AQAL: A Foundation Model Approach to Predicting Neurodiverse Brain Activity

Abstract

Contents

1.Introduction

Autism spectrum conditions affect approximately 1 in 36 children worldwide (CDC, 2023). Understanding how autistic brains process sensory information differently from neurotypical brains is critical for designing inclusive environments, educational strategies, and clinical interventions. However, functional neuroimaging studies are expensive, slow, and inaccessible to most practitioners and families.

Recent advances in brain encoding models — neural networks trained to predict brain activity from sensory stimuli — offer a new paradigm. State-of-the-art multimodal transformers can now predict cortical surface activity from vision, audition, and language inputs with meaningful accuracy across subjects.

We asked: Can a brain encoding model, combined with population-level connectivity data from autism neuroimaging, predict how a neurodiverse brain would respond to arbitrary stimuli?

AQAL is our answer. Rather than requiring individual brain scans, AQAL uses a proprietary statistical transform trained on a large multi-site fMRI corpus to convert neurotypical brain predictions into neurodiverse variants, making neuroscience-informed predictions accessible through a web API.

1.1 Contributions

1. A CPU-based statistical transform that converts neurotypical predictions to neurodiverse predictions using connectivity effect sizes from 871 subjects across 20 clinical sites.
2. A connectivity analysis pipeline identifying 820 significant inter-regional differences between ASD and TD groups across 4,950 tested connections.
3. A sensory profiling system that maps brain-level divergence onto seven functional networks (visual, auditory, motor, language, social, default mode, salience).
4. A publicly accessible platform (API + web interface) enabling real-time neurodiverse brain prediction from text, audio, or video input.

2.Related Work

2.1 Brain Encoding Models

Brain encoding models learn a mapping from stimulus features to neural activity. Early approaches used linear ridge regression on hand-crafted features (Huth et al., 2016; de Heer et al., 2017). More recent work leverages deep neural network representations, with models like BrainLM (Ortega Caro et al., 2023) and BrainBERT (Thomas et al., 2023) achieving increasingly accurate predictions across modalities.

AQAL builds on this lineage with a proprietary multimodal architecture that accepts vision, audio, and language inputs simultaneously and predicts activity across 20,484 cortical surface vertices using a deep transformer encoder.

2.2 Autism Neuroimaging

Large-scale multi-site autism neuroimaging consortia have provided resting-state fMRI data from over 1,100 individuals, enabling population-level connectivity analysis. Studies using these datasets have consistently identified differences in functional connectivity between ASD and TD groups, particularly in the default mode network, salience network, and limbic regions (Hull et al., 2017; Holiga et al., 2019).

2.3 Gap

No prior work has combined a multimodal brain encoding model with population-level autism connectivity data to generate stimulus-specific neurodiverse brain predictions. AQAL bridges this gap.

3.Methods

3.1 System Architecture

Stimulus (text / audio / video)
        │
        ▼
┌───────────────────────┐
│   Feature Extraction  │  Proprietary multimodal encoders
└──────────┬────────────┘
           │
           ▼
┌───────────────────────┐
│   AQAL Brain Encoder  │  8-layer transformer, 177M params
│   (NT Prediction)     │  → 20,484 vertices × T timesteps
└──────────┬────────────┘
           │
           ▼
┌───────────────────────┐
│   ND Transform        │  Trained on 871 multi-site subjects
│   (ASD Prediction)    │  Per-vertex scale + shift
└──────────┬────────────┘
           │
           ▼
┌───────────────────────┐
│   Analysis &          │  Divergence metrics, sensory profile,
│   Visualization       │  connectivity mapping, interpretation
└───────────────────────┘

3.2 AQAL Brain Encoder

The AQAL brain encoder is a proprietary multimodal transformer trained to predict cortical surface activity from sensory inputs. The model architecture consists of:

• Feature extractors: Dedicated encoders for text (6 layers), video (2 layers), and audio (2 layers) built on state-of-the-art foundation models
• Feature projectors: Per-modality MLPs mapping to a common hidden dimension of 1,152
• Temporal smoothing: Gaussian kernel (size 9) for hemodynamic response alignment
• Transformer encoder: 8 layers, 18 heads, head dimension 64, FF dimension 3,072
• Subject layers: Low-rank output adaptation (bottleneck 2,048) with 0.1 dropout
• Output space: 20,484 vertices on a standard cortical surface mesh (10,242 per hemisphere)

The model operates at 2 Hz temporal resolution. Modality dropout of 0.3 enables predictions from any subset of input modalities.

3.3 Training Dataset

We trained the neurodiverse transform on a large-scale multi-site autism neuroimaging corpus comprising 871 resting-state fMRI scans:

All scans were preprocessed using a standard fMRI preprocessing pipeline with band-pass filtering (0.01–0.1 Hz) and no global signal regression. We extracted mean timeseries from 100 cortical parcels using a validated cortical parcellation, computed Pearson correlation matrices, and applied Fisher's z-transform for variance stabilization, producing 4,950 unique connectivity features per subject.

3.4 Statistical Group Comparison

For each of the 4,950 unique connectivity pairs, we performed an independent-samples t-test comparing ASD and TD groups. 820 of 4,950 connections (16.6%) were significant at p < 0.05.

The most affected regions ranked by normalized effect size:

These findings are consistent with the autism neuroimaging literature, which identifies limbic hyperconnectivity and default mode network alterations as hallmarks of ASD (Hull et al., 2017).

3.5 Neurodiverse Transform

We derive a proprietary per-vertex transform that converts neurotypical predictions to neurodiverse predictions. For each of 20,484 cortical vertices v:

1. Identify the cortical ROI r(v) to which vertex v belongs.
2. Compute the normalized effect size e_r for that ROI from connectivity t-tests.
3. Compute scale and shift factors using a proprietary mapping function.
4. Apply: &ycirc;_v^ND = &ycirc;_v^NT × scale_v + shift_v

Of 20,484 vertices, 18,067 (88.3%) are affected by the transform, reflecting the widespread but heterogeneous nature of connectivity differences in autism.

3.6 Sensory Profiling

We map vertex-level divergence between NT and ND predictions onto seven canonical functional brain networks:

Scores are normalized to [0, 1] by dividing by the maximum network divergence, providing an interpretable summary of which processing domains are most differently activated.

3.7 Infrastructure

All inference runs on CPU. No GPU is required.

4.Results

4.1 Connectivity Analysis

Analysis of 871 subjects across 20 clinical sites revealed 820 significant connections out of 4,950 tested (16.6%, p < 0.05). The limbic system showed the largest effect sizes, consistent with known emotional regulation differences. Default mode network showed bilateral temporal alterations. Visual cortex alterations align with documented sensory processing variations.

4.2 Network-Level Differences

4.3 Transform Coverage

The neurodiverse transform affects 18,067 of 20,484 cortical vertices (88.3%), with scale factors ranging from 0.85 to 1.15 and shift factors from \u22120.05 to 0.01. The transform is denser in temporal and limbic regions and sparser in primary motor cortex, matching the known neuroanatomy of autism.

4.4 Real-Time Performance

5.Applications

5.1 Sensory Accessibility Auditing

AQAL can process video of physical spaces (classrooms, offices, retail environments) and generate second-by-second sensory stress predictions for neurodiverse individuals, including flagged high-stress moments and actionable accessibility recommendations.

5.2 Educational Accommodation Design

By profiling how specific educational content differently activates neurodiverse brain networks, AQAL can inform content pacing, optimal break timing, and modality-specific accommodations.

5.3 Clinical Communication

The sensory profile output provides a quantitative, brain-data-grounded framework for communicating how a neurodiverse individual might experience specific stimuli — replacing subjective descriptions with measurable network-level predictions.

6.Limitations

1. No GPU fine-tuning. The neurodiverse transform is a statistical approximation, not a directly fine-tuned neural network. GPU fine-tuning on neurodiverse fMRI data would likely improve accuracy but was not feasible due to compute constraints.
2. Indirect mapping. The transform maps connectivity-level group differences onto vertex-level predictions. This assumes resting-state connectivity differences generalize to task-evoked activity — a reasonable but imperfect assumption.
3. Group-level, not individual. AQAL produces an average neurodiverse prediction. Autism is a spectrum, and individual variation is substantial.
4. Uncorrected p-values. The 820 significant connections are reported at p < 0.05 without multiple comparison correction. With FDR correction, the number decreases.
5. Dataset limitations. The training corpus is heterogeneous (20 sites, varying protocols) and skewed toward males. Results may not generalize equally to all populations.

7.Future Work

• GPU fine-tuning of the AQAL encoder directly on neurodiverse fMRI data
• Individual calibration via a 5-minute sensory assessment to personalize predictions
• Expanded training data from additional consortia to reach 10,000+ subjects
• Additional neurodivergent conditions (ADHD, sensory processing disorder, anxiety)
• EEG integration for real-time, portable brain monitoring
• Clinical validation studies comparing AQAL predictions against observed behavioral responses

8.Conclusion

AQAL demonstrates that combining a state-of-the-art multimodal brain encoding model with population-level autism neuroimaging data can produce meaningful neurodiverse brain predictions from arbitrary sensory stimuli. While the current statistical transform is an approximation, it makes neuroscience-informed predictions accessible without requiring individual brain scans or GPU infrastructure. The identification of 820 significant connectivity differences across 871 subjects provides a robust empirical foundation, and the system's real-time performance enables practical applications in accessibility, education, and clinical communication.

By releasing the platform publicly, we aim to make computational neurodiversity research accessible to researchers, educators, clinicians, and families who lack access to neuroimaging facilities.

References

Appendix A: Model Parameters

Appendix B: Training Corpus Demographics

Appendix C: System Availability