Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification
Three-dimensional convolutional neural networks (3D CNNs) have been widely applied to analyze Alzheimer’s disease (AD) brain images for a better understanding of the disease progress or predicting the conversion from cognitively impaired (CU) or mild cognitive impairment status. It is well-known tha...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/12/2/467 |
| _version_ | 1797443387309162496 |
|---|---|
| author | Xin Xing Muhammad Usman Rafique Gongbo Liang Hunter Blanton Yu Zhang Chris Wang Nathan Jacobs Ai-Ling Lin |
| author_facet | Xin Xing Muhammad Usman Rafique Gongbo Liang Hunter Blanton Yu Zhang Chris Wang Nathan Jacobs Ai-Ling Lin |
| author_sort | Xin Xing |
| collection | DOAJ |
| description | Three-dimensional convolutional neural networks (3D CNNs) have been widely applied to analyze Alzheimer’s disease (AD) brain images for a better understanding of the disease progress or predicting the conversion from cognitively impaired (CU) or mild cognitive impairment status. It is well-known that training 3D-CNN is computationally expensive and with the potential of overfitting due to the small sample size available in the medical imaging field. Here we proposed a novel 3D-2D approach by converting a 3D brain image to a 2D fused image using a Learnable Weighted Pooling (LWP) method to improve efficient training and maintain comparable model performance. By the 3D-to-2D conversion, the proposed model can easily forward the fused 2D image through a pre-trained 2D model while achieving better performance over different 3D and 2D baselines. In the implementation, we chose to use ResNet34 for feature extraction as it outperformed other 2D CNN backbones. We further showed that the weights of the slices are location-dependent and the model performance relies on the 3D-to-2D fusion view, with the best outcomes from the coronal view. With the new approach, we were able to reduce 75% of the training time and increase the accuracy to 0.88, compared with conventional 3D CNNs, for classifying amyloid-beta PET imaging from the AD patients from the CU participants using the publicly available Alzheimer’s Disease Neuroimaging Initiative dataset. The novel 3D-2D model may have profound implications for timely AD diagnosis in clinical settings in the future. |
| first_indexed | 2024-03-09T12:55:18Z |
| format | Article |
| id | doaj.art-d0f00ac1cda0410ba96185ca59f255e9 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-09T12:55:18Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-d0f00ac1cda0410ba96185ca59f255e92023-11-30T22:00:52ZengMDPI AGElectronics2079-92922023-01-0112246710.3390/electronics12020467Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image ClassificationXin Xing0Muhammad Usman Rafique1Gongbo Liang2Hunter Blanton3Yu Zhang4Chris Wang5Nathan Jacobs6Ai-Ling Lin7Department of Computer Science, University of Kentucky, Lexington, KY 40506, USAKitware Inc., Clifton Park, NY 12065, USADepartment of Computing and Cyber Security, Texas A & M University-San Antonio, San Antonio, TX 78224, USADepartment of Computer Science, University of Kentucky, Lexington, KY 40506, USADepartment of Computer Science, University of Kentucky, Lexington, KY 40506, USADepartment of Computer Science, University of Missouri, Columbia, MO 65211, USADepartment of Computer Science & Engineering, Washington University in St. Louis, St. Louis, MO 63130, USADepartment of Radiology, University of Missouri, Columbia, MO 65212, USAThree-dimensional convolutional neural networks (3D CNNs) have been widely applied to analyze Alzheimer’s disease (AD) brain images for a better understanding of the disease progress or predicting the conversion from cognitively impaired (CU) or mild cognitive impairment status. It is well-known that training 3D-CNN is computationally expensive and with the potential of overfitting due to the small sample size available in the medical imaging field. Here we proposed a novel 3D-2D approach by converting a 3D brain image to a 2D fused image using a Learnable Weighted Pooling (LWP) method to improve efficient training and maintain comparable model performance. By the 3D-to-2D conversion, the proposed model can easily forward the fused 2D image through a pre-trained 2D model while achieving better performance over different 3D and 2D baselines. In the implementation, we chose to use ResNet34 for feature extraction as it outperformed other 2D CNN backbones. We further showed that the weights of the slices are location-dependent and the model performance relies on the 3D-to-2D fusion view, with the best outcomes from the coronal view. With the new approach, we were able to reduce 75% of the training time and increase the accuracy to 0.88, compared with conventional 3D CNNs, for classifying amyloid-beta PET imaging from the AD patients from the CU participants using the publicly available Alzheimer’s Disease Neuroimaging Initiative dataset. The novel 3D-2D model may have profound implications for timely AD diagnosis in clinical settings in the future.https://www.mdpi.com/2079-9292/12/2/467efficient trainingAlzheimer’s diseasedeep learning3D-to-2D module |
| spellingShingle | Xin Xing Muhammad Usman Rafique Gongbo Liang Hunter Blanton Yu Zhang Chris Wang Nathan Jacobs Ai-Ling Lin Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification Electronics efficient training Alzheimer’s disease deep learning 3D-to-2D module |
| title | Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification |
| title_full | Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification |
| title_fullStr | Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification |
| title_full_unstemmed | Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification |
| title_short | Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification |
| title_sort | efficient training on alzheimer s disease diagnosis with learnable weighted pooling for 3d pet brain image classification |
| topic | efficient training Alzheimer’s disease deep learning 3D-to-2D module |
| url | https://www.mdpi.com/2079-9292/12/2/467 |
| work_keys_str_mv | AT xinxing efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT muhammadusmanrafique efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT gongboliang efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT hunterblanton efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT yuzhang efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT chriswang efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT nathanjacobs efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification AT ailinglin efficienttrainingonalzheimersdiseasediagnosiswithlearnableweightedpoolingfor3dpetbrainimageclassification |