It should be noted that these modern imaging methods involve sophisticated instrumentation and equipment, which employ high-speed electronics and computers for data collection. Spatial-temporal image data occur in a broad range of medical imaging applications. It is now common for patients to be imaged multiple times, either by repeated imaging with a single modality, or by imaging with different modalities. It is also common for patients to be imaged dynamically, i.e., to have sequences of images acquired, often at many frames per second. The ever increasing amount of image data acquired makes it more and more desirable to relate more than one statistical tool to assist in extracting relevant clinical information.
Our research is concerned with the spatial-temporal data mining motivated by analyzing data from our “Neuromuscular Electrical Stimulation” experiment. We develop an efficient procedure for mining spatial-temporal data – Longitudinal Analysis with Self-Registration (LASR, pronounced “laser”). This new procedure is a statistical ensemble built on following modern or newly developed components: (1) data segmentation for separating heterogeneous data and for distinguishing outliers, (2) automatic approaches for spatial and temporal data registration, (3) statistical smoothing mapping for identifying “activated” regions based on generalized false discovery rate (FDR) controlled p-maps/movies from “large-p-small-n” data sets. Our new procedure should be applicable to other types of spatial-temporal data sets beyond those from the NMES experiment. It has the potential to be used in the analysis of time-series images and functional images such as those from fMRI.
