There were good reasons for MIP to be widely adopted: it is parameter free, fast and straightforward to use as it is implemented in NIH Fiji/ImageJ 7 and in many other programmes.ĭespite its widespread adoption, MIP presents a major drawback: it produces a highly discontinuous index map. The image of levels is called the index map while the image made of intensity values corresponding to that index map is called the composite image or the projection. MIP consists in retrieving the level of maximum intensity along the z axis for each x, y position. In practice, more than 80% of the biology community who acquire 3D volumes use maximum intensity projection (MIP), one of the simplest Z projection methods, to reduce a 3D stack into a single 2D image as reflected in a survey (see ‘Methods’ section). Finally, because it produces a single 2D image with no need of an interactive interface. Second, because resolution is always lower in z than in x/ y in fluorescence microscopes. First, because it contains the highest amount of information. In this case, a projection in the z direction is often preferred. In contrast, a 3D stack contains all pieces of information needed to reconstruct a focused 2D image.

This is because at high resolution, those objects cannot hold within the depth of field of a microscope and therefore a single 2D image acquisition often suffers from being only partially in focus. Image stacks are also useful to image flattish or so-called 2.5D objects such as an epithelium, a monolayered cell culture, a membrane within an in vivo tissue sample or a flat biological structure such as cultured neurons. A variety of software programme as Voxx 6, NIH’s Fiji/ImageJ 7 or VTK 8 propose such options with an interactive visualization. Optionally, a detection step can be introduced before the projection to focus on the reconstruction of layers of interest 5. In this case, a single projection cannot be satisfactory and an interactive rendering tool must be used to obtain successive projections of the data in arbitrary directions using transparency filters. Unlike a single 2D image, which can be rendered and interpreted straightforwardly on a screen, a 3D stack needs to be processed to be displayed.Īn image stack can be employed to image a full 3D object. Furthermore, it is common nowadays to image the same preparation repetitively at various depths to obtain a three-dimensional (3D) image under the form of a stack of two-dimensional (2D) images. Cell biologists routinely use fluorescence microscopy to observe spatial organization of organelles, co-localization of vesicles or to study the impact of gene expression or small compounds on cell morphology 1, 2, 3, 4.

Convert 3d image to 2d in matlab