Publications by authors named "Heba M Afify"
Article Synopsis
- The paper discusses the need for improved diagnostic approaches for ear diseases, highlighting the limitations of traditional clinical diagnoses by otolaryngologists.
- It proposes using convolutional neural networks (CNNs) enhanced with Bayesian hyperparameter optimization to analyze otoscopic images classified into four types: normal, myringosclerosis, earwax plug, and chronic otitis media.
- The proposed method achieved high performance metrics with an accuracy of 98.10%, sensitivity of 98.11%, specificity of 99.36%, and positive predictive value of 98.10%, aiming to create an automated tool for better detection and classification of ear diseases.
The external and middle ear conditions are diagnosed using a digital otoscope. The clinical diagnosis of ear conditions is suffered from restricted accuracy due to the increased dependency on otolaryngologist expertise, patient complaint, blurring of the otoscopic images, and complexity of lesions definition. There is a high requirement for improved diagnosis algorithms based on otoscopic image processing.
View Article and Find Full Text PDFComputational predictions for protein sequences of COVID-19 virus via machine learning algorithms.
Med Biol Eng Comput
September 2021
The rapid spread of coronavirus disease (COVID-19) has become a worldwide pandemic and affected more than 15 million patients reported in 27 countries. Therefore, the computational biology carrying this virus that correlates with the human population urgently needs to be understood. In this paper, the classification of the human protein sequences of COVID-19, according to the country, is presented based on machine learning algorithms.
View Article and Find Full Text PDFArticle Synopsis
- - Lung segmentation in chest CT scans is crucial for identifying lung cancer and involves separating tumors from lung tissue, often utilizing advanced algorithms like fully convolutional networks (FCNs) for accuracy.
- - The study applied a V.Net-inspired FCN to a lung CT cancer scan database, conducting experiments with 64 training and 32 testing images, which included steps of data preprocessing, data augmentation, and neural network implementation.
- - The proposed system achieved an impressive average Dice score coefficient of 80% for tumor regions of interest and 98% for surrounding lung tissues, demonstrating its effectiveness in 3D lung segmentation and tumor estimation.