Skip to main content
Search

University Children’s Hospital Basel

1 related articles

Subscribe to updates
  1. Scoliosis CT labels without hand annotation: bending healthy spines into pathological shapes lifts vertebra Dice from 0.765 to 0.842

    Vertebra segmentation in severe scoliosis has been limited by scarce annotated pathological CT, since public data contain almost no scoliotic cases. An anatomy-aware synthesis pipeline jointly deforms public healthy spine CT scans and their segmentation labels into scoliosis-like volumes, then trains nnU-Net CNN and PRIMUS models under matched compute. On 18 in-house adolescent scoliosis CTs, the best CNN setting raised mean Dice from 0.765 to 0.842 and vertebra success rate from 73.2% to 81.6%; on 55 surgeon-defined apical labels, mean Dice rose from 0.720 to 0.854.

You’re all caught up