Abstract
Transfer learning has been successfully used in the early diagnosis of Alzheimer’s disease (AD). In these methods, data from one single or multiple related source domain(s) are employed to aid the learning task in the target domain. However, most of the existing methods utilize data from all source domains, ignoring the fact that unrelated source domains may degrade the learning performance. Also, previous studies assume that class labels for all subjects are reliable, without considering the ambiguity of class labels caused by slight differences between early AD patients and normal control subjects. To address these issues, we propose to transform the original binary class label of a particular subject into a multi-bit label coding vector with the aid of multiple source domains. We further develop a robust multi-label transfer feature learning (rMLTFL) model to simultaneously capture a common set of features from different domains (including the target domain and all source domains) and to identify the unrelated source domains. We evaluate our method on 406 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database with baseline magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) data. The experimental results show that the proposed rMLTFL method can effectively improve the performance of AD diagnosis, compared with several state-of-the-art methods.
| Original language | English |
|---|---|
| Pages (from-to) | 1-16 |
| Number of pages | 16 |
| Journal | Brain Imaging and Behavior |
| DOIs | |
| Publication status | Accepted/In press - 2018 Mar 27 |
Fingerprint
Keywords
- Alzheimer’s disease (AD)
- Feature learning
- Multi-label learning
- Transfer learning
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging
- Neurology
- Cognitive Neuroscience
- Clinical Neurology
- Cellular and Molecular Neuroscience
- Psychiatry and Mental health
- Behavioral Neuroscience
Cite this
Robust multi-label transfer feature learning for early diagnosis of Alzheimer’s disease. / Alzheimer’s Disease Neuroimaging Initiative.
In: Brain Imaging and Behavior, 27.03.2018, p. 1-16.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Robust multi-label transfer feature learning for early diagnosis of Alzheimer’s disease
AU - Alzheimer’s Disease Neuroimaging Initiative
AU - Cheng, Bo
AU - Liu, Mingxia
AU - Zhang, Daoqiang
AU - Shen, Dinggang
PY - 2018/3/27
Y1 - 2018/3/27
N2 - Transfer learning has been successfully used in the early diagnosis of Alzheimer’s disease (AD). In these methods, data from one single or multiple related source domain(s) are employed to aid the learning task in the target domain. However, most of the existing methods utilize data from all source domains, ignoring the fact that unrelated source domains may degrade the learning performance. Also, previous studies assume that class labels for all subjects are reliable, without considering the ambiguity of class labels caused by slight differences between early AD patients and normal control subjects. To address these issues, we propose to transform the original binary class label of a particular subject into a multi-bit label coding vector with the aid of multiple source domains. We further develop a robust multi-label transfer feature learning (rMLTFL) model to simultaneously capture a common set of features from different domains (including the target domain and all source domains) and to identify the unrelated source domains. We evaluate our method on 406 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database with baseline magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) data. The experimental results show that the proposed rMLTFL method can effectively improve the performance of AD diagnosis, compared with several state-of-the-art methods.
AB - Transfer learning has been successfully used in the early diagnosis of Alzheimer’s disease (AD). In these methods, data from one single or multiple related source domain(s) are employed to aid the learning task in the target domain. However, most of the existing methods utilize data from all source domains, ignoring the fact that unrelated source domains may degrade the learning performance. Also, previous studies assume that class labels for all subjects are reliable, without considering the ambiguity of class labels caused by slight differences between early AD patients and normal control subjects. To address these issues, we propose to transform the original binary class label of a particular subject into a multi-bit label coding vector with the aid of multiple source domains. We further develop a robust multi-label transfer feature learning (rMLTFL) model to simultaneously capture a common set of features from different domains (including the target domain and all source domains) and to identify the unrelated source domains. We evaluate our method on 406 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database with baseline magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) data. The experimental results show that the proposed rMLTFL method can effectively improve the performance of AD diagnosis, compared with several state-of-the-art methods.
KW - Alzheimer’s disease (AD)
KW - Feature learning
KW - Multi-label learning
KW - Transfer learning
UR - http://www.scopus.com/inward/record.url?scp=85044462330&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85044462330&partnerID=8YFLogxK
U2 - 10.1007/s11682-018-9846-8
DO - 10.1007/s11682-018-9846-8
M3 - Article
C2 - 29589326
AN - SCOPUS:85044462330
SP - 1
EP - 16
JO - Brain Imaging and Behavior
JF - Brain Imaging and Behavior
SN - 1931-7557
ER -