综合神经网络和进化算法方法用于肝纤维化病阶段:人工智能可以降低患者成本吗?
Ali Nazarizadeh1, Touraj Banirostam1, Taraneh Biglari1
1Department of Computer Engineering Central Tehran Branch, Islamic Azad University Tehran Iran.
概括
这项研究引入了一种新的人工神经网络 (ANN) 方法,使用基于学习的教学优化 (TLBO) 算法来预测肝纤维化阶段. 优化的模型以更少的患者特征实现了高精度,简化了诊断.
科学领域:
- 医疗信息学 医疗信息学
- 机器学习 机器学习
- 肝病学 肝病学是一种肝病学.
背景情况:
- 准确的肝纤维化阶段确定对于患者的管理至关重要.
- 肝脏活检,目前的黄金标准,是侵入性的.
- 预测纤维化的非侵入性方法非常受欢迎.
研究的目的:
- 开发和评估一个人工神经网络 (ANN) 模型与基于教学优化 (TLBO) 算法集成.
- 在献血者和型肝炎患者中预测肝纤维化的阶段.
- 为了减少准确的纤维化预测所需的输入特征的数量.
主要方法:
- 使用的机器学习分类方法:多层感知子 (MLP),天真贝叶斯式 (NB),决策树和深度学习.
- 应用合成少数群体过量采样技术 (SMOTE) 来处理数据集不平衡.
- 集成的MLP与TLBO算法进行优化特征选择和预测.
主要成果:
- 使用TLBO提出的MLP模型仅使用7个特征,实现了0.891的诊断准确度.
- 12个特征的标准MLP的准确性为0.903.
- SMOTE应用程序改善了大多数方法的诊断准确性,不包括贝叶斯网络模型.
结论:
- 针对TLBO优化的MLP模型提供了一种简单的方法,具有较少的特征要求和与更复杂的方法相比的准确性.
- 基于决策树的深度学习模型显示了12个特征的最高准确性.
- 拟议的方法表明了准确,不那么侵入性肝纤维化病阶段的潜力.
更多相关视频
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
8.2K
08:41Novel In Vivo Micro-Computed Tomography Imaging Techniques for Assessing the Progression of Non-Alcoholic Fatty Liver Disease
Published on: March 24, 2023
1.1K
相关概念视频
Ultrasound II: Endoscopic Ultrasound and FibroScan
101
Endoscopic Ultrasound (EUS) and FibroScan are valuable diagnostic tools in gastroenterology and hepatology, each with specific applications and techniques.
Endoscopic Ultrasound (EUS):
Endoscopic Ultrasound (EUS):
101
Liver Regeneration
3.3K
The liver is an important organ in vertebrates that plays an essential role in metabolism. It is also responsible for storing and redistributing nutrients such as carbohydrates, fats, and vitamins in the body. Additionally, the liver releases bile salts which are critical for digesting food and eliminating toxic metabolites from the body.
Cells of Liver
The liver comprises four major types of cells— hepatocytes, stellate, Kupffer, and sinusoidal endothelial cells. The hepatocytes are...
Cells of Liver
The liver comprises four major types of cells— hepatocytes, stellate, Kupffer, and sinusoidal endothelial cells. The hepatocytes are...
3.3K
