Artificial Intelligence in Identifying Dental Implant Systems on Radiographs.

Health care is entering a new era where data mining is applied to artificial intelligence. The number of dental implant systems has been increasing worldwide. Patient mobility from different dental offices can make identification of implants for clinicians extremely challenging if there are no past available records, and it would be advantageous to use a reliable tool to identify the various implant system designs in the same practice, as there is a great need for identifying the systems in the field of periodontology and restorative dentistry. However, there have not been any studies devoted to using artificial intelligence/convolutional neural networks to classify implant attributes. Thus, the present study used artificial intelligence to identify the attributes of radiographic images of implants. An average accuracy rate of over 95% was achieved with various machine learning networks to identify three implant manufacturers and their subtypes placed during the past 9 years.

Phenotypic identification of periodontal Prevotella intermedia/nigrescens group isolates validated by MALDI-TOF mass spectrometry.

The accuracy of a phenotypic scheme to recognize periodontal Prevotella intermedia/nigrescens group clinical isolates on primary isolation culture plates was assessed with matrix-assisted laser desorption-ionization time-of-flight mass spectrometry (MALDI-TOF MS). A total of 84 fresh subgingival isolates from 23 chronic periodontitis patients were presumptively recognized on anaerobically-incubated enriched Brucella blood agar primary isolation plates as P. intermedia/nigrescens based on their dark-pigmented colony morphology, brick-red autofluorescence under long-wave ultraviolet light, and a negative fluorescence test for lactose production. The presumptive P. intermedia/nigrescens clinical isolates were subjected to MALDI-TOF MS analysis using Bruker MALDI Biotyper analytic software containing mass spectra for P. intermedia and Prevotella nigrescens in its reference library of bacterial protein profiles. Using a ≥1.7 log score agreement threshold, 60 (71.4%) of the presumptive P. intermedia/nigrescens clinical isolates were confirmed as either P. intermedia (25 isolates) or P. nigrescens (35 isolates). All isolates with a <1.7 log score were also identified as P. intermedia or P. nigrescens from the top choice designated on the MALDI Biotyper most likely species identification list. These MALDI-TOF MS findings document the ability of the phenotypic scheme to correctly recognize most periodontal P. intermedia/nigrescens group clinical isolates on primary isolation culture plates.