In-Field-of-View Thermal Image Calibration System for Medical Thermography Applications

Abstract

Medical thermography has become ever more accessible to hospitals, medical research, and clinical centers with the new generation of thermal cameras, which are easier to use and lower in cost. Some diagnostic techniques using thermal cameras are now regarded as standardized, such as the cold challenge test for Raynaud's phenomenon. The future for medical thermography appears to be improved accuracy, standardization, and establishment as a mainstream medical imaging methodology. Medical thermography standardization, quantitative measurements, image comparison, and multi-center research trials all require thermal cameras to provide a demonstrably traceable, accurate, and reliable temperature output. To this end, the National Physical Laboratory (NPL) has developed a multi-fixed-point source that serves as an in-image calibration system, thereby providing a reliable means for radiometric image validation. An in-field-of-view fixed-point validation system for thermal imaging has successfully been developed, tested, and validated at NPL and has undergone field trials at three clinical centers in the UK. The sources use the phase change plateaux of gallium-zinc eutectic, gallium, and ethylene carbonate. The fixed-point sources have an estimated cavity emissivity of greater than 0.998, a plateau longevity of nominally 3 h at ambient conditions, a stability of 0.1 • C, or better, over that period, a repeatability of 0.1 • C or better, and an estimated temperature uncertainty of ±0.4 • C (k = 2). In this article, the source specifications and design as well as testing, validation, and field trial results are described in detail.

Calibration and validation of thermal imagers

“…These are repeatable to ~0.1 °C and are small enough to be located in the image plane. In 2006 the performance of the sources was validated in trials in clinical centres around the UK [9]. • Temperature measurement "best practice" has been provided through a teaching module on the IR thermography in medicine course run by the University of Glamorgan [10].…”

“…Although certainly better than no reference source, the blackbody used, because it relied on a contact thermometer, could be prone to long term drift and would need regular recalibration to maintain performance and in particular traceability to ITS-90. A more radical approach was followed in the developments reported here [9]. Instead of relying on a contact thermometer sensor to provide traceability to ITS-90, the blackbody sources constructed were based upon fixed-point materials.…”

“…The usefulness of these fixed points is immediately apparent as they spanned the temperature range of most ordinary clinical procedures. Each cavity required a particular initiation procedure (reported in [9]) but the performance was similar and flat melting and freezing plateaux (within 0.1 °C) of 1-3 hours duration were obtained. The assigned radiometric temperature uncertainty of each fixed-point was ±0.4°C (k=2), but this could, with care, be improved.…”

A Roadmap for Thermal Metrology

“…The quality of both these techniques is strongly dependent on good thermal measurements. (For example, the medical sector, in particular, is increasingly requiring quantitative thermal image measurements to facilitate cross-center data exchange and as an aid to more reliable diagnosis [6]). …”

Core and local skin temperature: 3–24 months old toddlers and comparison to adults

“…Some data are historical and are no longer applicable as infants anthropometry has changed significantly in the last three decades [41]. Some utilised measuring techniques are questionable, some used thermistor probes attached to the skin with adhesive tape that effectively increases the temperature in the area [30], while others used thermal cameras at a long distance [33] without reference temperature points for the correction of errors [42]. Furthermore current thermal comfort models do not apply to children [34,36].…”