Identification, Quantification, And Characterization Of Morphological Patterns in the Mid-thigh

Regional Tissue Identification, Quantification and Characterization

Members: Azubuike Okorie, Chelsea Harris, Keni Zheng, Sokratis Makrogiannis

The quantification of regional tissue components and their longitudinal changes is a key parameter in epidemiological studies, because it may help to monitor changes in body composition and function. The purpose of this project is to develop an automated method to delineate and quantify the main tissue components in the middle thigh anatomical site from volumetric MRI data.

Our objective is to delineate the main tissues in mid-thigh, namely subcutaneous fat, inter-muscular fat, muscle, cortical and endosteal bone components. The subcutaneous fat is segmentation using active contours, and the inter-muscular fat is separated from muscle using clustering methods. Finally the bone is identified by shape recognition. The quantification of each tissue involves computation of volume, average area, and first and second order intensity statistics. The method was validated versus reference quantification on CT imaging data of the same subjects produced by a semi-manual workflow.

Fig. 1: Overview of the thigh quantification method.

Fig. 2: Performance evaluation against a semi-manual CT-based quantification.

 


References

  1. K. Zheng, Harris, C. E., Jennane, R., and Makrogiannis, S., Integrative blockwise sparse analysis for tissue characterization and classification, Artificial Intelligence in Medicine, vol. 107, p. 101885, 2020.
  2. S. Makrogiannis, Okorie, A., Biswas, T., and Ferrucci, L., Multi-atlas-based tissue identification in the lower leg using pQCT, in SPIE Medical Imaging 2020: Image Processing, 2020.
  3. T. Biswas and Makrogiannis, S., Learning Graph Cut Class Prototypes for Thigh CT Tissue Identification, in Advances in Visual Computing, Cham, 2019.
  4. S. Makrogiannis, Okorie, A., Biswas, T., and Ferrucci, L., Shape Modeling and Atlas-Based Segmentation for Identification of Lower Leg Tissues in pQCT, in 2019 IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2019.
  5. S. Makrogiannis, Boukari, F., and Ferrucci, L., Automated skeletal tissue quantification in the lower leg using peripheral quantitative computed tomography, Physiological Measurement, vol. 39, p. 035011, 2018.
  6. S. Makrogiannis, Fishbein, K. W., Moore, A. Z., Spencer, R. G., and Ferrucci, L., Image-Based Tissue Distribution Modeling for Skeletal Muscle Quality Characterization., IEEE Trans Biomed Eng, vol. 63, no. 4, pp. 805-13, 2016.
  7. S. Makrogiannis and Ferrucci, L., Software system for computing material and structural properties of bone and muscle in the lower extremity from pQCT, Proc. SPIE 9112, Sensing Technologies for Global Health, Military Medicine, and Environmental Monitoring IV, vol. 9112. pp. 911216-911216-6, 2014.
  8. S. Makrogiannis, Caturegli, G., Davatzikos, C., and Ferrucci, L., Computer-aided assessment of regional abdominal fat with food residue removal in CT., Acad Radiol, vol. 20, pp. 1413–1421, 2013.
  9. S. Makrogiannis, Ramachandran, R., Chia, C. W., and Ferrucci, L., Automated abdominal fat quantification and food residue removal in CT, in Mathematical Methods in Biomedical Image Analysis (MMBIA), 2012 IEEE Workshop on, 2012.
  10. S. Makrogiannis, Serai, S., Fishbein, K. W., Schreiber, C., Ferrucci, L., and Spencer, R. G., Automated quantification of muscle and fat in the thigh from water-, fat-, and nonsuppressed MR images., J Magn Reson Imaging, vol. 35, pp. 1152–1161, 2012.