Detrended Partial Cross Correlation for Brain Connectivity Analysis
Authors: Jaime Ide, Fábio Cappabianco, Fabio Faria, Chiang-shan R. Li
NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We use a rich simulated f MRI dataset to validate the proposed method, and apply it to a real f MRI dataset in a cocaine dependence prediction task. We show that, compared to extant methods, the DPCCA-based approach not only distinguishes short and long memory functional connectivity but also improves feature extraction and enhances classification accuracy. |
| Researcher Affiliation | Academia | Jaime S Ide Yale University New Haven, CT 06519 jaime.ide@yale.edu Fabio A Cappabianco Federal University of Sao Paulo S.J. dos Campos, 12231, Brazil cappabianco@unifesp.br Fabio A Faria Federal University of Sao Paulo S.J. dos Campos, 12231, Brazil ffaria@unifesp.br Chiang-shan R Li Yale University New Haven, CT chiang-shan.li-yale.edu |
| Pseudocode | Yes | Algorithm 1 DPCCA+CCA Input: Time series {Xt IRm, t = 1, 2, ..., N}, where m is the number of vectors and N is the number of time points; time range srange with k values Output: Connectivity matrix F C : [m m] and associated matrices |
| Open Source Code | No | The paper does not provide any explicit statement about releasing its own source code or a link to a repository for the methodology described. |
| Open Datasets | Yes | We use f MRI simulation data Net Sim [20] previously developed for the evaluation of network modeling methods. |
| Dataset Splits | Yes | The classification task consists of predicting the class membership, cocaine dependence (CD) and healthy control (HC), given each individual s f MRI data. After initial preprocessing (Section 2.2), we extract average time series within the frontoparietal circuit of 6 regions 3 (Figure 1), and compute the different cross-correlation measures. These coefficients are used as features to train and test (leave-one-out cross-validation) a set of popular classifiers available in scikit-learn toolbox [16] (version 0.18.1), |
| Hardware Specification | No | The paper does not provide specific details about the computational hardware (e.g., CPU, GPU models, or memory) used for running the experiments. It only mentions the fMRI scanner used for data collection. |
| Software Dependencies | Yes | Functional MRI data was pre-processed with standard pipeline using Statistical Parametric Mapping 12 (SPM12) (Wellcome Department of Imaging Neuroscience, University College London, U.K.). |
| Experiment Setup | Yes | For the DPCCA coefficients, we test both peak values DPCCAmax as well as the rich temporal profiles DPCCAF ull. Finally, we also include the brain activation maps (Section 2.2.1) as feature set, thus allowing comparison with popular f MRI classification softwares such as PRONTO (http://www.mlnl.cs.ucl.ac.uk/pronto/). Features are summarized in Table 2. ... DPCCAIso 135-180 Isomap with 9-12 components and 30 neighbors DPCCAAuto E 30-45 Autoencoders with 2-3 hidden layers, 5-20 neurons, batch=100, epoch=1000 |