Method for quantifying caries

1. A method for quantifying caries, executed at least in part on data processing hardware, the method comprising:

generating a digital image of a tooth, the image comprising intensity values for a region of pixels corresponding to the tooth, gum, and background;

automatically extracting at least one lesion area from sound regions of a tooth by,

identifying tooth regions and extracting suspicious lesion areas in the identified tooth regions,

selecting a first subset of the extracted suspicious lesion areas,

removing false positives in the first subset of the extracted suspicious lesion areas of the tooth by applying an additional false positive detection to the first subset of the extracted suspicious lesion areas to leave a smaller second subset of the first subset, and

identifying remaining suspicious lesion areas not in the first subset and suspicious lesion areas in the second subset as the extracted at least one lesion area;

identifying an adjacent sound region that is adjacent to the extracted at least one lesion area;

reconstructing intensity values for tooth tissue within the extracted at least one lesion area according to values in the adjacent sound region; and

quantifying the condition of the caries using the reconstructed intensity values and intensity values from the lesion area.

2. The method of claim 1 , wherein generating the digital image of the tooth comprises obtaining a fluorescence image of the tooth.

3. The method of claim 1 , wherein generating the digital image of the tooth comprises:

obtaining a fluorescence image of the tooth;

obtaining a reflectance image of the tooth; and

combining image data for the fluorescence and reflectance images.

4. The method of claim 1 further comprising identifying one or more tooth regions using thresholding.

5. The method of claim 1 , wherein extracting the suspicious lesion area comprises using a marker-controlled watershed algorithm.

6. The method of claim 1 , wherein extracting the suspicious lesion area comprises using a morphological bottom-hat based method along with multi-resolution and surface reconstruction techniques.

7. The method of claim 1 , wherein removing false positives comprises locating interproximal regions, selecting the first subset of the extracted suspicious lesion areas by identifying suspicious lesion areas in the interproximal regions, and removing interproximal false positives to determine the second subset.

8. The method of claim 1 , wherein the step of reconstructing intensity values is performed by a process consisting of one or more of bilinear interpolation, surface fit, and interpolation by solving Laplace's equation.

9. The method of claim 1 , wherein quantifying the condition of the caries comprises calculating the fluorescence loss of the lesion area or calculating the area of the lesion area.

10. A computer program embodied on a non-transitory computer readable medium for use in quantifying caries, the program comprising executable instructions that when loaded on a computer, causes the computer to:

generate a digital image of a tooth, the image comprising intensity values for a region of pixels corresponding to the tooth, gum, and background;

automatically extract at least one lesion area from sound regions of a tooth by,

identifying tooth regions in the digital image,

extracting suspicious lesion areas in the identified tooth regions, and

locating interproximal regions in the identified tooth regions,

identifying one or more suspicious lesion areas connected to the interproximal regions,

removing false positives in the one or more suspicious lesion areas connected to the interproximal regions by applying an additional interproximal caries detection process, and identifying remaining suspicious lesion areas connected to the interproximal regions and suspicious lesion areas not connected to the interproximal regions as the extracted at least one lesion area;

identify an adjacent sound region that is adjacent to the extracted at least one lesion area;

reconstruct intensity values for tooth tissue within the at least one extracted lesion area according to values in the adjacent sound region; and

quantify the condition of the caries using the reconstructed intensity values and intensity values from the at least one extracted lesion area.

11. The computer program of claim 10 wherein generating the digital image of the tooth comprises obtaining a fluorescence image of the tooth.

12. The computer program of claim 10 , wherein generating the digital image of the tooth comprises:

obtaining a fluorescence image of the tooth;

obtaining a reflectance image of the tooth; and

combining image data for the fluorescence and reflectance images.

13. The computer program of claim 10 further comprising identifying one or more tooth regions using thresholding.

14. The computer program of claim 10 , wherein extracting the suspicious lesion area comprises using a marker-controlled watershed algorithm.

15. The computer program of claim 10 , wherein extracting the suspicious lesion area comprises using a morphological bottom-hat based method along with multi-resolution and surface reconstruction techniques.

16. The computer program of claim 10 , wherein the step of reconstructing intensity values is performed by a process consisting of one or more of bilinear interpolation, surface fit, and interpolation by solving Laplace's equation.

17. The computer program of claim 10 , wherein quantifying the condition of the caries comprises calculating the fluorescence loss of the lesion area or calculating the area of the lesion area.

18. The computer program of claim 10 , wherein locating interproximal regions comprises locating interproximal regions that have clear demarcation with the steps of:

applying a distance transformation to a binary image of the digital image of the tooth to locate the pixel with the largest distance measured from the boundaries of the identified tooth regions in the binary image;

assigning the identified tooth region that is connected to the located pixel as a first object;

assigning the identified tooth region that is not connected to the located pixel as a second object; and

defining the interproximal regions to be the pixels in the background having the same distance to the first and second objects.

19. The computer program of claim 10 , wherein locating interproximal regions comprises locating interproximal regions that have no clear demarcation with the steps of:

defining an origin point in a binary image of the tooth;

casting a fan of ray lines from the origin point in a plurality of angles;

defining a contour line at points at which each ray line first encounters the boundary between tooth and background areas,

determining internal and external markers;

applying a marker-controlled watershed transformation to a gradient image of a grayscale version of the digital image with the internal and external markers to form first and second groups of water basins; and

taking the pixels having the same distance to the first and second groups of water basins as interproximal regions.

20. The computer program of claim 10 , wherein locating interproximal regions comprises locating interproximal regions that have no clear demarcation with the steps of:

applying a distance transformation to a binary image of the digital image of the tooth, to form a distance image in which each pixel value represents the closest distance of that pixel to the background of the teeth;

determining internal and external markers using the distance image;

applying a marker-controlled watershed transformation to a gradient image of a grayscale version of the digital image with the internal and external markers to form two groups of water basins; and

taking the pixels having the same distance to the two groups of basins as interproximal regions.