Keywords:
Computer applications, Oncology, Musculoskeletal bone, MR-Diffusion/Perfusion, Neural networks, Experimental, Computer Applications-Detection, diagnosis, Segmentation, Experimental investigations, Cancer, Image registration, Image verification
Authors:
V. K. Anand1, G. KRISHNAMURTHI1, R. Balaji2; 1Chennai/IN, 2CHENNAI, TAMILNADU/IN
DOI:
10.1594/ecr2018/C-2274
Aims and objectives
Multiple Myeloma(MM) is a malignancy of plasma cells.
Expansion of single clone of plasma cell derived from B cells in bone marrow leads to multiple myeloma.
Patients present with anemia,
immunosuppression,
renal failure,
hypercalcemia,
and bone destruction with pathologic fractures [1]. Conventional radiography was used in the diagnosis but was hampered by low detection limit and sensitivity.
Among all imaging modalities for imaging marrow infiltration,
Magnetic Resonance (MR) imaging is most sensitive and specific.
T1-weighted (T1),
T2- weighted (T2) and Diffusion Weighted (DW) imaging sequences are used for imaging marrow involvement.
DW imaging is used to assess bone marrow due to its sensitivity to cell density,
the relative content of fat and marrow cells,
water content and bone marrow perfusion.
The signal intensity of DW imaging relies on stochastic Brownian motion or self-diffusion of water molecules at microscopic level within tissues and hence provides information on cellularity of tissue [2,
3,
4].
Objective of this study was to build a MR image based classifier to detect myelomatous lesions.
Classification of volume as healthy or ”diseased with myeloma” has previously been performed with Computed Tomography images [5].
This work presents classification of patients with MM versus healthy individuals using MR images of pelvic bone.
To the best of our knowledge this is first work that attempts the aforementioned task using multi modal MR images.