adding paper

Author: manishsaggar1

images/papers/sptmsfmri.png

@@ -1,15 +1,15 @@
 ---
 layout: paper
-title: Predicting Task Activation Maps from Resting-State Functional Connectivity using Deep Learning
-image: /images/papers/rdoc-dl.png
-authors: Madsen S., Uddin L.Q., Mumford J.A., Barch D.M., Fair D.A., Gotlib I.H., Poldrack, R.A., Kuceyeski A., Saggar M.
+title: Concurrent single-pulse (sp) TMS/fMRI to reveal the causal connectome in healthy and patient populations
+image: /images/papers/sptmsfmri.png
+authors: Glick C., Gajawelli N., Sun Y., Badami F., Saggar M. Etkin A.
 year: 2024
-ref: Madsen S., Uddin L.Q., Mumford J.A., Barch D.M., Fair D.A., Gotlib I.H., Poldrack, R.A., Kuceyeski A., Saggar M. (2024) BioRxiv
+ref: Glick C., Gajawelli N., Sun Y., Badami F., Saggar M. Etkin A. (2024) BioRxiv
 journal: "BioRxiv"
-doi: 10.1101/2024.09.10.612309
-github: https://github.com/braindynamicslab/dl-task-contrast-prediction
-pdf: /pdfs/papers/rdoc-dl.pdf
+doi: 10.1101/2024.09.25.614833
+github: https://github.com/braindynamicslab/sptmsfmri
+pdf: /pdfs/papers/sptmsfmri.pdf
 ---
 
 # Abstract
-Recent work has shown that deep learning is a powerful tool for predicting brain activation patterns evoked through various tasks using resting state features. We replicate and improve upon this recent work to introduce two models, BrainSERF and BrainSurfGCN, that perform at least as well as the state-of-the-art while greatly reducing memory and computational footprints. Our performance analysis observed that low predictability was associated with a possible lack of task engagement derived from behavioral performance. Furthermore, a deficiency in model performance was also observed for closely matched task contrasts, likely due to high individual variability confirmed by low test-retest reliability. Overall, we successfully replicate recently developed deep learning architecture and provide scalable models for further research.
+Neuroimaging and cognitive neuroscience studies have identified neural circuits linked to anxiety, mood, and trauma-related symptoms and focused on their interaction with the medial prefrontal default mode circuitry. Despite these advances, developing new neuromodulatory treatments based on neurocircuitry remains challenging. It remains unclear which nodes within and controlling these circuits are affected and how their impairment is connected to psychiatric symptoms. Concurrent single-pulse (sp) TMS/fMRI offers a promising approach to probing and mapping the integrity of these circuits. In this study, we present concurrent sp-TMS/fMRI data along with structural MRI scans from 152 participants, including both healthy and depressed individuals. The sp-TMS was administered to 11 different cortical sites, providing a dataset that allows researchers to investigate how brain circuits are modulated by spTMS.

papers/_posts/2024-09-14-rdoc-dl.md

@@ -0,0 +1,15 @@
+---
+layout: paper
+title: Predicting Task Activation Maps from Resting-State Functional Connectivity using Deep Learning
+image: /images/papers/rdoc-dl.png
+authors: Madsen S., Uddin L.Q., Mumford J.A., Barch D.M., Fair D.A., Gotlib I.H., Poldrack, R.A., Kuceyeski A., Saggar M.
+year: 2024
+ref: Madsen S., Uddin L.Q., Mumford J.A., Barch D.M., Fair D.A., Gotlib I.H., Poldrack, R.A., Kuceyeski A., Saggar M. (2024) BioRxiv
+journal: "BioRxiv"
+doi: 10.1101/2024.09.10.612309
+github: https://github.com/braindynamicslab/dl-task-contrast-prediction
+pdf: /pdfs/papers/rdoc-dl.pdf
+---
+
+# Abstract
+Recent work has shown that deep learning is a powerful tool for predicting brain activation patterns evoked through various tasks using resting state features. We replicate and improve upon this recent work to introduce two models, BrainSERF and BrainSurfGCN, that perform at least as well as the state-of-the-art while greatly reducing memory and computational footprints. Our performance analysis observed that low predictability was associated with a possible lack of task engagement derived from behavioral performance. Furthermore, a deficiency in model performance was also observed for closely matched task contrasts, likely due to high individual variability confirmed by low test-retest reliability. Overall, we successfully replicate recently developed deep learning architecture and provide scalable models for further research.