Validation of ThermoHuman automatic thermographic software for assessing foot temperature before and after running

Highlights

• Analysis using ThermoHuman resulted in a reduction of 86% of the time.

• ThermoHuman presents in 12% of the images an error in its delimitation.

• Differences between manual and automatic definition has a small effect size.

• Reliability between ThermoHuman and both manual procedures was excellent.

Abstract

The aim of the study was to evaluate an automatic thermographic software package (ThermoHuman®) for assessing skin temperature on the soles of the feet before and after running and to compare it with two manual definitions of the regions of interest (ROIs). 120 thermal images of the soles of the feet of 30 participants, at two measurement points (before and after running 30 min) and on two measurement days were analyzed. Three different models of thermographic image analyses were used to obtain the mean temperature of 9 ROIs: A) ThermoHuman (automatic definition of ROIs using ThermoHuman® software), B) Manual (manual delimitation of ROIs by proportion criteria), and C) Manual-TH (manual delimitation of ROIs in an attempt to replicate the regions analyzed by ThermoHuman). ThermoHuman resulted in an 86% reduction in time involved compared to manual delimitation. Fourteen of the 120 images (12%) presented some error in one or more of the ROI delimitations. Although the three procedures presented significant differences between them (53% in the comparison between ThermoHuman and Manual, 47% between ThermoHuman and Manual-TH, and 28% between Manual and Manual-TH), all differences had a small effect size (ES 0.2–0.4) or lower (ES < 0.2). Bland-Altman plots showed similar 95% limits of agreement between the three procedures before and after running. Intraclass correlation coefficient analysis of the three procedures presented excellent reliability (ICC>0.8). In conclusion, ThermoHuman® software was observed to be time-saving for image analysis with excellent reliability. 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.

Introduction

Infrared thermography is a non-invasive and accurate technique for measuring skin temperature (de Andrade Fernandes et al., 2014; Hildebrandt et al., 2010). The popularity of its human applications has increased due to technological advances that result in lower instrumentation purchasing costs (Priego Quesada et al., 2017a; Priego Quesada and Carpes, 2019). Its human applications include detecting injuries and pathologies (Hildebrandt et al., 2010; Ring and Ammer, 2012), studying the effect of clothing and equipment (Fournet and Havenith, 2017; Priego Quesada and Carpes, 2019), and thermoregulatory analyses in different environmental and sport scenarios (Formenti et al., 2017). Although infrared thermography presents some advantages over other methods for measuring skin temperature, as it is a distance technique with the possibility of defining large regions of interest (ROIs) in order to obtain a more robust mean temperature, it does have some limitations (de Andrade Fernandes et al., 2014; Priego Quesada et al., 2017b). One of them is the long time required to analyze thermal images due to the manual definition of the ROIs used in most cases (de Andrade Fernandes et al., 2014; Priego Quesada et al., 2017b).

Defining the ROIs can be undertaken either manually or using software with automatic selection. For manual definition, researchers define their own criteria to ensure accuracy usually based on anatomical proportions or body segments (Fernández-Cuevas et al., 2015; Priego Quesada et al., 2017b). However, this method of definition may affect the reliability of measurements due to the human error (Fernández-Cuevas et al., 2015; Priego Quesada et al., 2017b). For this reason, automatic software could be a desirable option. Software packages commonly process thermal images using masks, which provide an outline of the regions of interest (Lahiri et al., 2012). Although most of these software packages are programed by researchers and are not accessible to everyone, there are now some commercial automatic analysis software packages available for human applications. ThermoHuman®, for example, is an online software package that uses machine learning algorithms to define ROIs (Fernández-Cuevas et al., 2017). However, there is a lack of evidence about the differences between the manual method and the automatic definition of the ROIs.

Traditionally, mean temperature has been obtained from multiple ROIs of the foot (Gil-Calvo et al., 2017; Wang et al., 2018), which could be affected by size or differences in delimitation (Priego Quesada et al., 2015). In baseline temperatures, the toes of the foot can present similar temperatures to surrounding room temperature which complicates the determination of its ROIs (Gil-Calvo et al., 2017). It is, therefore, interesting to investigate the differences between these types of analysis in sports scenarios in which the increase of hot spots at the region of interest may increase due to vasodilation (Formenti et al., 2013; Priego Quesada et al., 2015), especially as the foot presents a higher skin temperature increase (between 5 and 9 °C) after exercise compared to other proximal body regions (between 1 and 5 °C) (Formenti et al., 2017; Gil-Calvo et al., 2019). The aim of this study was, therefore, to compare the manual and automatic definition (using ThermoHuman® software) of ROIs on the soles of the foot, before and after running. It was hypothesized that the automatic procedure would present a shorter time required for the analysis than a manual procedure. In addition, greater differences were expected in the values obtained between the automatic and manual procedure with similar ROIs, but with anatomical criteria to define them, than if the manual procedure tried to copy the delimitation performed by the automatic software.