MANAS Lab
Welcome to MANAS (Models, Algorithms, Neuroscience, Affect, pSychiatry) Lab. Our research interests are broadly
- To use computational models and tools to understand brain function and behavior, particularly in mood and anxiety disorders.
- To use these insights, to design better experiments and treatments, and thus complete the loop.
Team
Ryan Thomas Philips (PI)
I am an Assitant Professor at the School of Arts and Sciences, Azim Premji University (APU) and a computational neuroscientist by training. I am interested in designing theoretical-model-inspired experiments that probe brain function, particularly in mood and anxiety disorders. I completed my doctoral studies at the Indian Institute of Technology Madras (IITM), with Prof. V. Srinivasa Chakravarthy as my thesis guide. My postdoctoral training was at the Section on Neurobiology of Fear & Anxiety, at the National Institute of Mental Health (NIMH/NIH) with Dr. Christian Grillon & Dr. Monique Ernst as my mentors. My current projects include: exploration-exploitation under anxiety; functional connectivity of BNST & CEA using resting state 7T fMRI; the role of mental fatigue in threat biases; investigating valence and uncertainty circuits using task based fMRI.
Research
Key Intellectual Question:
‘How does anxiety impact decision-making and metacognitive parameters, and what is the relationshing between these parameters and physiological and fMRI neural signals?’
Directions and Approach
With this key question in mind, the lab has 2 major directions. The first direction is to build a robust computation-driven understanding of decision making and metacognition in anxiety, specifically, by isolating the influence of independent parameters. Fear and anxiety are normal adaptive responses to threat. Anxiety is considered pathological when it is either excessive or inappropriate to the context. The influence of anxiety goes beyond the subjective feelings and autonomic reactivity. Indeed, anxiety powerfully affects motivated behavior. Most experimental tasks, used to probe behavior in anxiogenic environments or in response to aversive cues, usually focus on only one aspect of behavior. For example, previous studies have targeted cognitive constructs like working memory (Balderston et al., 2016; Balderston et al., 2017), response inhibition (Grillon et al., 2017), or attention. Each of these cognitive processes contributes to decision making, the backbone of motivated behavior. I would like to focus on a holistic approach that evaluates many decision-making facets. Our approach would be to study and model induced anxiety in both healthy volunteers and anxiety patients and examine an array of parameters such as learning rate, exploration vs exploitation parameter, inverse temperature and loss aversion/gain seeking.
The second related direction is understanding brain networks involved in fear and anxiety, specifically BNST and CeA related fMRI circuits. The Bed Nucleus of the Stria Terminalis (BNST or BST) and the Central Amygdala (CeA) are two interconnected sub-regions of the central extended amygdala that have been associated with fear and anxiety. Initial basic studies in rodents suggested that the BNST mediates anxiety-like responses to sustained threats, whereas the CeA mediates fear-like responses to more acute threats (Davis et al., 2010). Lesions of or pharmacological interventions on the BNST affect ‘anxiety’ responses whereas those to the CeA, affect ‘fear’ responses (Davis et al., 2010). A similar dissociation between the functional role of the CeA and BNST has also been proposed in humans based on pharmacological, clinical, and neuroimaging studies. However, recently this distinction has been challenged. In humans, a few translational studies have shown no differential activation of the CeA and the BNST to discrete and diffused cues. These findings have been bolstered by a meta-analysis that did not support the CeA-BNST dissociation (Chavanne & Robinson, 2021). Our approach would be to analyze resting state and task-based fMRI data to investigate these circuits, using a dynamic time warping machine- learning pipeline I have recently developed (Philips et al., 2021).
Importance and Consequences
There is much need to pursue anxiety-related research in an India specific context. On one hand, in 2017, 197·3 million people had mental disorders in India, including 45·7 million with depressive disorders and 44·9 million with anxiety disorders (Sagar et al., 2020). This makes anxiety disorders one of the most prevalent and debilitating psychiatric disorders, with a mental disorder crude disability-adjusted life-years (DALY) of 19%. On the other hand, there is a deficit in mental health professionals. It is estimated that there are 0.75 Psychiatrists per 100,000 people, while the desirable number is above 3 (Garg et al. 2019). A computation-driven understanding of anxiety disorders and underlying neural circuits would aide efforts toward precision psychiatry, digital health initiatives, novel cognitive therapies, and innovative drug treatment strategies. My goal is to advance along this path in my future research career as a principal investigator.
The overall end goal of this research is to open new avenues for understanding decision-making perturbations in anxiety and to provide a basis for the optimization and development of novel treatment strategies. Specifically, understanding individual differences in the decision-making parameters could give clinicians: 1) the tools to better understand anxiety subtypes and provide actionable biomarkers, 2) the ability to move beyond a population level, trial-and-error approach to treatment.
Publications
Gorka, Adam X; Philips, Ryan T.; Torrisi, Sam; Manbeck, Adrienne ; Goodwin, Madeline; Ernst, Monique; Grillon, Christian “Periaqueductal gray matter and medial prefrontal cortex reflect negative prediction errors during differential conditioning” Social Cognitive and Affective Neuroscience [Accepted]
Gorka, Adam X; Philips, Ryan T.; Torrisi, Sam; Claudino, Leonardo; Foray, Katherine; Ernst, Monique; Grillon, Christian “The posterior cingulate cortex reflects the impact of anxiety on drift rates during cognitive processing.” Biological Psychiatry: CNNI (2022). https:// doi.org/10.1016/j.bpsc.2022.03.010.
Philips, Ryan T.; Torrisi, Sam; Gorka, Adam X; Grillon, Christian; Ernst, Monique “Dynamic time warping identifies functionally distinct fMRI resting state cortical networks specific to VTA and SNc: a proof of concept” Cerebral Cortex 32(6), 1142-1151. https://doi.org/10.1093/cercor/bhab273
Ernst, Monique; Gowin, Joshua L.; Gaillard, Claudie; Philips, Ryan T.; Grillon, Christian “Sketching the Power of Machine Learning to Decrypt a Neural Systems Model of Behavior.” Brain sciences 9, no. 3 (2019): 67. https://doi.org/10.3390/brainsci9030067
Philips, Ryan T.; Sur, Mriganka; Chakravarthy, V. Srinivasa “The influence of astrocytes on the width of orientation hypercolumns in visual cortex: A computational perspective.” PLoS computa- tional biology 13.10 (2017): e1005785. https://doi.org/10.1371/journal.pcbi.1005785
Philips, Ryan T.; Chakravarthy, V. Srinivasa “A global orientation map in the primary visual cortex (V1): Could a self organizing model reveal its hidden bias?” Frontiers in neural circuits 10 (2017): 109. https://doi.org/10.3389/fncir.2016.00109
Philips, Ryan T.; Chhabria, Karishma; Chakravarthy, V. Srinivasa “Vascular Dynamics Aid a Coupled Neurovascular Network Learn Sparse Independent Features: A Computational Model.” Frontiers in neural circuits 10 (2016): 7. https://doi.org/10.3389/fncir.2016.00007
Philips, Ryan T.; Chakravarthy, V. Srinivasa “The mapping of eccentricity and meridional angle onto orthogonal axes in the primary visual cortex: An activity-dependent developmental model.” Frontiers in computational neuroscience 9 (2015): 3. https://doi.org/10.3389/fncom.2015.00003
Work with us
If you would like to work with us please fill out this google form