Clinical ultraviolet dosimetry with a CCD monochromator array spectroradiometer

Single monochromator charge-coupled device (CCD) array spectroradiometers have the advantage of ease of use and speed compared with double grating instruments. Their inherently inferior stray-light rejection, however, can critically affect their accuracy in phototherapy and research-related dosimetry applications. This paper shows that without adequate correction the HR4000 (Ocean Optics Inc., Dunedin, USA) array device can overestimate the CIE erythema-weighted irradiance of common phototherapy sources and solar simulator beams by over 100%. A software stray-light correction (Ylianttila et al 2005 Photochem. Photobiol. 81 333–41), using the measured slit function of the HR4000, has been applied to spectra acquired from sources used in phototherapy and photobiology (PUVA, UV21, TL01 and solar simulator). The resulting corrected erythema-weighted irradiance measurements from the HR4000 are within 10% of those from a DM150 double grating spectroradiometer (Bentham Instruments Ltd, Reading, UK). A simple model is considered for combining estimates of measurement uncertainties. The importance of exposure bracketing to improve the dynamic range of the HR4000 is illustrated, along with the difficulty in making direct comparison of spectral values between two instruments due to wavelength scale uncertainties. Comparison with a double grating instrument in a solar simulator beam is examined here as a basis for validating CCD array device measurements. The study demonstrates that the HR4000 array spectroradiometer can provide an adequate level of accuracy for common phototherapy and photobiology applications only where a suitable stray-light correction is carefully applied and where the instrument's effective dynamic range is improved.