Medical imaging devices, methods, and systems
Devices, methods, and systems for providing optical imaging to detect and characterize anatomical and/or physiological indicators, such as, rheumatoid arthritis, and devices, methods and systems for computer aided detection and diagnosis of tomographic images. Embodiments for optimizing machine classification of tissue samples are described. Embodiments for using machine classification techniques to classify indicators present in optical tomographic images are described.
1. A method of classifying, with respect to a predefined disease, an optical tomographic image of a living sample tissue, comprising:
scanning light into tissue samples and capturing trans-illumination data from the tissue samples;
using optical tomographic imaging to generate a first set of images of the tissue samples from the trans-illumination data, including diseased tissue samples having the predefined disease and healthy tissue samples not having the predefined disease;
extracting, from the first set of images, a plurality of features selected from the group consisting of volumetric features and projection dependent features, the features representing optical properties of a sample tissue;
statistically analyzing each of the features and selecting responsively to a result of the statistically analyzing, a subset of the features that provides greater predictive accuracy as to the presence of the disease than other features when applied to a predefined classifier;
scanning light into further tissue samples and capturing further trans-illumination data from the further tissue samples;
using optical tomographic imaging to generate a second set of images of the further tissue samples from the further trans-illumination data; and
using the subset of features with the predefined classifier to classify the further tissue samples as having the having the predefined disease or not having the predefined disease based on the second set of images,
wherein the images in the first and second sets include, for each tissue sample, multiple structured images combined into a planar image from which the at least one of the features is extracted.
2. The method of claim 1 , wherein the statistically analyzing includes applying an evolution algorithm in which an initial feature combination is chosen as the first generation parents; all possible mutants are generated by adding, dropping, or replacing one feature from the parent combination; of all the mutants and their parent, the combination with the highest Youden index become the parent in next generation; and repeating until the Youden index stops improving.
3. The method of claim 1 , further comprising reconstructing at least one image of optical parameters of the tissue.
4. The method of claim 3 , wherein the reconstructing includes reconstructing raw absorption and scattering data as at least one three-dimensional image of absorption and scattering coefficients in the tissue.
5. The method of claim 4 , wherein the reconstructing of the three-dimensional image of absorption and scattering coefficients in the tissue includes a prediction model of transmission data exiting from the tissue.
6. The method of claim 5 , wherein the prediction model of transmission data includes solving at least one frequency-domain equation of radiative transfer.
7. The method of claim 6 , wherein the solution of at least one frequency-domain equation of radiative transfer includes establishing at least one partially-reflective boundary condition.
8. The method of claim 4 , wherein the reconstructing of the three-dimensional image of absorption and scattering coefficients in the tissue includes a model-based iterative image reconstruction algorithm.
9. The method of claim 8 , wherein the reconstruction includes using at least one transmission data normalized to a sum of each detection point over all illumination points.
10. The method of claim 8 , wherein the reconstruction includes using a PDE constrained sequential quadratic programming (SQP) algorithm.
11. The method of claim 10 , wherein using a PDE-constrained SQP algorithm includes a preconditioning of a Hessian matrix.
12. The method of claim 11 , wherein the preconditioning of the Hessian matrix includes a linear coordinates transformation (LCT) of absorption and scattering coefficients in the tissue.
13. The method of claim 12 , wherein the reconstruction of the PDE-constrained SQP algorithm includes assigning values of 0.3 cm′ and 8 cm′ as an initial guess of absorption and scattering coefficients, respectively, in the tissue.
14. The method of claim 4 , wherein the reconstructing of the three-dimensional data of absorption and scattering coefficients in the tissue includes an unstructured computation mesh of the tissue.
15. The method of claim 3 , further comprising reconstructing hemoglobin concentration and/or oxygen saturation as a three-dimensional image of optical parameter coefficients in the tissue.
16. The method of claim 3 , wherein reconstructing the at least one image of optical parameters includes using at least one transmission data normalized to a sum of all detection points over all illumination points, divided by the total number of measurements.
17. The method of claim 16 , wherein the reconstructing includes using a preconditioning of a gradient or Hessian matrix, which makes the image reconstruction process numerically more stable.
18. The method of claim 17 , wherein the preconditioning of the gradient or Hessian matrix includes a linear transformation of variables (LTV) for the reconstruction parameters.
19. The method of claim 3 , wherein the reconstructing includes using a generalized radial basis function (RBF) based smoothing operator that eliminates “grid effects” due to variation in the cell size of unstructured grids.
20. The method of claim 1 , wherein the predefined disease is rheumatoid arthritis.
21. The method of claim 1 , wherein the planar image is a projection of a three-dimensional optical property map.
22. The method of claim 1 , wherein at least one of the features is a mixture model.
23. The method of claim 22 , wherein the at least one of the features is a Gaussian mixture model.
24. The method of claim 22 , wherein the at least one of the features is multiple components of a Fourier decomposition of the planar image.
25. The method of claim 1 , wherein the planar image is generated from a structured set of raw images representing optical property maps of transverse, sagittal, or coronal sections through a tissue sample projected on a two-dimensional surface.
26. The method of claim 1 , wherein the first set of images represent absorption or scattering coefficients derived from frequency domain diffuse optical tomography and the features include ones derived from the first set of the images by parameterization of curvilinear function fits.
27. The method of claim 1 , wherein the first set of images represent absorption or scattering coefficients derived from frequency domain diffuse optical tomography and the features include spectral coefficients derived from the first set of images.
28. The method of claim 1 , wherein the first set of images represent absorption or scattering coefficients derived from frequency domain diffuse optical tomography and the features include modulation frequencies used to derive the first set of images.
29. The method of claim 1 , wherein the first set of images represent at least one of oxy-hemoglobin concentration, deoxy-hemoglobin concentration, total blood volume, and oxygen saturation derived from frequency domain diffuse optical tomography and the features include features derived from the first set of images by parameterization of curvilinear function fits.
30. The method of claim 1 , wherein the predefined classifier includes one of a nearest neighbor method (KNN), a discriminate analysis (DA), a support vector machine (SVM) method and a self-organizing maps (SOMs) method.
31. An apparatus that classifies, with respect to a predefined disease, an optical tomographic image of a living sample tissue, the apparatus comprising:
a processor programmed to implement a classifier based on data obtained by
scanning light into tissue samples and capturing trans-illumination data from the tissue samples,
using optical tomographic imaging to generate a first set of images of the tissue samples from the trans-illumination data, including diseased tissue samples having the predefined disease and healthy tissue samples not having the predefined disease,
extracting, from the first set of images, a plurality of features selected from the group consisting of volumetric features and projection dependent features, the features representing optical properties of a sample tissue, and
statistically analyzing each of the features and selecting responsively to a result of the statistically analyzing, a subset of the features that provides greater predictive accuracy as to the presence of the disease than other features when applied to a predefined classifier;
a scanning light source that scans further light into further tissue samples; and
a photodetector that captures further trans-illumination data from the further tissue samples;
wherein the processor is further programmed to
use optical tomographic imaging to generate a second set of images of the further tissue samples from the further trans-illumination data, and
use the subset of features with the predefined classifier to classify the further tissue samples as having the having the predefined disease or not having the predefined disease based on the second set of images,
wherein the images in the second set includes, for each tissue sample, multiple structured images combined into a planar image from which the at least one of the features is extracted.