期刊
ARTIFICIAL INTELLIGENCE REVIEW
卷 -, 期 -, 页码 -出版社
SPRINGER
DOI: 10.1007/s10462-023-10531-2
关键词
Cluster; Clustering; Traumatic brain injury; Acquired brain injury; Systematic literature review; Survey
The increasing number of people with traumatic brain injury (TBI) presents a challenge for practitioners in designing the rehabilitation process due to their multiple deficits. This study focuses on analyzing the clustering algorithms used to group TBI patients and aims to determine the purposes, deficits, algorithms, features, data pre-processing techniques, parameters, and efficiency/effectiveness achieved by these algorithms.
While the number of people suffering from traumatic brain injury (TBI) has increased considerably in recent years, the multiple deficits of these patients makes designing the rehabilitation process a challenge for practitioners. They need to group similar patients, due to their features and/ or diseases in order to assign them to the same clinically significant group to facilitate the design of appropriate rehabilitation activities. The information used to group the patients depends on the type of patient as well as the possible groups to be formed. This work focuses on studying how grouping patients with TBI has been carried out so far by means of clustering algorithms. The main interest in grouping TBI patients is the need to address this heterogeneity to create clinical guidelines or rehabilitation activities for individual groups and detect the characteristic features of each group. This study's main aims are: (1) to determine the purposes of the clustering algorithms developed for TBI patients, (2) to identify the normally considered deficits, (3) to determine the most commonly used clustering algorithms, (4) to identify the types of features usually employed for TBI clustering, (5) to analyse the data pre-processing techniques applied, (5) to identify the parameters chosen when running a clustering algorithm for TBI patients, and (6) to determine the efficiency/effectiveness achieved by clustering algorithms.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据