This paper presents the results of an early work on different concepts and technologies that can be used to recognize the context in Mixed Reality delivered by means of a head-mounted display through spatial reasoning. We focused on reasoning about the detected relationships between physical objects in a mixed reality scene through a head-mounted display (e.g., the Microsoft Hololens 2). We propose a solution implemented through a distributed architecture, where an Artificial Intelligence (AI) server is connected to a Mixed Reality head-mounted display. This server hosts a hybrid reasoning approach with two modules: an object detection module and a knowledge-based module built using ontologies. The first module identifies objects in the video stream from the head-mounted display. Another module deployed on that device then localizes the objects in a 3D coordinate system and sends the result back to the ontological knowledge-based module. This module performs an inference on the object locations for spatial reasoning. Finally, we have performed an evaluation of this solution in a prototype that provides users with details about spatial relationships between geometric shapes in the Mixed Reality scene.