14.06.2017     13:00–14:30

A3
Statistical modelling and data mining

Type:
Advanced educational workshop
Room:

Statistical modelling of morphological and functional parameters potentially enables a significantly higher efficiency of complex patient specific modelling and model based CAOS. The increase of efficiency is related e.g. to a reduced requirement of e.g. acquisition and/or time consuming interactive segmentation of patient specific image data. Moreover, annotated statistical models enable the efficient and even automatic fusion of patient specific morphology with generic apriori knowledge models on functional aspects such as biomechanics as well as on surgical procedures, instruments or implants respectively. Whereas early work suffered from limited availability of statistical knowledge on pathological situations (e.g. 3D image data on anatomy), the broader clinical application of CAOS techniques also led to a more systematic digital archiving and availability of big amount of digital (3D) data on pathologies providing a more solid basis for statistical modelling. The workshop is conducted by specialists with many years of personal experience in statistical modelling in CAOS, providing examples from different perspectives.

14.06.2017
13:00–14:30

A3
Statistical modelling and data mining

Type:
Advanced educational workshop
Room:

Statistical modelling of morphological and functional parameters potentially enables a significantly higher efficiency of complex patient specific modelling and model based CAOS. The increase of efficiency is related e.g. to a reduced requirement of e.g. acquisition and/or time consuming interactive segmentation of patient specific image data. Moreover, annotated statistical models enable the efficient and even automatic fusion of patient specific morphology with generic apriori knowledge models on functional aspects such as biomechanics as well as on surgical procedures, instruments or implants respectively. Whereas early work suffered from limited availability of statistical knowledge on pathological situations (e.g. 3D image data on anatomy), the broader clinical application of CAOS techniques also led to a more systematic digital archiving and availability of big amount of digital (3D) data on pathologies providing a more solid basis for statistical modelling. The workshop is conducted by specialists with many years of personal experience in statistical modelling in CAOS, providing examples from different perspectives.


13:00–13:20 Statistical shape modeling for CT-based automated tha planning
Yoshinobu Sato (Osaka/JP)


13:20–13:40 Comprehensive machine-learning based approach for mulsculoskeletal image analysis
Guoyan Zheng (Bern/CH)


13:40–14:00 The construction of statistical shape models and their application to bone surface reconstructions in the knee
Christoph Hänisch (Aachen/DE)


14:00–14:20 Interactive image segmentation using statistical shape models
Marcel Lüthi (Basel/CH)


14:20–14:30 Discussion