arXiv subjects

Diogo Sousa

Publications and source records attributed to Diogo Sousa.

3 recordsLinked to original sources

arXiv2026

A Regional Dopamine-Serotonin Control Model of Conscious State

This article proposes a dopamine-serotonin theory of conscious state control. Regional dopamine signaling controls a candidate intensity coordinate and serotonergic signaling changes a candidate coordinate of felt complexity and significance within the capacity of the neural substrate. A receptor model, a regional bottleneck and a cumulative intensity response with an upper collapse gate connect these coordinates. Extensions cover sleep, seizures, regional demand, species differences, development, terminal collapse and receptor priming. A fixed regional benchmark and a weight-cancelling sensitivity identity provide quantitative tests across perturbations. Exploratory PPMI analyses examine REM timing and regional motor response. Restricted waiting for sustained REM uses 10,537 nights from 137 participants, with formulation variation in 36. A formulation contrast gives +1.220 minutes per 100 mg/day LEDD substitution at fixed total levodopa LEDD (95% interval 0.425 to 2.014, Holm-adjusted p = 0.0546 across six tests). Individual trends reduce the estimate to +0.905 minutes (-0.037 to 1.847), and removing the largest information contributor gives +0.482 minutes (-0.308 to 1.272). Estimates stay positive after deleting the two or three largest contributors, but all four intervals span zero. Conventional REM phase and latency coefficients are positive in both temporal models. A regional motor comparison in 254 participants finds an uncertain incremental prediction gain from a regional imaging response term (log-score difference $-7.38\times10^{-5}$, 95% interval $[-3.72,3.68]\times10^{-3}$). These outcomes measure sleep and motor behavior, rather than receptor signaling or conscious experience. The REM pattern motivates a prospective entry-mechanism test.

q-bio.NC↗

arXiv2026

A Physics-Informed Statistical Learning Model for Long-Term Fragmentation Cloud Propagation

This paper introduced a Hierarchical Generative Density Model (HGDM) for the long-term propagation of orbital fragmentation clouds. Validation against high-fidelity Monte Carlo simulations showed that the proposed surrogate accurately reproduces the dominant multidimensional structures of propagated clouds while consistently outperforming classical band-formation approximations based on independent angular variables. Accurate cloud reconstructions were obtained using only a few hundred to a few thousand propagated fragments, yielding reductions exceeding two orders of magnitude in computational cost and three orders of magnitude in storage requirements; future work will investigate its application to large-scale debris-environment evolution and collision-cascade simulations associated with the Kessler syndrome.

math.NA↗

arXiv2025

Performance of Large Language Models in Supporting Medical Diagnosis and Treatment

The integration of Large Language Models (LLMs) into healthcare holds significant potential to enhance diagnostic accuracy and support medical treatment planning. These AI-driven systems can analyze vast datasets, assisting clinicians in identifying diseases, recommending treatments, and predicting patient outcomes. This study evaluates the performance of a range of contemporary LLMs, including both open-source and closed-source models, on the 2024 Portuguese National Exam for medical specialty access (PNA), a standardized medical knowledge assessment. Our results highlight considerable variation in accuracy and cost-effectiveness, with several models demonstrating performance exceeding human benchmarks for medical students on this specific task. We identify leading models based on a combined score of accuracy and cost, discuss the implications of reasoning methodologies like Chain-of-Thought, and underscore the potential for LLMs to function as valuable complementary tools aiding medical professionals in complex clinical decision-making.

cs.CL↗