ABSTRACT

This paper presents a novel approach for sinusitis detection by means of thermography. So far, the role of passive static thermography (PST) for diagnosing the sinusitis has been remarked by many studies. But, it has not been proved as an effective approach for sinusitis detection. With this aim, the automatic and active dynamic thermography (ADT) based approach has been proposed. The study has four steps: (a) data acquisition, PST and ADT of 19 control subjects and 16 sinusitis patients; (b) automatic extraction of maxillary regions; (c) thermal data processing; and (d) thermal data analysis. The effectiveness of ADT and PST approach has been evaluated by comparing their outcomes with physical examinations. Consequently, PST shows failure in the detection of maxillary sinusitis, whereas the outcomes of ADT highly correlate with physical examinations. The frequency analysis of ADT data shows 90% accuracy for sinusitis detection with sensitivity and specificity of 77.27% and 95.83%, respectively. In ADT, it has been observed that the thermal profiles of sinusitis patients are significantly different from those of control group during self-warming phase. Hence, the study shows the encouraging results towards the application of ADT in the diagnosis of sinusitis.

摘要

本文提出了一种通过热成像检测鼻窦炎的新方法。到目前为止，被动静态热成像 (PST) 在诊断鼻窦炎方面的作用已被许多研究所提及。但是，它尚未被证明是鼻窦炎检测的有效方法。为此，提出了基于自动和主动动态热成像 (ADT) 的方法。该研究分为四个步骤：(a) 19 名对照受试者和 16 名鼻窦炎患者的数据采集、PST 和 ADT；(b) 自动提取上颌区域；(c) 热数据处理；(d) 热数据分析。ADT 和 PST 方法的有效性已通过将其结果与体检结果进行比较来评估。因此，PST 显示无法检测上颌窦炎，而 ADT 的结果与体检高度相关。ADT 数据的频率分析显示鼻窦炎检测准确率为 90%，灵敏度和特异性分别为 77.27% 和 95.83%。在 ADT 中，已经观察到鼻窦炎患者在自我升温阶段的热分布与对照组的显着不同。因此，该研究显示了 ADT 在鼻窦炎诊断中应用的令人鼓舞的结果。