CAMOS: Coupled Oscillatory State-Space Model for Multimodal Clinical Time-Series
Abstract
Longitudinal clinical cohorts are multimodal, irregularly sampled and pervasively incomplete: in ADNI, positron emission tomography and cerebrospinal fluid assays are absent from roughly half of all visits. Linear state-space models handle irregular sampling gracefully but treat a missing modality by masking the input, leaving the transition operator untouched. We prove that this is a representational limitation: the latent state of any linear state-space layer whose transition operator does not depend on the availability pattern is an additive function of the availability indicators, so no such layer can represent an interaction between two modalities being jointly present or jointly absent. We propose CAMOS, which gives each modality a bank of second-order oscillators coupled through a matrix that sits inside the differential equation and is gated by availability, so the transition operator itself becomes a function of which measurements were taken. Coupling invalidates the analysis of uncoupled oscillatory models, and we restore it: a per-channel Gershgorin budget makes the effective stiffness positive definite uniformly over all 2M availability patterns and all gaps, an energy argument charges amplification to availability transitions rather than sequence length, and a channel factorization preserves exact associative parallel scans. On ADNI, CAMOS outperforms uncoupled oscillatory state-space models and clinical fusion models on same-visit staging, landmark prediction and longitudinal forecasting, and under zero-shot transfer to OASIS-3 it is the only model that avoids collapse to the majority class.
Related
- When Does Text Inform? Benchmarking Information-Theoretic Metrics for Multimodal Time-Series Forecasting
- Transport statistics in asymmetric coupled chaotic dots in series: Brownian motion model
- Robust Fault Detection in Mechanical Multimodal Time Series via Self-Supervised Cross-Modal Reconstruction
- Evaluating Time-Series Foundation Models and Multimodal Dietary Context for CGM Forecasting
- Sentinel-1 SAR time series for forest phenology: multimodal deep learning with meteorological fusion across deciduous broadleaf and evergreen needleleaf forests
BibTeX
@misc{rahman2026camos,
title = {{CAMOS: Coupled Oscillatory State-Space Model for Multimodal Clinical Time-Series}},
author = {Maxx Richard Rahman and Mostafa Hammouda and Wolfgang Maass},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.39484},
eprint = {2609.39484},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.39484},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - CAMOS: Coupled Oscillatory State-Space Model for Multimodal Clinical Time-Series
AU - Maxx Richard Rahman
AU - Mostafa Hammouda
AU - Wolfgang Maass
PY - 2026
DA - 2026/09/30
PB - arXiv
DO - 10.48550/arxiv.2609.39484
UR - https://arxiv.org/abs/2609.39484
AB - Longitudinal clinical cohorts are multimodal, irregularly sampled and pervasively incomplete: in ADNI, positron emission tomography and cerebrospinal fluid assays are absent from roughly half of all visits. Linear state-space models handle irregular sampling gracefully but treat a missing modality by masking the input, leaving the transition operator untouched. We prove that this is a representational limitation: the latent state of any linear state-space layer whose transition operator does not depend on the availability pattern is an additive function of the availability indicators, so no such layer can represent an interaction between two modalities being jointly present or jointly absent. We propose CAMOS, which gives each modality a bank of second-order oscillators coupled through a matrix that sits inside the differential equation and is gated by availability, so the transition operator itself becomes a function of which measurements were taken. Coupling invalidates the analysis of uncoupled oscillatory models, and we restore it: a per-channel Gershgorin budget makes the effective stiffness positive definite uniformly over all 2^M availability patterns and all gaps, an energy argument charges amplification to availability transitions rather than sequence length, and a channel factorization preserves exact associative parallel scans. On ADNI, CAMOS outperforms uncoupled oscillatory state-space models and clinical fusion models on same-visit staging, landmark prediction and longitudinal forecasting, and under zero-shot transfer to OASIS-3 it is the only model that avoids collapse to the majority class.
ER -