Segmentation and network construction of a muscle biopsy image. (a) Fluorescence image corresponding to a control biopsy showing collagen VI content including the endomysium and perimysium (green), slow fibers (red) and fast fibers (black). The white square delimits the region of interest used in this case. (b) Segmented image enabling all fibers to be identified. The fibers are considered objects (white boundary lines) whose geometric properties can be calculated. (c) This information is used to produce a muscle network where each fiber is represented as a node, and two nodes are connected if two fibers are adjacent in the muscle biopsy. (d) Detail of the segmented image showing objects corresponding to the fibers (green boundary lines). The area of each object is termed A2 for subsequent calculations of the index of fibrosis. NDICIA identifies ‘slow’ and ‘fast’ fibers shown in red and black, respectively. (e) The objects are expanded in a linear manner until they reach the adjacent objects. This allows each fiber’s neighbors to be identified. A1 is the area of the expansion for each object, and is also used for calculation of the fibrosis index for each image. (f) Detail of the network corresponding to the region in (e). The network is formed by slow and fast nodes shown as red and black dots, respectively. (g,h) Region of interest selected from NA (BNA01-1, g) and MD (QD54-1, h) biopsies analyzed in this study. The collagen VI content (green), slow fibers (red) and fast fibers (black) are marked. Scale bar, 500 mm.
Sáez et al. BMC Medicine 2013 11:77 doi:10.1186/1741-7015-11-77