Models Research About GitHub
Neuro-AI Research Platform

Intelligence at the
edge of biology.

Arxelos deploys research-grade AI models at the intersection of neuroscience and deep learning. Each model is a working artifact — built to demonstrate, not just describe.

Models 3
Status Building
Stack PyTorch + FastAPI

Three models. One thesis.

Each model sits at the intersection of AI capability and biological insight — deployed, demo-ready, and built to withstand technical scrutiny.

MODEL 01
BUILDING

Brain Tumor MRI Classifier

Medical Computer Vision

4-class classification on 7,023 MRI scans. 96.2% test accuracy. The fast win — already built, now being productionized with a clean upload UI and cloud deployment.

CNN FastAPI Docker TensorFlow
MODEL 02
PLANNED

Virtual Lesions Visualizer

Neuro-AI / Interpretability

Ablate specific layers and channels in VGG-16 or ResNet-50. Measure whether prediction deficits mirror biological visual deficits — prosopagnosia, akinetopsia, achromatopsia.

PyTorch Grad-CAM Saliency Maps Interactive
MODEL 03
PLANNED

Medical Literature Q&A

NLP / Retrieval-Augmented Generation

Ingest PubMed abstracts, build a retrieval pipeline, generate domain-specific answers with source citations. Medical RAG with cited sources — not another generic chatbot.

LangChain ChromaDB Transformers FastAPI

Neuro-AI Research Track

An 8-week independent study mapping biological neural systems to deep learning architectures. Every lab produces a deployable artifact.

Phase 1

Bio-Electric Foundation

Weeks 1–2

Neuroanatomy, Hodgkin–Huxley neuron models, Leaky Integrate-and-Fire simulations, and mapping the human visual stream against CNN feature hierarchies.

Phase 2

Systems & Computational Principles

Weeks 3–5

Sparse coding, synaptic plasticity, STDP learning rules, Hopfield networks, place cell representations, and RL agents with biologically-inspired state encodings.

Phase 3

High-Level Cognition & Synthesis

Weeks 6–8

Language processing parallels, mechanistic interpretability, circuit analysis, and a capstone BCI neural signal decoder using EEG motor imagery data.

Key Papers

The theoretical backbone — primary literature driving each phase of the study.

  • Felleman & Van Essen (1991)
    Distributed Hierarchical Processing in Primate Cortex
  • Hodgkin & Huxley (1952)
    Quantitative Description of Membrane Current
  • Olshausen & Field (1996)
    Emergence of Simple-Cell Receptive Fields
  • Bi & Poo (1998)
    Synaptic Modifications in Hippocampal Neurons
  • Moser, Kropff & Moser (2008)
    Place Cells, Grid Cells & Spatial Representation
  • Hassabis et al. (2017)
    Neuroscience-Inspired Artificial Intelligence

Built by Aryan Patel

MS in Artificial Intelligence at Northeastern University. Background in deep learning, computer vision, and NLP. Building toward the intersection of neuroscience and machine intelligence.