A novel methodology using high resolution thermal imaging to detect vertebral fractures in children with osteogenesis imperfecta

Salis Alexandra Fane De; Reza Saatchi; Paul Dimitri

doi:10.1530/endoabs.45.OC6.5

OC6.5

Prev Next

Section Contents Cite

Endocrine Abstracts (2016) 45 OC6.5 | DOI: 10.1530/endoabs.45.OC6.5

BSPED2016 Oral Communications Oral Communications 6- Endocrine (9 abstracts)

A novel methodology using high resolution thermal imaging to detect vertebral fractures in children with osteogenesis imperfecta

Alexandra Fane De Salis ¹ , Reza Saatchi ² & Paul Dimitri ^2,

Views

Author affiliations

¹The University of Sheffield, Sheffield, UK; ²Sheffield Hallam University, Sheffield, UK; ³Sheffield Children’s NHS Foundation Trust, Sheffield, UK.

Background: Vertebral compression fractures are common in children with osteogenesis imperfecta. Current imaging methods for fracture detection (X-ray and DXA) use ionising radiation. High Resolution Thermal imaging (HRTI) is a non-invasive, non-ionising method that detects infrared radiation energy emissivity to an accuracy of 0.04 °C, providing a quantitative and qualitative map of temperature distribution. Given that the alteration in blood flow in vertebral fractures acutely and chronically results in temperature change we hypothesised that HRTI may detect thermal variation in vertebral fractures in patients with OI.

Methodology: Prior to routine DXA and X-ray scans, twelve consented participants aged 5–18 with OI and known vertebral fractures had metal discs (diameter 0.85 cm) fixed at 4 cm to the right of each spinous process.

Thermal imaging of each subject’s back was performed for two minutes at a capture rate of thirty frames per second using the Flir T630sc HRTI camera. Genant semiquantative (GQ20) visual grading was used to define vertebral fracture position and severity on DXA or radiographs and then correlated with the metal disc postion on thermal images (figure). TI recordings were compared to DXA and X-ray images using MATLAB® (Matrices Laboratory: version 8.3, R2014a). The temperature percentage change (TPC) of fractured and non-fractured vertebrae relative to adjacent skin areas (region of reference:ROR) was analysed using paired t-tests.

Results: Mean age of the patients was 11.7 (3.9). Two-thirds of the cohort had OI-type I, with the remaining classed as type III or IV. TPC between fractured thoracic vertebrae (n=11) and the ROR was significant (1.27% (1.07): P=0.007, CI:0.45 to 2.09) whereas TPC between the non-fractured vertebrae and ROR was not (0.97% (1.01): P=0.064, CI:−0.08 to 2.03). Lumbar fractures (n=3) were too few to draw conclusions. TPC of the least severe fractures (GQ20-D1) was also significant (1.14% (1.14): P=0.01, CI:0.36, 1.92), implying that the least severe fractures could be identified on TI.

Conclusions: Preliminary work suggests HRTI may be a novel non-radiation approach to identifying vertebral fractures in OI. Larger studies are needed to substantiate current findings and to determine whether fracture severity and position influence thermal emissivity.