OPEN Research Support
head

MSc, Radiographer
Janni Jensen
Research & Innovation Unit of Radiology, University of Southern Denmark (SDU)


Project management
Project status    Closed
 
Data collection dates
Start 01.05.2021  
End 02.05.2022  
 



X-ray real world AI validation (X-ValidAIt)

Short summary

The aim of this study is to estimate accuracy and precision of RBhip (an artificial intelligence software) for measuring anatomic angles and joint space width in evaluation of femoroacetabular deformities and osteoarthritis of the hip.


Rationale

Patients with persistent, non-traumatic hip pain often undergo pelvic radiography as part of their diagnostic workup. The most common diagnoses include osteoarthritis of the hip and anatomical deformities, such as hip dysplasia and femoroacetabular impingement (FAI) of both pincer- and CAM-type. Measurements taken of a standardized posterior-anterior radiograph can quantify the anatomical configuration of the pelvis and help to confirm or revoke above-mentioned diagnoses. However, the aforementioned measurements are not always routinely performed and patients may be left with symptoms for years before the right diagnosis is made.

A study of patients with hip dysplasia found that the correct diagnosis is delayed with 42 months on average (range: 0-204 months) and more than three confrontations (range: 0-11) with the healthcare system was needed. Several of the measurements, are associated with subjective variability and studies have shown that the reliability is low to moderate. It has furthermore been reported that pelvic rotation and/or tilt during the radiographic procedure may affect the measured values.

Early diagnosis may postpone, reduce or even prevent the onset of osteoarthritis. To ensure that patients with persistent hip pain can get a timely diagnosis, there is a need to introduce these measurements as a standard when reading pelvic radiographs and routinely assess the radiograph for pelvic rotation and tilt.

RBhip is a software program with algorithms incorporated for automatic measurement of LCEA, the AI, the Alpha angle and the width of the femoroacetabular joint. Based on those measurements a diagnosis of osteoarthritis of the hip and dysplasia is offered if present. RBhip is able to offer a diagnosis immediately after the pelvic radiograph has been taken, thus ensuring a prompt diagnosis benefitting the patient. Prior to introducing RBhip to clinical practice research is needed to ensure that accuracy and precision is comparable to standard clinical practice.


Description of the cohort

Retrospectively collected pelvic radiographs


Data and biological material

Radiographic measurements of Lateral center edge angle, Acetabular index, Alpha angle and Foramen obturator index.


Collaborating researchers and departments

Department of Orthopedic Surgery, Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark

  • Professor Søren Overgaard, MD, Dr. Med

Research & Innovation Unit of Radiology, University of Southern Denmark (SDU)

  • Benjamin Schnack Rasmussen, MD, PhD
  • Associate professor, Head of research Ole Graumann, MD, PhD

Radiobotics, Copenhagen, Denmark

  • Director of Clinical Operations, Michael Lundemann, MSc, PhD
  • Chief Scientific Officer, Martin Axelsen, MSc, PhD