A group of researchers led by Lara Requena Bueno and Jose Ignacio Priego Quesada, belonging to the Research Group in Sports Biomechanics (GIBD) of the University of Valencia, has independently published a scientific article on the validation of the ThermoHuman software. This article, entitled "Validation of ThermoHuman automatic thermographic software for assessing foot temperature before and after running" has been published in the Journal of Thermal Biology (impact factor JCR 2019 2,361), in the same line as other works already published about ThermoHuman method (Fernández-Cuevas et al., 2012; Fernández-Cuevas et al., 2016).

Infrared thermography allows two methods of analysis to be carried out: the qualitative one (with visual and intuitive reading of colors) and the quantitative one (thanks to the temperature data contained in the thermogram). This second method allows an objective analysis of the temperature measurements, but has the disadvantage of the excessive time that is used and the lack of reliability associated with manual analysis (Kroese et al., 2018). The authors of this current article aimed to compare both methods in three ways (ThermoHuman software, manual analysis, and manual analysis mimicking the regions established by ThermoHuman).

"ThermoHuman resulted in an 86% reduction in time involved compared to manual delimitation"

Requena-Bueno et al. (2020)

In conclusion, the ThermoHuman software was found to be a valid method that saves time in image analysis with excellent reliability results.

The publication in details

Requena-Bueno et al. (2020) analyzed 120 feet soles thermal images from 30 subjects at 4 different times (before and after running on two different days). Measurement tests consisted of 30 minutes continuous running at 80% of their maximum aerobic speed (MAS) on a treadmill with a 1% slope. The thermograms were obtained using a FLIR E60bx infrared camera, with a resolution of 320x240 pixels.

The measurements were carried out in the laboratory following the TISEM protocol (Gomes Moreira et al., 2017). Before taking the thermal images, the subjects waited 10 minutes sitting with their legs in a horizontal position. The images were taken from a distance of 1 m and perpendicular to the soles of the feet. An anti-reflection panel was placed behind the feet so as to minimize the effects of reflected temperature of the surroundings and to isolate the images of the feet.

As we can seen in figure 1, the thermal images were analyzed using three different approaches: (A) manual analysis (B) automatic analysis with ThermoHuman software and (C) manual analysis imitating ThermoHuman segmentation of the regions of interest (ROI). Each feet sole was divided into 9 ROIs and statistical analyzes were performed to confirm the normality of the data, the differences between methods, the effect size and the correlation coefficients (reliability and reproducibility).

Figure 1. Three methods of analysis (A) manual (B) ThermoHuman (C) manual imitating ThermoHuman (adapted from Requena-Bueno et al., 2020)

The results obtained by the authors show that analysis using ThermoHuman resulted in a reduction of the time. In addition to that, the 88.4% of the images processed had a perfect automatic ROI detection. The 3 analysis approaches showed differences between them, but with a small effect size. Therefore, the reliability between ThermoHuman and manual procedures is excellent.

The limitation described by some studies about the application of infrared thermography in humans and the long time required for the quantitative analysis (Marins J, et al., 2014; Priego-Quesada et al., 2017b), is solved with the use of ThermoHuman software, as it saves 86% of the time spent in the analysis. That is to say: from 8 minutes per image we went to less than 1 minute per thermogram.

Thermal image analysis with ThermoHuman software obtained excellent reliability result (ICC 0.96)

Requena-Bueno et al., 2020

The analyzes performed with ThermoHuman software have a slightly higher intra-class correlation (ICC) (0.96 vs 0.94) than the manual coefficients (Figure 2). Indicating excellent reliability and reproducibility for both procedures. It is important to highlight that, although results suggest that ThermoHuman and manual methods are both valid in themselves, combining them is not recommended due to the differences observed between them.

Figure 2. Adapted from Requena-Bueno and collaborators (2020)

Likewise, it should be mentioned that the results also indicate a margin of improvement in the effectiveness of the ThermoHuman recognition algorithms, since 12% of the images showed some type of error in the delimitation of the ROIs.

Conclusions

This publication helps to demonstrate and consolidate the validity, reliability and reproducibility of the ThermoHuman software. It emphasizes the improvement that it supposes, both in saving time (86% less) and in the excellent results of reliability and reproducibility (ICC of 0.96). In this sense, it reinforces the message and mission of ThermoHuman: aimed at helping professionals who work with thermography and humans, saving time, ensuring validity and reliability, and facilitating the management and visualization of results. However, it is evident the need to continue improving the algorithms to increase the efficiency of automatic recognition, a task on which we continue working.

_____

On behalf of ThermoHuman team, we acknowledge the work performed by GIBD and the University of Valencia in this independent publication.

CONTACT WITH US

Fernández-Cuevas, I., Marins, J. C., Gómez Carmona, P. M., García-Concepción, M. Á., Arnáiz Lastras, J., & Sillero Quintana, M. (2012, 5-8 September). Reliability and reproducibility of skin temperature of overweight subjects by an infrared thermography software designed for human beings. Paper presented at the XII Congress EAT 2012, Porto, Portugal.

Fernández-Cuevas, I., Marins, J. C., Arnáiz Lastras, J., Gómez Carmona, P., & Sillero Quintana, M. (2016). Validity, Reliability, and Reproducibility of Skin Temperature in Healthy Subjects Using Infrared Thermography. In P. Humbert, H. Maibach, F. Fanian & P. Agache (Eds.), Agache’s Measuring the Skin (pp. 1311-1318). Cham: Springer International Publishing.

Gomes Moreira, D., Costello, J.T., Brito, C.J., Adamczyk, J.G., Ammer, K., Bach, A.J.E., Costa, C.M.A., Eglin, C., Fernandes, A.A., Fern ́andez-Cuevas, I., Ferreira, J.J.A., Formenti, D., Fournet, D., Havenith, G., Howell, K., Jung, A., Kenny, G.P., Kolosovas- Machuca, E.S., Maley, M.J., Merla, A., Pascoe, D.D., Priego Quesada, J.I., Schwartz, R.G., Seixas, A.R.D., Selfe, J., Vainer, B.G., Sillero-Quintana, M. (2017). Thermographic imaging in sports and exercise medicine: a Delphi study and consensus statement on the measurement of human skin temperature. J. Therm. Biol. 69, 155–162. https://doi.org/10.1016/j.jtherbio.2017.07.006.

Kroese, L. F., Sneiders, D., Kleinrensink, G. J., Muysoms, F., & Lange, J. F. (2018). Comparing different modalities for the diagnosis of incisional hernia: a systematic review. Hernia, 22(2), 229-242.

Marins, J. C. B., Moreira, D. G., Cano, S. P., Quintana, M. S., Soares, D. D., de Andrade Fernandes, A., ... & dos Santos Amorim, P. R. (2014). Time required to stabilize thermographic images at rest. Infrared Physics & Technology, 65, 30-35.

Priego Quesada, J. I., Kunzler, M. R., & Carpes, F. P. (2017). Methodological aspects of infrared thermography in human assessment. In Application of Infrared Thermography in Sports Science (pp. 49-79). Springer, Cham.

Requena-Bueno, L., Priego-Quesada, J. I., Jimenez-Perez, I., Gil-Calvo, M., & Pérez-Soriano, P. (2020). Validation of ThermoHuman automatic thermographic software for assessing foot temperature before and after running. Journal of Thermal Biology, 92, 102639.