X-ray jaws image enhancement and analysis for numerical evaluation of the degree of bone regeneration

Numerical analysis of jaws bone internal structure is needed to assess the progress in treatment of bone defects, and consequently help surgeons in selecting the appropriate material and procedure to regenerate the bone. One of the methods of assessment of the progress in the bone defects treatment is histomorphometric analysis, which however is of limited use in routine procedures due to its invasiveness (it involves taking a segment out of the bone). In most situations, surgeons use classical radiography to assess the progress in bone regeneration, but also this analysis may be insufficient. This paper describes a newly developed algorithm for quantitative computer analysis of intra-oral radiographs. The bone under treatment and healthy (intact) reference regions are manually marked by surgeon on X-ray image. Then the images inside these regions of interest (ROI) are preprocessed, normalized, filtered using a novel PCA-based filter and thresholded with the threshold adjusted on the basis of the reference image properties. The relative number of above-threshold pixels in the ROI representing the newly regenerated bone is proposed to be a quantitative measure of progress in the bone regeneration processes. Results of statistical analysis of example 117 radiographs taken for 59 patients at 3 different times after the surgery are briefly presented to show the usefulness of the method for objective evaluation of the progress in bone reconstruction.