New Computerized Method in Measuring the Sagittal Bowing of Femur from Plain Radiograph-A Validation Study

Chen-Kun Liaw, Yen-Po Chen, Tai-Yin Wu, Chiou-Shann Fuh, Ruey-Feng Chang

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)


BACKGROUND: Mismatch of intramedullary nails with the bowing of femur is a frequent clinical finding. Previous studies showed inconsistent results.

METHODS: We present an algorithm of region growing territory method to get the radii of the anterior bowing of femur. We also tested it on ten radiographs. Plain radiographs of the lateral view of femur from five men and five women taken between January and August 2014 in Taipei Hospital were chosen randomly. The curvature of femur outline and medullary canal were measured for three times each. Radii of curvature of whole femur, proximal, middle and distal parts were calculated and analyzed.

RESULTS: The coefficient of variation of the 240 measurements ranged from 0.007 to 0.295 and averaged 0.088. The average radii of curvature of the whole, proximal, middle, and distal femur were 1318 mm, 752 mm, 1379 mm, and 599 mm, respectively. At the distal part of the femur, the radius of curvature of the femur outline (452 mm) was smaller than the medullary canal (746 mm) (p < 0.05). Women's femur was straighter than men's when we compared the whole length (1435 mm vs. 1201 mm, p < 0.05). The radii we calculated were smaller than the current intramedullary nails.

CONCLUSION: The results showed that the inter-observer and intra-observer differences are acceptable, support the impression that different bowing conditions existed for Asians as compared to Caucasians, and also indicate the mismatch of current instruments to the curvature of femur.

Original languageEnglish
JournalJournal of Clinical Medicine
Issue number10
Publication statusPublished - Oct 3 2019


Dive into the research topics of 'New Computerized Method in Measuring the Sagittal Bowing of Femur from Plain Radiograph-A Validation Study'. Together they form a unique fingerprint.

Cite this