Differences

This shows you the differences between two versions of the page.

Link to this comparison view

research [2017/11/13 15:08]
rgrupp3@johnshopkins.edu
research [2017/11/13 15:31] (current)
rgrupp3@johnshopkins.edu
Line 357: Line 357:
  
 ===== Anatomical Model Completion with Statistical Shape Models ===== ===== Anatomical Model Completion with Statistical Shape Models =====
 +{{ :research:model_extrap_heat_map_preview.png?240|}}
 The model completion projects aim to estimate an entire anatomical structure when only partial knowledge is available. The model completion projects aim to estimate an entire anatomical structure when only partial knowledge is available.
  
Line 369: Line 370:
  
 Leave-one-out analysis has shown that skull and mandible surfaces may be extrapolated to within 1.5 mm (RMS surface distance) when over 50% of the structure is unknown. For the orthopedic cases, leave-one-out analysis shows that an entire pelvis surface may be estimated to within 2.5 mm (RMS surface distance) when the partial CT consists of only the acetabular joint region and the superior 5% of the iliac crest. Leave-one-out analysis has shown that skull and mandible surfaces may be extrapolated to within 1.5 mm (RMS surface distance) when over 50% of the structure is unknown. For the orthopedic cases, leave-one-out analysis shows that an entire pelvis surface may be estimated to within 2.5 mm (RMS surface distance) when the partial CT consists of only the acetabular joint region and the superior 5% of the iliac crest.
- 
-TODO insert figure. 
  
 Publications: Publications:
research.txt · Last modified: 2017/11/13 15:31 by rgrupp3@johnshopkins.edu




ERC CISST    LCSR    WSE    JHU