3D modeling of detected pleural thickenings through thin plate spline interpolation

Pleuramesothelioma is a malignant tumor. A computer-assisted diagnosis system shall provide physicians with the assessment of the detected pleural thickening from the 3D CT data. This paper describes a new, improved method to assess the size of detected thickening using both the 2D and 3D thin plate spline. First, a coordinate transformation was applied. Next, set of landmarks was selected as input for the thin plate spline. The final step is the numerical area integration. The results show that the calculated area was more accurate than the former pixel counting technique, and promise further automatic development.