ABSTRACT
New Zealand has long been praised for the effectiveness of its COVID-19 elimination strategy. It resulted in fewer COVID-19-related deaths, better economic recovery, and less stringent policy measures within its borders compared with other OECD countries, which opted for mitigation or suppression. However, since September 2021, the rising number of infections has not been contained anymore by the contact tracing and self-isolation system in place and the government has shifted towards a policy strategy similar to suppression to manage the crisis. In this case study, we analyse the factors that led the government to switch policy and discuss why elimination became unsustainable to manage the COVID-19 epidemic in New Zealand. Results showed that the socioeconomic and political factors, along with the appearance of new variants and a delayed vaccination program, were accountable for the switch in strategy. This switch allows the country to better adapt to the evolving nature of the disease and to address the social and economic repercussions of the first year of measures. Our conclusion does not disregard elimination as an appropriate initial strategy to contain this pandemic in the absence of a vaccine or treatment, but rather suggests that borders cannot remain closed for long periods of time without creating social, economical, and political issues.
ABSTRACT
While Deep Learning (DL) is often considered the state-of-the art for Artificial Intel-ligence-based medical decision support, it remains sparsely implemented in clinical practice and poorly trusted by clinicians due to insufficient interpretability of neural network models. We have approached this issue in the context of online detection of epileptic seizures by developing a DL model from EEG signals, and associating certain properties of the model behavior with the expert medical knowledge. This has conditioned the preparation of the input signals, the network architecture, and the post-processing of the output in line with the domain knowledge. Specifically, we focused the discussion on three main aspects: (1) how to aggregate the classification results on signal segments provided by the DL model into a larger time scale, at the seizure-level; (2) what are the relevant frequency patterns learned in the first convolutional layer of different models, and their relation with the delta, theta, alpha, beta and gamma frequency bands on which the visual interpretation of EEG is based; and (3) the identification of the signal waveforms with larger contribution towards the ictal class, according to the activation differences highlighted using the DeepLIFT method. Results show that the kernel size in the first layer determines the interpretability of the extracted features and the sensitivity of the trained models, even though the final performance is very similar after post-processing. Also, we found that amplitude is the main feature leading to an ictal prediction, suggesting that a larger patient population would be required to learn more complex frequency patterns. Still, our methodology was successfully able to generalize patient inter-variability for the majority of the studied population with a classification F1-score of 0.873 and detecting 90% of the seizures.
