Computer-assisted periodontal classification vs manual classification among dental students: a comparative pilot study.

OBJECTIVE: The objective of the study was to assess whether computer-assisted periodontal diagnosis can improve the accuracy and homogeneity of classification results obtained by dental students using the 2017 classification of periodontal diseases. METHOD AND MATERIALS: All final year predoctoral dental students from two dental schools were invited to participate in the study. Participants who volunteered for the study were randomly divided into two groups for digital or manual diagnosis, and each participant classified 48 cases. A group of three experienced periodontists provided the reference or gold standard diagnosis. RESULTS: Overall, 27 students completed the evaluation of all cases; 14 students comprised the digital application group and 13 the manual group. The accuracy of the classification results compared with the gold standard committee was 82% for the digital group compared to 50% of the manual group in terms of the extent of gingivitis; 71% vs 56% for the stage of periodontitis; 67% vs 62% for grade of periodontitis; 76% vs 63% for extent of periodontitis; and 43% vs 30% for overall diagnosis accuracy of periodontitis cases respectively. CONCLUSIONS: Computer-assisted classification using newly developed software, within the boundaries of this study, was shown to be a sensible support tool for dental practitioners to use when diagnosing periodontal disease. This digital tool can the clinicians' accuracy of diagnosis primarily in the extent and staging of periodontitis.

Size agnostic change point detection framework for evolving networks.

Changes in the structure of observed social and complex networks can indicate a significant underlying change in an organization, or reflect the response of the network to an external event. Automatic detection of change points in evolving networks is rudimentary to the research and the understanding of the effect of such events on networks. Here we present an easy-to-implement and fast framework for change point detection in evolving temporal networks. Our method is size agnostic, and does not require either prior knowledge about the network's size and structure, nor does it require obtaining historical information or nodal identities over time. We tested it over both synthetic data derived from dynamic models and two real datasets: Enron email exchange and AskUbuntu forum. Our framework succeeds with both precision and recall and outperforms previous solutions.