Measuring Dependencies between Biological Signals with Self-supervision, and its Limitations

Evangelos Sariyanidi, John D Herrington, Lisa D Yankowitz, Pratik Chaudhari, Theodore D. Satterthwaite, Casey J. Zampella, Robert T Schultz, Russell T. Shinohara, Birkan Tunc

Track: Track 1: Original Research/Position/Education/Attention Track

Keywords: time series, dependence, statistical, biological, biology

TL;DR: Exposing linear or non-linear statistical dependencies in biological signals

Abstract: Measuring the statistical dependence between observed signals is a primary tool for scientific discovery. However, biological systems often exhibit complex non-linear interactions that currently cannot be captured without a priori knowledge regarding the nature of dependence. We introduce a new approach for measuring statistical dependence, namely concurrence, which is inspired by the observation that if two signals are dependent, then one should be able to distinguish between temporally aligned vs. misaligned segments extracted from them. Concurrence can become a standard tool for scientific analyses across fields, as it is, to our knowledge, the first approach that can expose relationships across a wide spectrum of signals (fMRI, physiological and behavioral data) and extract scientifically relevant differences without ad-hoc parameter tuning or large datasets. However, dependencies due to extraneous factors remain an open problem, thus researchers should validate that exposed relationships truly pertain to the question(s) of interest.

Submission Number: 66