Carestream Health, Inc.

Jiayong Yan

Victor C Wong

Liangliang Pan

Wei Wang

Lixing Shi

Yingqian Wu

Liwei Song

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; extracting a lesion area from sound tooth regions by identifying tooth regions, extracting suspicious lesion areas, and removing false positives; identifying an adjacent sound region that is adjacent to the extracted lesion area; reconstructing intensity values for tooth tissue within the 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.

Method for quantifying caries

A method for locating one or more interproximal tooth regions in a digital tooth image. The method can be executed at least in part on data processing hardware. The method includes generating the digital tooth image from a fluorescence image of one or more teeth and a reflectance image of the one or more teeth, so as to combine image data from the fluorescence and reflectance images. The digital tooth image has intensity values for pixels corresponding to the one or more teeth and background. The method identifies one or more tooth regions by processing the digital tooth image and locates the one or more interproximal tooth regions according to the one or more identified tooth regions.

Method for locating an interproximal tooth region

A method for identifying tooth regions. The method includes generating a first threshold image from a first tooth image by selecting intensity data values higher than a first predetermined threshold value c; generating a second threshold image from a second tooth image by selecting intensity data values higher than a second predetermined threshold value c; generating a preliminary tooth regions image that defines at least a first tooth region from the intersection of the first and second threshold images; generating a reference binary image from the first image by selecting intensity data values higher than a third predetermined threshold value c, wherein threshold value cexceeds c; and generating a refined tooth regions image from at least the first tooth region in the preliminary tooth regions image. The first tooth region is connected to objects in the reference binary image.

Method for identifying a tooth region

A method for intraoral imaging obtains a digital image of one or more teeth and detects first and second boundaries. At each of the first and second boundaries, there is calculated a boundary ratio of mean gray-scale values for the tooth area on one side of the boundary to mean gray-scale values for background areas on the other side. The calculated boundary ratios are stored. A third ratio of the mean gray-scale values for the tooth area near the first boundary to the mean gray-scale values for the tooth area near the second boundary is calculated and stored. A vector is formed and stored that contains at least the calculated boundary ratios and the third ratio. The tooth surface is classified as either smooth or occlusal according to the stored vector. The image data is processed according to the classification and processing results are reported.

Method for tooth surface classification

A method for extracting a carious lesion area from sound regions of a tooth. In one embodiment, the method includes generating a digital image of the tooth; identifying tooth regions; extracting a suspicious lesion area by using a marker-controlled watershed algorithm or by using a morphological bottom-hat based method along with a multi-resolution surface reconstruction method; and removing false positives. In another embodiment, the method includes generating a digital image of the tooth comprising obtaining a fluorescence image of the tooth, obtaining a reflectance image of the tooth, and combining image data for the fluorescence and reflectance images; identifying tooth regions; extracting suspicious lesion areas; and removing false positives.

Method for extracting a carious lesion area

A method for identification of caries, executed at least in part on data processing hardware, obtains an original digital tooth image that has a plurality of color channels and generates an adjusted image by adjusting intensity values of the original digital tooth image to a range between a minimum value and a maximum value, wherein the adjusted image has at least a green channel image Iwgreen, a red channel image Iwred, and a blue channel image Iwblue. One or more tooth regions are segmented from gum and background regions within the adjusted image according to a relationship between two or more of the images Iwgreen, Iwred, and Iwblue to each other. One or more caries lesions is identified according to pixel intensity values from within the one or more segmented tooth regions, and the one or more tooth regions and the displayed and identified caries regions are highlighted.

Growing community of inventors

Shanghai, China

Jiayong Yan