This work presents Structural Medicine v1. 7. 1, a predictive framework for disease progression based on time-dependent structural dynamics derived from longitudinal cognitive data. Conventional models describe disease progression as a monotonic increase in decay rate. In contrast, this study demonstrates that progression follows a two-stage structural process: variance instability → directional collapse bias Using ADNI longitudinal data, structural persistence F (t) F (t) F (t) is derived from normalized MMSE scores, and the time-dependent decay rate λ (t) (t) λ (t) is computed. The analysis shows that: The distribution of λ (t) (t) λ (t) widens with diagnostic severity, indicating increasing variance instability Variance peaks occur slightly more often before diagnostic conversion than expected under a random baseline (66. 8% vs 64. 0% ± 1. 4%, Z = 2. 07), providing a weak but non-random anticipatory signal Directional asymmetry AAA, defined as the imbalance between positive and negative fluctuations, shifts significantly toward collapse-biased values at the LMCI stage Both mean and median statistics confirm that this shift reflects a distribution-level structural transition rather than outlier effects These results indicate that structural disease progression is not governed by decay magnitude alone, but by the emergence of directional bias in fluctuations. Variance instability acts as an early but weak signal, while directional asymmetry defines the onset of irreversible collapse dynamics. This framework provides a minimal, testable link between fluctuation structure and clinical transition, and suggests a new approach to early detection based on structural dynamics rather than static biomarkers. The ADNI dataset is not redistributed and must be obtained independently.
Koji Okino (2026) studied this question.