Tamoghna Chattopadhyay

The paper predicts how preprocessing affects fMRI connectivity estimates using prior scans, showing personalized models outperform average predictions.

The paper evaluates temporal complexity measures in resting-state fMRI to identify disease-related changes, finding that real data show greater outliers than randomized controls.

This study compares hidden Markov models (HMMs) and vector autoregressive (VAR) models in resting-state fMRI to determine if apparent state transitions reflect discrete dynamics or continuous activity, finding that VAR often outperforms HMMs in continuous cases.

The paper examines how dementia severity affects resting-state fMRI network responses to preprocessing, finding that greater severity correlates with reduced temporal concentration after nuisance regression.

The paper examines if temporal variability in brain connectivity improves prediction of two-year clinical severity in mild cognitive impairment compared to static connectivity measures.

This paper investigates why performance plateaus in 3D MRI transformers for brain disease classification and finds that representation collapse, not task difficulty, causes the plateau.

The paper uses polygenic scores from GWAS to identify spatial genetic effects on brain structure across multiple scales in MRI data.

This study examines how the spectral distribution of discriminative signals in MRI transformers affects their scaling behavior with increasing training data, finding that distributed representations improve more with larger datasets.

The paper maps the focal and generalized effects of common genetic variants on human brain structure using polygenic scores and MRI data.

This paper develops a lifespan normative model of brain white matter microstructure using diffusion MRI data from 54,583 individuals, enabling detection of microstructural deviations in diseases like Alzheimer's and 22q11.2 deletion syndrome.

The paper uses machine learning to predict autopsy-confirmed neuropathology from clinical, neuroimaging, and CSF biomarkers, finding that different modalities best predict specific pathologies.

This study compares deep learning methods to predict cognitive decline using brain MRI and clinical data over two years, showing the value of integrating image and tabular data.

This paper generates synthetic brain MRI data using a Diffusion Transformer to augment neuroimaging datasets and improve dementia classification.

This paper generates synthetic diffusion MRI scalar maps from T1-weighted images using RevGAN, showing they retain microstructural features and perform well in downstream tasks.

This study evaluates deep learning models that predict multiple dementia-related neuropathologies using MRI, clinical, and genetic data, showing the potential of multimodal AI for non-invasive profiling.

The paper generates synthetic diffusion MRI maps using diffusion models and evaluates their realism and utility for improving AI classifiers.

This study evaluates 2D and 3D diffusion models for generating synthetic DTI maps, finding 3D models perform better in downstream tasks like sex and dementia classification.

This study introduces MINT, a framework that jointly models white matter microstructure and macrostructure to better detect anomalies in MCI and dementia, showing improved sensitivity compared to traditional diffusion metrics.

This study compares explainable AI models for Alzheimer's disease classification using MRI scans, evaluating interpretability methods and generalization across different populations.

This paper introduces DNT, a variational autoencoder framework for detecting white matter abnormalities in Alzheimer's disease using macro- and microstructural data.

This paper evaluates CNN models for brain age prediction and dementia classification using diffusion MRI, showing improved performance after dataset harmonization across Indian and North American cohorts.

This paper harmonizes brain MRI data from multiple sources using 3D CycleGANs and shows improved accuracy in brain age prediction compared to raw data.

This study uses a variational autoencoder to generate synthetic tractography-based bundle templates from a small dataset, showing improved shape capture compared to standard atlases.

This study compares anatomical and diffusion MRI for Parkinson's disease detection using deep convolutional neural networks, finding that dMRI-based models show promise for classification.

This study evaluates how different defacing algorithms affect brain-age prediction using 3D CNNs, finding significant differences in accuracy based on defacing methods.

This paper uses a ConvVAE to reduce the dimensionality of white matter tractography data, identifies 6 tracts with significant differences between Alzheimer's patients and controls, and preserves distances in a 6-dimensional space.

This paper uses deep 3D CNNs combining anatomical and diffusion MRI to predict dementia severity and brain age, finding that diffusion MRI improves performance over anatomical MRI alone.

This paper predicts brain amyloid positivity from T1-weighted MRI and derived tissue maps using 3D CNNs and transfer learning, achieving 76% balanced accuracy.

FiberNeat filters false positive and implausible streamlines from tractograms using unsupervised methods, improving tract quality and tractometry.