Abstract

Abstract. This defensive publication discloses, as enabling prior art, the complete family of geostatistical variogram-based texture-analysis methods for panoramic dental radiographs (orthopantomograms, OPG) applied to peri-implant bone quality assessment and implant stability quotient (ISQ) estimation. Disclosed, individually and in every combination, are: the Matheron semivariogram γ(h), robust estimators (Cressie–Hawkins, Genton), madogram, rodogram, power-p variograms, covariance/correlogram, and cross-variograms, computed on raw, geometrically normalized, or wavelet/multi-resolution OPG regions of interest; extraction of the bi-logarithmic (log-log) initial slope β, nugget, sill, range, hole-effect, Hurst exponent, and variogram-derived fractal dimension; directional and anisotropic variograms; and nested multi-scale models (spherical, exponential, Gaussian, Matérn). Further disclosed is the mapping of any such descriptor set to a continuous ISQ value (resonance frequency analysis, RFA), to Lekholm–Zarb bone-quality classes (I–IV), or to bone-implant-contact and failure-risk thresholds, by linear/ridge/PLS/SVM/random-forest/gradient-boosting/MLP/CNN models, alone or fused with fractal-dimension, GLCM, wavelet, and Trabecular Bone Score descriptors. The full Cartesian design space {preprocessing × estimator × parameter set × spatial scope × downstream model} is published expressly to constitute prior art under 35 U.S.C. §102/§103 and EPC Articles 54/56 against any later-filed application directed to variogram-based OPG bone or implant-stability analysis. The stationarity violation intrinsic to panoramic acquisition, which constrains the entire family, is analyzed. No convolutional-neural-network feature-extraction architecture, learned-embedding method, or generative implant-design method is disclosed; these are reserved.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Download

Share

COinS