RH-BrainFS: Regional Heterogeneous Multimodal Brain Networks Fusion Strategy
Authors: Hongting Ye, Yalu Zheng, Yueying Li, Ke Zhang, Youyong Kong, Yonggui Yuan
NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we perform a series of experiments to evaluate the effectiveness of the proposed RH-Brain FS method. First, we provide the detailed experimental settings in Sec. 4.1. Then, we perform comparison experiments on all datasets to compare the performance of different methods in Sec. 4.2. Finally, we perform some ablation studies of the main modules and hyperparameters in the proposed RH-Brain FS method in Sec. 4.3. |
| Researcher Affiliation | Academia | Hongting Ye 1, Yalu Zheng 1, Yueying Li 1, Ke Zhang 1, Youyong Kong 1,2 , Yonggui Yuan 3 1Jiangsu Provincial Joint International Research Laboratory of Medical Information Processing School of Computer Science and Engineering, Southeast University 2Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education, China 3Department of Psychosomatics and Psychiatry, Zhongda Hospital School of Medicine, Southeast University {yehongting, 220212084, 230228504, kylenz, kongyouyong}@seu.edu.cn yygylh2000@sina.com |
| Pseudocode | No | The paper describes the model architecture and processes in detail but does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | 2The codes are available at https://github.com/Yedaxia1/RH-Brain FS. |
| Open Datasets | Yes | We evaluate our RH-Brain FS method on two different classification tasks investigating structure-function fusion. 1) The gender classification task on Human Connectome Project (HCP) dataset [43], which contains 560 female samples and 479 male samples. 2) The Major Depressive Disorder (MDD) diagnosis task on the hospital datasets [17, 18], including the Affiliated Zhongda Hospital of Southeast University (Zhongda hospital) and the Second Affiliated Hospital of Xinxiang Medical University (Xinxiang hospital). |
| Dataset Splits | Yes | In this study, we evaluate all the methods using 10-fold cross-validation with the same partition of training and testing splits. |
| Hardware Specification | Yes | All our experiments are implemented in Py Torch and trained on one NVIDIA 3090. |
| Software Dependencies | No | The paper mentions that experiments are 'implemented in Py Torch' but does not specify the version number of Py Torch or any other software dependencies. |
| Experiment Setup | Yes | The initial learning rate is set to 5e-4 and the dropout rate is set to 0.3. Also we utilize a early stop mechanism that 300 epochs patience in total 500 epochs. In the RH-Brain FS model, we set the k-hop in the subgraph sampling to 1, the number of bottlenecks Nb to 4, the number of attention heads in the Transformer to 4, and the total number of network layers to 2. |