PROJECT TITLE :

Multiple Vital-Sign-Based Infection Screening Outperforms Thermography Independent of the Classification Algorithm

ABSTRACT:

Goal: Thermography-based mostly infection screening at international airports plays an important role within the prevention of pandemics. However, studies show that thermography suffers from low sensitivity and specificity. To achieve higher screening accuracy, we developed a screening system based mostly on the acquisition of multiple important-signs. This multimodal approach will increase accuracy, but introduces the need for subtle classification methods. This paper presents a comprehensive analysis of the multimodal approach to infection screening from a Machine Learning perspective. Methods: We tend to conduct an empirical study applying six Classification Algorithms to measurements from the multimodal screening system and comparing their performance among every different, additionally as to the performance of thermography. Additionally, we tend to offer an data theoretic read on the employment of multiple vital-signs for infection screening. The classification ways are tested using the same clinical knowledge, which has been analyzed in our previous study using linear discriminant analysis. A total of ninety two subjects were recruited for influenza screening using the system, consisting of fifty seven inpatients diagnosed to have seasonal influenza and 35 healthy controls. Results: Our study revealed that the multimodal screening system reduces the misclassification rate by more than fifty% compared to thermography. At the same time, not one of the multimodal classifiers required more than six ms for classification, which is negligible for practical functions. Conclusion: Among the tested classifiers k-nearest neighbors, support vector machine and quadratic discriminant analysis achieved the highest cross-validated sensitivity score of ninety three%. Significance: Multimodal infection screening may be able to deal with the shortcomings of thermography.