Anthropometry of Human Muscle Using Segmentation Techniques and 3D Modelling: Applications to Lower Motor Neuron Denervated Muscle in Spinal Cord Injury

Abstract

This chapter describes a novel approach to determining muscle anthropometry using medical imaging and processing techniques to evaluate and quantify: (1) progression of atrophy in permanent muscle lower motor neuron (LMN) denervation in humans and (2) muscle recovery as induced by functional electrical stimulation (FES). Briefly, we used three-dimensional reconstruction of muscle belly and bone images to study the structural changes occurring in these tissues in paralyzed subjects after complete lumbar-ischiatic spinal cord injury (SCI). These subjects were recruited through the European project RISE, an endeavour designed to establish a novel clinical rehabilitation method for patients who have permanent and non-recoverable muscle LMN denervation in the lower extremities. This chapter describes the use of anthropometric techniques to study muscles in several states: healthy, LMN denervated-degenerated not stimulated, and LMN denervated-stimulated. Here, we have used medical images to develop three-dimensional models, including computational models of activation patterns induced by FES. Shape, volume and density changes were measured on each part of the muscles studied. Changes in tissue composition within both normal and atrophic muscle were visualized by associating the Hounsfield unit values of fat and connective tissue with different colours. The minimal volumetric element (voxel) is approximately ten times smaller than the volume analyzed by needle muscle biopsy. The results of this microstructural analysis are presented as the percentage of different tissues (muscle, loose and fibrous connective tissue, fat) in the total volume of the rectus femoris muscle; the results display the first cortical layer of voxels that describe the muscle epimisium directly on the three-dimensional reconstruction of the muscle. These analyses show restoration of the muscular structure after FES. The three-dimensional approach used in this work also allows measurement of geometric changes in LMN denervated muscle. The computational methods developed allow us to calculate curvature indices along the muscle’s central line in order to quantify changes in muscle shape during the treatment. The results show a correlation between degeneration status and changes in shape; the differences in curvature between control and LMN denervated muscle diminish with the growth of the latter. Bone mineral density of the femur is also measured in order to study the structural changes induced by muscle contraction and current flow. Importantly, we show how segmented data can be used to build numerical models of the stimulated LMN denervated muscle. These models are used to study the distribution of the electrical field during stimulation and the activation patterns.