CLOOSE: An open-source platform for optical closed-loop experiments

Brain-Computer Interfaces (BCI) have catalyzed advancements in both clinical applications and basic neuroscience research. However, technical barriers such as steep learning curves and complex synchronization requirements often impede their widespread adoption. In response to the increasing demand for optical BCI experiments and to the technical challenges associated with their implementation, we introduce CLOOSE (Closed-loop Optical Open-Source Experiments), a versatile platform designed to standardize, simplify, and accelerate closed-loop experiments with functional imaging data. CLOOSE interfaces easily with any frame-based imaging system via TCP/IP, allowing for real-time data streaming and distributed computation. Benchmark tests validate CLOOSE’s real-time accuracy in image registration, signal processing, and analysis at imaging frequencies ≥1 kHz, making it the first optical BCI compatible with voltage indicators. Throughout the paper we showcase CLOOSE’s versatility in supporting several neurofeedback paradigms: from single neuron to population dynamics, multiplane imaging, and online (and offline) z-tracking. CLOOSE’s functionality is easily accessible to users with minimal coding experience through an intuitive graphical interface for experiment setup and real-time performance monitoring. By significantly lowering the barrier to performing technically demanding closed-loop experiments, CLOOSE advances the streamlining, standardization, and democratization of online image analysis and neurofeedback BCI paradigms, opening new avenues for precise manipulation and real-time readout of neuronal activity in health and disease.