Search for: All records

We consider the problem of a swarm of nanobots detecting and treating human cancer that is diffuse, dispersed in a region with multiple separate cancer sites in need of treatment. We present a mathematical model of nanobots and their colloidal environment that is inspired by actual chemotactic nanoparticles, involving agents noisily following chemical gradients (both attractively and repellently, depending on the chemical).We present three incrementally sophisticated algorithms that describe additional chemical payloads that agents carry onboard, beyond the cancer-treating drug K, as well as the rules for when agents drop their payloads: Algorithm KM, in which agents simply ascend naturally existing chemical M signals that surround cancer sites; KMA, in which agents themselves amplify these natural signals by dropping chemical A payloads upon reaching a site; and KMAR, in which agents choose to either amplify the signal by dropping chemical A or counteract/reduce the signal by dropping chemical R, according to the current unsatisfied demand of the site.We present simulation results for all of the algorithms, across a set of distinct cancer arrangements, that track both the achieved treatment success as well as the time/duration of the treatment. KM has generally successful treatment unless the natural M-signals are weak, in which case the treatment progresses too slowly. KMA demonstrates a significant speedup in treatment time (over KM), but also a drop in success except for the most concentrated cancer patterns. KMAR has relatively optimal performance across all types of cancer patterns, demonstrating robustness and adaptability in its mechanisms for nanobot coordination. 

In this work, we propose a neural mechanistic theory onhow animals plan hierarchically for decision making. • How to learn hierarchyfrom experience? • How to learn hierarchical subgoal values? • How to plan over learned hierarchy and subgoal values? 

Deploying motile nanosized particles, also known as “nanobots,” in the human body promises to improve selectivity in drug delivery and reduce side effects. We consider a swarm of nanobots locating a single cancerous region and treating it by releasing an onboard payload of drugs at the site. At nanoscale, the computation, communication, sensing, and locomotion capabilities of individual agents are extremely limited, noisy, and/or nonexistent. We present a general model to formally describe the individual and collective behavior of agents in a colloidal environment, such as the bloodstream, for cancer detection and treatment by nanobots. This includes a feasible and precise model of agent locomotion, inspired by actual nanoparticles that, in the presence of an external chemical gradient, move toward areas of higher concentration by means of self-propulsion. We present two variants of our model: the first assumes an endogenous chemical gradient fixed over time and centered at the cancer site; the second is a more speculative, dynamic variant in which agents themselves create and amplify a gradient centered at the cancer site. In both settings, agents sense the gradient and ascend it noisily, locating the cancer site more quickly than by simple Brownian motion. For the first variant, we present simulation and analytical results to bound the time it takes for agents to reach the cancer site. For the second, simulations highlight collective benefit from agent-generated signaling, showing significant runtime improvement via chemical signal amplification. 

Hierarchical decisions in natural environments require processing uncertainty across multiple levels, but existing models struggle to explain how animals perform flexible, goal-directed behaviors under such conditions. Here we introduce CogLinks, biologically grounded neural architectures that combine corticostriatal circuits for reinforcement learning and frontal thalamocortical networks for executive control. Through mathematical analysis and targeted lesion, we show that these systems specialize in different forms of uncertainty, and their interaction supports hierarchical decisions by regulating efficient exploration, and strategy switching. We apply CogLinks to a computational psychiatry problem, linking neural dysfunction in schizophrenia to atypical reasoning patterns in decision making. Overall, CogLink fills an important gap in the computational landscape, providing a bridge from neural substrates to higher cognition. 

Human decision-making involves model-free and model-based reinforcement learning (RL) strategies, largely implemented by prefrontal-striatal circuits. Combining human brain imaging with neural network modelling in a probabilistic reversal learning task, we identify a unique role for the mediodorsal thalamus (MD) in arbitrating between RL strategies. While both dorsal PFC and the striatum support rule switching, the former does so when subjects predominantly adopt model-based strategy, and the latter model-free. The lateral and medial subdivisions of MD likewise engage these modes, each with distinct PFC connectivity. Notably, prefrontal transthalamic processing increases during the shift from stable rule use to model-based updating, with model-free updates at intermediate levels. Our CogLinks model shows that model-free strategies emerge when prefrontal-thalamic mechanisms for context inference fail, resulting in a slower overwriting of prefrontal strategy representations - a phenomenon we empirically validate with fMRI decoding analysis. These findings reveal how prefrontal transthalamic pathways implement flexible RL-based cognition. 

Electrophysiological implants enable exploration of the relationship between neuronal activity and behavior. These technologies evolve rapidly, with multiple iterations of recording systems developed and utilized. Chronic implants must address a litany of complications, including retention of high signal-to-noise ratio in probes and the ability to withstand excess force over the experimental period. To overcome these issues, we designed a chronic implant for rats. Our comprehensive protocol optimizes the entire implant process, from assembling and testing the probes (Neuropixels) to implantation. In addition to addressing the complications previously mentioned, our implant can vertically adjust probes with micron precision and is constructed using modular components, allowing it to be easily modified for various research contexts, electrophysiological recording systems, headstages, and probe types. 

Optimal decision-making requires consideration of internal and external contexts. Biased decision-making is a transdiagnostic symptom of neu- ropsychiatric disorders. We created a computational model demonstrating how the striosome compartment of the striatum constructs a context- dependent mathematical space for decision-making computations, and how the matrix compartment uses this space to define action value. The model explains multiple experimental results and unifies other theories like reward prediction error, roles of the direct versus indirect pathways, and roles of the striosome versus matrix, under one framework. We also found, through new analyses, that striosome and matrix neurons increase their synchrony during difficult tasks, caused by a necessary increase in dimensionality of the space. The model makes testable predictions about individual differences in disorder susceptibility, decision-making symptoms shared among neuropsychiatric disorders, and differences in neuropsychiatric disorder symptom presenta- tion. The model provides evidence for the central role that striosomes play in neuroeconomic and disorder-affected decision-making. 

Optimal decision-making requires consideration of internal and external contexts. Biased decision-making is a transdiagnostic symptom of neuropsychiatric disorders. We created a computational model demonstrating how the striosome compartment of the striatum constructs a context-dependent mathematical space for decision-making computations, and how the matrix compartment uses this space to define action value. The model explains multiple experimental results and unifies other theories like reward prediction error, roles of the direct versus indirect pathways, and roles of the striosome versus matrix, under one framework. We also found, through new analyses, that striosome and matrix neurons increase their synchrony during difficult tasks, caused by a necessary increase in dimensionality of the space. The model makes testable predictions about individual differences in disorder susceptibility, decision-making symptoms shared among neuropsychiatric disorders, and differences in neuropsychiatric disorder symptom presentation. The model provides evidence for the central role that striosomes play in neuroeconomic and disorder-affected decision-making. 

Poster 

Poster