Graph analytics methods have evinced significant interest in recent years. Their applicability to real world complex systems is currently limited by the challenges with inferring effective graph representations of the high dimensional, noisy, nonlinear and transient dynamics from limited time series outputs, as well as in extracting statistical quantifiers that capture the salient structure of the inferred graphs for change detection. In this paper, we present an approach based on spectral graph theory to detect changes in complex dynamic systems using a single realization of time series data collected under specific, common types of transient conditions, such as intermittency. We introduce a statistic, \(\gamma_{k}\), based on the spectral content of the inferred graph. We show that \(\gamma_{k}\) statistic under high dimensional dynamics converges to a normal distribution, and employ the parameters of this distribution to construct a procedure to detect qualitative changes in the coupling structure of a dynamical system. Experimental investigations suggest that \(\gamma_{k}\) statistic by itself is able to detect changes with modified area under curve (mAUC) of about 0.96 (for numerical simulation tests), and can, by itself, achieve true positive rate of about 40% for detecting seizures from EEG signals. In addition, by incoporating this statistic with random forest, one of the best seizure detection methods, the seizure detection rate of random forest method is improved by 5% in 35% of the subjects. These studies on the network inferred from EEG signals suggest that \(\gamma_{k}\) can capture the salient structural changes in the physiology of the process and can therefore serve as an effective feature for detecting seizures from EEG signal.