RESUMEN
Increased temperature in humans is the symptom of many infectious diseases and it is thus an important diagnostic tool. Infrared temperature measurement methods have been developed and applied over long periods due to their advantage of non-contact and fast measurements. This study deals with a statistical evaluation of the possibilities and limitations of infrared/thermographic human temperature measurement. A short review of the use of infrared temperature measurement in medical applications is provided. Experiments and statistics-based evaluation to confirm the expected accuracy and limits of thermography-based human temperature measurement are introduced. The results presented in this study show that the standard deviation of the thermographic measurement of the eyes maximum temperature was 0.4-0.9 °C and the mean values differences from the armpit measurement were up to 0.5 °C, based on the used IR camera, even though near ideal measurement conditions and permanent blackbody correction were used. It was also shown that a certain number of outliers must be assumed in such measurements. Extended analyses including simulations of true negative/false positive, sensitivity/specificity and receiver operating characteristics (ROC) curves are presented. The statistical evaluation as well as the extended analyses show that maximum eyes temperature is more relevant than a forehead temperature examination.
Asunto(s)
Rayos Infrarrojos , Termografía , Humanos , Termografía/métodos , Temperatura , Temperatura Corporal , Fiebre/diagnósticoRESUMEN
This study is focused on the quantitative estimation of defect depth by applying pulsed thermal nondestructive testing. The majority of known defect characterization techniques are based on 1D heat conduction solutions, thus being inappropriate for evaluating defects with low aspect ratios. A novel method for estimating defect depth is proposed by taking into account the phenomenon of 3D heat diffusion, finite lateral size of defects and the thermal reflection coefficient at the boundary between a host material and defects. The method is based on the combination of a known analytical model and a non-linear fitting (NLF) procedure. The algorithm was verified both numerically and experimentally on 3D-printed polylactic acid plastic samples. The accuracy of depth prediction using the proposed method was compared with the reference characterization technique based on thermographic signal reconstruction to demonstrate the efficiency of the proposed NLF method.
RESUMEN
Pulsed thermography is a common technique for nondestructive testing (NDT) of materials. This study presents the apparent effusivity method for the quantitative evaluation of coating thickness in a one-sided thermal NDT procedure. The proposed algorithm is based on determining a threshold value of apparent effusivity, which can be found for particular coating-on-substrate structures. It has been found that the square root of the time at which the apparent effusivity curve reaches this threshold is proportional to the coating thickness. The efficiency of the proposed approach is demonstrated by analytical modeling and experimentation performed on thermally-sprayed coatings.