Deep learning-based defect detection in film-coated tablets using a convolutional neural network

International Journal of Pharmaceutics, Journal Year: 2025, Volume and Issue: unknown, P. 125220 - 125220

Published: Jan. 1, 2025

Language: Английский

---

Pneumonia Disease Detection Using Chest X-Rays and Machine Learning

Algorithms, Journal Year: 2025, Volume and Issue: 18(2), P. 82 - 82

Published: Feb. 3, 2025

Pneumonia is a deadly disease affecting millions worldwide, caused by microorganisms and environmental factors. It leads to lung fluid build-up, making breathing difficult, leading cause of death. Early detection treatment are crucial for preventing severe outcomes. Chest X-rays commonly used diagnoses due their accessibility low costs; however, detecting pneumonia through challenging. Automated methods needed, machine learning can solve complex computer vision problems in medical imaging. This research develops robust model the early using chest X-rays, leveraging advanced image processing techniques deep algorithms that accurately identify patterns, enabling prompt diagnosis treatment. The CNN from ground up ResNet-50 pretrained study uses RSNA challenge original dataset comprising 26,684 array images collected unique patients (56% male, 44% females) build pneumonia. data made (31.6%) non-pneumonia (68.8%), providing an effective foundation training evaluation. A reduced size was examine impact both versions were tested with without use augmentation. models compared existing works, model’s effectiveness one another, augmentation on performance examined. overall best accuracy achieved scratch, no augmentation, 0.79, precision 0.76, recall 0.73, F1 score 0.74. However, model, lower accuracy, found be more generalizable.

Language: Английский

---

COVID-19 detection from lung CT-Scans using a fuzzy integral-based CNN ensemble

Computers in Biology and Medicine, Journal Year: 2021, Volume and Issue: 138, P. 104895 - 104895

Published: Oct. 1, 2021

Language: Английский

---

Deep Ensemble Learning for the Automatic Detection of Pneumoconiosis in Coal Worker’s Chest X-ray Radiography

Journal of Clinical Medicine, Journal Year: 2022, Volume and Issue: 11(18), P. 5342 - 5342

Published: Sept. 12, 2022

Globally, coal remains one of the natural resources that provide power to world. Thousands people are involved in collection, processing, and transportation. Particulate dust is produced during these processes, which can crush lung structure workers cause pneumoconiosis. There no automated system for detecting monitoring diseases miners, except specialist radiologists. This paper proposes ensemble learning techniques pneumoconiosis disease chest X-ray radiographs (CXRs) using multiple deep models. Three (simple averaging, multi-weighted majority voting (MVOT)) were proposed investigate performances randomised cross-folds leave-one-out cross-validations datasets. Five statistical measurements used compare outcomes three investigations on integrated approach with state-of-the-art approaches from literature same dataset. In second investigation, combination was marginally enhanced averaging a robust model, CheXNet. However, third model elevated accuracies 87.80 90.2%. The investigated results helped us identify framework outperformed others, achieving an accuracy 91.50% detection

Language: Английский

---

Brain Tumor Classification based on Improved Stacked Ensemble Deep Learning Methods

Asian Pacific Journal of Cancer Prevention, Journal Year: 2023, Volume and Issue: 24(6), P. 2141 - 2148

Published: June 1, 2023

Brain Tumor diagnostic prediction is essential for assisting radiologists and other healthcare professionals in identifying classifying brain tumors. For the diagnosis treatment of cancer diseases, classification accuracy are crucial. The aim this study was to improve ensemble deep learning models classifing tumor increase performance structure by combining different model develop a with more accurate predictions than individual models.Convolutional neural networks (CNNs), which made up single algorithm called CNN model, foundation most current methods illness images. combined create method. However, compared machine algorithm, accurate. This used stacked technology. data set obtained from Kaggle included two categories: abnormal & normal brains. trained three models: VGG19, Inception v3, Resnet 10.The 96.6% binary (0,1) have been achieved Loss cross entropy, Adam optimizer take into consideration stacking models.The can be improved over framework.

Language: Английский

---