Renewing the theoretical framework of intensity correlation analysis to evaluate blood-brain barrier permeability with spectral fluorescence microscopy in mice

In the context of induced blood-brain barrier (BBB) opening, estimating the parenchymal occurrence and concentration of different-sized molecules is an essential aspect for the optimization of focused and non-focused ultrasound protocols. Separate estimations of colocalization and correlation of signals in images partially address these aspects. However, the simultaneous analysis of localization and intensity of two signals holds promise to better understand the mechanistic players in ultrasound-induced BBB permeability. A new theoretical framework is proposed for intensity correlation and uncorrelation analysis (ICUA) enabling the spatial characterization of two signal distributions into 11 behaviors. ICUA was applied to a mice model of BBB opening induced by non-focused ultrasounds followed by 1kDa and 70kDa fluorophore injections and based on (1) parenchymal signal intensities and (2) injected blood concentration. Results showed that the spatial distribution of correlated intensities was strongly influenced by the pressure field pattern of unfocused US. Notably, when normalized by parenchymal signal intensities, the smaller molecule was found in higher intensities compared to the larger molecule in over 30% of the colocalization volume. Conversely, when normalized by injected concentration, the larger molecule was found to have accumulated in 30% of the brain volume for the group with the highest small molecule signal intensity. This two-step approach addressed how the fluorophores were spatially scattered and how this spatial organization was dependent on the efficiency of BBB opening. Overall, the analysis of the spatial localization and intrinsic parameters defining the ICUA helped estimate two physical parameters of BBB opening: size exclusion and duration of opening. Furthermore, the complementary analysis normalized by the injected concentration offered a new potential to explore complex BBB mechanisms such as clearance and binding. These results could help model the efficiency of US to open the BBB to different sizes of molecules and thus optimize the settings to reach the expected BBB opening for drug candidates.