Image based assessment for dental treatment monitoring

A method of dental treatment monitoring, the method comprising:

• accessing a captured two-dimensional (2D) image of a patient's teeth taken at a particular time during a course of dental treatment;
• generating a first teeth mask for the captured 2D image, wherein the first teeth mask is configured to distinguish between tooth and non-tooth regions of the captured 2D image;
• accessing an expected three-dimensional (3D) model representing an expected configuration of the patient's teeth at the particular time;
• determining a set of virtual camera parameters of a virtual camera for the expected 3D model corresponding to the captured 2D image;
• generating an expected 2D image from the expected 3D model by mapping points from a surface of the expected 3D model to points on an image plane of the virtual camera;
• generating a second teeth mask for the expected 2D image, wherein the second teeth mask is configured to distinguish between tooth and non-tooth regions of the expected 2D image;
• aligning the first teeth mask with the second teeth mask, wherein aligning the first teeth mask with the second teeth mask includes taking a distance transform of the first teeth mask, and determining a minimal-cost overlay between the second teeth mask and the distance transform of the first teeth mask;
• comparing the captured 2D image with the expected 2D image, with the first and second teeth masks aligned, to determine whether a configuration of the patient's teeth is within a threshold level of correspondence to the expected configuration of the patient's teeth;
• and
• providing an indication that the dental treatment is proceeding as expected or not proceeding as expected based on the determination as to whether the configuration of the patient's teeth is within the threshold level of correspondence.

A non-transitory computer-readable storage medium storing a set of instructions capable of being executed by one or more processors that, when executed, causes the one or more processors to perform a method comprising:

• accessing a captured two-dimensional (2D) image of a patient's teeth taken at a particular time during a course of dental treatment;
• generating a first teeth mask for the captured 2D image, wherein the first teeth mask is configured to distinguish between tooth and non-tooth regions of the captured 2D image;
• accessing an expected three-dimensional (3D) model representing an expected configuration of the patient's teeth at the particular time;
• determining a set of virtual camera parameters of a virtual camera for the expected 3D model corresponding to the captured 2D image;
• generating an expected 2D image from the expected 3D model by mapping points from a surface of the expected 3D model to points on an image plane of the virtual camera;
• generating a second teeth mask for the expected 2D image, wherein the second teeth mask is configured to distinguish between tooth and non-tooth regions of the expected 2D image;
• aligning the first teeth mask with the second teeth mask, wherein aligning the first teeth mask with the second teeth mask includes taking a distance transform of the first teeth mask, and determining a minimal-cost overlay between the second teeth mask and the distance transform of the first teeth mask;
• comparing the captured 2D image with the expected 2D image, with the first and second teeth masks aligned, to determine whether a configuration of the patient's teeth is within a threshold level of correspondence to the expected configuration of the patient's teeth;
• and
• providing an indication that the dental treatment is proceeding as expected or not proceeding as expected based on the determination as to whether the configuration of the patient's teeth is within the threshold level of correspondence.