.80% better than GastroVision, Genetic algorithm dataset 0.60%, 1.21% Modified U-Net, WCENet respectively.

Language: Английский

---

EUS-Guided Vascular Interventions: Recent Advances

Journal of Clinical Medicine, Journal Year: 2024, Volume and Issue: 13(16), P. 4835 - 4835

Published: Aug. 16, 2024

Endoscopic ultrasound (EUS)-guided vascular interventions were first reported in 2000 a study that evaluated the utility of EUS sclerotherapy esophageal varices. Currently, gastric variceal therapy and portosystemic pressure gradient (PPG) measurements are most widely utilized applications. Ectopic obliteration, splenic artery embolization, aneurysm/pseudoaneurysm treatment, portal venous sampling, shunt creation using some other emerging interventions. Since release American Gastroenterological Association (AGA)'s commentary 2023, which primarily endorses EUS-guided EUS-PPG measurement, several new studies have been published supporting use for various conditions. In this review, we present recent advances field, critically appraising trials.

Language: Английский

---

Progress in the Study of Surgical Methods of Liver Biopsy

Advances in Clinical Medicine, Journal Year: 2024, Volume and Issue: 14(12), P. 1192 - 1199

Published: Jan. 1, 2024

Language: Английский

---