Automated and Reliable Estimation of Patient Height From CT Scan Using an Artificial Intelligence Algorithm

Condition(s) studied

Study summary

The TAILLE study is a single-center, non-interventional (RIPH3), uncontrolled, non-randomized, open-label study.prospective study conducted at CHU Grenoble Alpes.

The goal is to evaluate the reliability of an AI algorithm developed as part of the TAILLE-pilot study to estimate height from abdominal CT scan data and relevant clinical data like age and gender.

The TAILLE-pilot study , is a pilot data study (RNIPH) aimed at developing artificial intelligence algorithms using a retrospective cohort, developed by our team. It is based on AI algorithms trained to predict height from abdominal CT scan volumes.

The TAILLE study therefore aims to prospectively evaluate the performance of this algorithm in a new cohort of adult patients scheduled for an abdominal CT scan. The TAILLE study hypothesize that height can be reliably estimated using this algorithm, potentially with greater accuracy at the individual level than self-reported height, the limitations of which have been documented in the literature.

As part of the TAILLE study, three methods will be compared of height estimation or measurement :

* CRA-measured height: height measured by a Clinical Research Associate (CRA), serving as the reference method;
* AI-estimated height: height estimated by the AI algorithm developed as part of the TAILLE-pilot study;
* Self-reported height: height reported by the patient.

Eligibility

Primary outcome measure(s)

• Assessment of the reliability of an AI algorithm developed as part of the TAILLE-pilot study to estimate height from abdominal CT scan and relevant clinical data, such as gender and sex — At enrollement
  Assessment of the differences between CRA-measured height and AI-estimated height (CRA = Clinical Research Associate). CRA-measured height is the reference

Trial sites (1)

More University Hospital, Grenoble trials in France