Outbreak detection model based on danger theory

Abstract

a b s t r a c tIn outbreak detection, one of the key issues is the need to deal with the weakness of early outbreak signals because this causes the detection model to have has less capability in terms of robustness when unseen outbreak patterns vary from those in the trained model. As a result, an imbalance between high detection rate and low false alarm rate occurs. To solve this problem, this study proposes a novel outbreak detection model based on danger theory; a bio-inspired method that replicates how the human body fights pathogens. We propose a signal formalization approach based on cumulative sum and a cumulative mature antigen contact value to suit the outbreak characteristic and danger theory. Two outbreak diseases, dengue and SARS, are subjected to a danger theory algorithm; namely the dendritic cell algorithm. To evaluate the model, four measurement metrics are applied: detection rate, specificity, false alarm rate, and accuracy. From the experiment, the proposed model outperforms the other detection approaches and shows a significant improvement for both diseases outbreak detection. The findings reveal that the robustness of the proposed immune model increases when dealing with inconsistent outbreak signals. The model is able to detect new unknown outbreak patterns and can discriminate between outbreak and non-outbreak cases with a consistent high detection rate, high sensitivity, and lower false alarm rate even without a training phase.

An algorithm applied to national surveillance data for the early detection of major dengue outbreaks in Cambodia

“…Their method was able to predict the occurrence of an outbreak 3.1 ± 2.2 weeks in advance and 2.9 ± 2.4 weeks in advance for the first and second pattern, respectively [21]. A Malaysian team developed an algorithm based on danger theory using climatic data in addition to referrals and dengue registered cases, with an average sensitivity of 77% and an average specificity of 99% [22].…”

“…The main limit of our study is that the algorithm used needs to be trained, which may cause a loss of robustness in early phases of surveillance, or if the outbreak pattern changes or differs significantly from previous years. The 2015 outbreak in Kampong Chhnang province, for instance, was reported later than usual and was not detected in a timely way by our algorithm [22]. Robustness could also be lower because the method requires a sample of outbreak and non-outbreak periods.…”

“…Robustness could also be lower because the method requires a sample of outbreak and non-outbreak periods. If any change occurs in the sensor’s outbreak definition, the algorithm would need to be reviewed or changed entirely [22]: Because of the learning phase, the algorithm cannot face rapid changes in epidemic patterns. Such changes, however, are unlikely in the short/medium term in Cambodia.…”

An Evaluation of Feature Selection Technique for Dendrite Cell Algorithm

“…After that, the attributes are normalized according DCA signal. The cumulative sum normalization technique is adopted to normalized DCA signal [10] .…”

Validation on an Enhanced Dendrite Cell Algorithm using Statistical Analysis

“…Then, a fully functioning real-time network intrusion detection system was implemented in the subsequent year [8]. Based on its success, DCA has been widely applied in various areas such as fault [9], intrusion [10], fraud [11], and outbreak detection [12]. The published results of these applications demonstrate that DCA performs well in terms of producing a high detection rate and lower false detection rate in comparison to other systems.…”