.png){kind=icon}
.png){kind=icon}
Return
Inflammatory Bowel Disease Detection Using Machine Learning Techniques
.png){kind=icon}
PRE
AI
DOI：10.34028/iajit/23/2/14.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Crohn's Disease (CD) is an Inflammatory Bowel Disease (IBD) has seen a sharp rise around 50% worldwide. Therefore, researchers started looking for alternative ways and started applying computation-based deep learning algorithms. We proposed Machine Learning (ML) and Deep Learning (DL) techniques for identification of complex patterns present in the DNA sequences with a primary goal to improve the accuracy. The current study presents key findings of the performance of a variety of ML models decision tree, Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), Convolutional Neural Networks (CNN), CNN+LSTM, CNN+BiLSTM hybrids, resent, Multi-Layer Perceptron (MLP), Gated Recurrent Units (GRU), Transformer-based models, auto encoder with Feedforward Neural Network and other models for the classification of the disease. We analyzed a gene expression dataset obtained from the NCBI. Each model is evaluated based on accuracy, Area Under the Curve-Receiver Operating Characteristic (AUC-ROC), precision, and kappa. The experimental results are compared with state-of-the-art approaches from the existing literature, demonstrating the effectiveness of the proposed model. Among all evaluated methods, the proposed Sequence Read Archive (EMAT) model achieves the highest performance, attaining an accuracy of 88.12%. The comparative analysis confirms that EMAT stands the best-performing model setting a new benchmark.
Keywords:
Bowel disease
crohn's disease
machine learning
Journal
I
IF:
1.1
Papers:
45
Citations:
815