Enhanced glaucoma detection using U-Net and U-Net+ architectures using deep learning techniques DOI Creative Commons
B. P. Pradeep Kumar, Pramod Rangaiah, Robin Augustine

и другие.

Photodiagnosis and Photodynamic Therapy, Год журнала: 2025, Номер unknown, С. 104621 - 104621

Опубликована: Июнь 1, 2025

This study compares multiple image processing and deep learning methods to demonstrate an enhanced approach glaucoma diagnosis. The focuses on noise reduction using median filtering optic disc segmentation utilizing the U-Net U-Net+ architectures. Capsule Networks were utilized for feature extraction Extreme Learning Machines (ELM) diagnostic classification. Three datasets evaluated, including DRISHTI-GS, DRIONS-DB, HRF, important parameters such as accuracy, sensitivity, specificity. findings revealed that reduced by 97.88%, with a peak signal-to-noise ratio of 44.99. beat in process Dice coefficient 0.8557, Jaccard index 0.7307, higher accuracy. suggested model has great scoring 99% 99.5% 98.5% HRF. These show approaches can increase diagnosis accuracy reliability, implications healthcare applications patient outcomes.

Язык: Английский

Histopathology-driven prostate cancer identification: A VBIR approach with CLAHE and GLCM insights DOI Creative Commons
Pramod Rangaiah, B. P. Pradeep Kumar, Robin Augustine

и другие.

Computers in Biology and Medicine, Год журнала: 2024, Номер 182, С. 109213 - 109213

Опубликована: Окт. 2, 2024

Язык: Английский

Процитировано

4
Enhanced perceptual wavelet packet features for spontaneous Kannada sentence recognition under uncontrolled conditions DOI
Mahadevaswamy Shanthamallappa, B. P. Pradeep Kumar

International Journal of Speech Technology, Год журнала: 2025, Номер unknown

Опубликована: Фев. 25, 2025

Язык: Английский

Процитировано

0
The Burn Grafting Image Reclamation Redefined with the Peak-Valley Approach DOI
B. P. Pradeep Kumar,

E. Naresh,

A. Ashwitha

и другие.

Critical Reviews in Biomedical Engineering, Год журнала: 2025, Номер 53(2), С. 21 - 35

Опубликована: Янв. 1, 2025

Burn injuries constitute a significant public health challenge, often necessitating the expertise of medical professionals for diagnosis. However, in scenarios where specialized facilities are unavailable, utility automated burn assessment tools becomes evident. Factors such as area, depth, and location play pivotal role determining severity. In this study, we present classification model diagnosis, leveraging machine learning techniques. Our approach includes an image reclamation system that incorporates peak valley algorithm, ensuring removal noise while consistently delivering high-quality results. By using skewness kurtosis, demonstrate substantial improvements diagnostic accuracy. proposed sources key features from enhanced grafting samples transformation, enabling computation BQs unique bin analysis to enhance reclamation. experimental results highlight efficiency gains, notably growing matching graft 14 images. The intended work involves creation model. utilizes support vector (SVM). evaluation will be conducted untrained catalogue, with specific focus on its effectiveness reclaiming images necessitate grafts distinguishing them those do not. holds promise sample emergency settings, thereby expediting more accurate diagnoses treatments acute injuries. This has latent save lives improve patient upshots traumas.

Язык: Английский

Процитировано

0
Enhanced glaucoma detection using U-Net and U-Net+ architectures using deep learning techniques DOI Creative Commons
B. P. Pradeep Kumar, Pramod Rangaiah, Robin Augustine

и другие.

Photodiagnosis and Photodynamic Therapy, Год журнала: 2025, Номер unknown, С. 104621 - 104621

Опубликована: Июнь 1, 2025

This study compares multiple image processing and deep learning methods to demonstrate an enhanced approach glaucoma diagnosis. The focuses on noise reduction using median filtering optic disc segmentation utilizing the U-Net U-Net+ architectures. Capsule Networks were utilized for feature extraction Extreme Learning Machines (ELM) diagnostic classification. Three datasets evaluated, including DRISHTI-GS, DRIONS-DB, HRF, important parameters such as accuracy, sensitivity, specificity. findings revealed that reduced by 97.88%, with a peak signal-to-noise ratio of 44.99. beat in process Dice coefficient 0.8557, Jaccard index 0.7307, higher accuracy. suggested model has great scoring 99% 99.5% 98.5% HRF. These show approaches can increase diagnosis accuracy reliability, implications healthcare applications patient outcomes.

Язык: Английский

Процитировано

0