Spatio-Temporal Clustering of Multi-Location Time Series to Model Seasonal Influenza Spread.

Although seasonal influenza disease spread is a spatio-temporal phenomenon, public surveillance systems aggregate data only spatially, and are rarely predictive. We develop a hierarchical clustering-based machine learning tool to anticipate flu spread patterns based on historical spatio-temporal flu activity, where we use historical influenza-related emergency department records as a proxy for flu prevalence. This analysis replaces conventional geographical hospital clustering with clusters based on both spatial and temporal distance between hospital flu peaks to generate a network illustrating whether flu spreads between pairs of clusters (direction) and how long that spread takes (magnitude). To overcome data sparsity, we take a model-free approach, treating hospital clusters as a fully-connected network, where arcs indicate flu transmission. We perform predictive analysis on the clusters' time series of flu ED visits to determine direction and magnitude of flu travel. Detection of recurrent spatio-temporal patterns may help policymakers and hospitals better prepare for outbreaks. We apply this tool to Ontario, Canada using a five-year historical dataset of daily flu-related ED visits, and find that in addition to expected flu spread between major cities/airport regions, we were able to illuminate previously unsuspected patterns of flu spread between non-major cities, providing new insights for public health officials. We showed that while a spatial clustering outperforms a temporal clustering in terms of the direction of the spread (81% spatial v. 71% temporal), the opposite is true in terms of the magnitude of the time lag (20% spatial v. 70% temporal).

SuPART: supervised projective adapted resonance theory for automatic quality assurance approval of radiotherapy treatment plans.

Radiotherapy is a common treatment modality for the treatment of cancer, where treatments must be carefully designed to deliver appropriate dose to targets while avoiding healthy organs. The comprehensive multi-disciplinary quality assurance (QA) process in radiotherapy is designed to ensure safe and effective treatment plans are delivered to patients. However, the plan QA process is expensive, often time-intensive, and requires review of large quantities of complex data, potentially leading to human error in QA assessment. We therefore develop an automated machine learning algorithm to identify 'acceptable' plans (plans that are similar to historically approved plans) and 'unacceptable' plans (plans that are dissimilar to historically approved plans). This algorithm is a supervised extension of projective adaptive resonance theory, called SuPART, that learns a set of distinctive features, and considers deviations from them indications of unacceptable plans. We test SuPART on breast and prostate radiotherapy datasets from our institution, and find that SuPART outperforms common classification algorithms in several measures of accuracy. When no falsely approved plans are allowed, SuPART can correctly auto-approve 34% of the acceptable breast and 32% of the acceptable prostate plans, and can also correctly reject 53% of the unacceptable breast and 56% of the unacceptable prostate plans. Thus, usage of SuPART to aid in QA could potentially yield significant time savings.