The accuracy of this artificial intelligence system in implant identification depends on the quality of the images provided. This system is optimized to work with periapical radiographs, so it is recommended to use this type of image in png, jpg or jpeg format. Whenever possible, images should be obtained using paralleling devices or similar tools.
Radiographs should be of good quality, with correct exposure, sharp focus, and sufficient contrast so that implants and surrounding structures, such as the platform and apex, are clearly visible. Images that are too dark, overexposed, or out of focus may hinder the analysis and affect the accuracy of the results.
It is recommended to upload radiographs directly from the computer, either by exporting them from your management software or through a screenshot. Photographs taken with a mobile phone of the computer screen introduce noise and may present distortions, reflections, or excessive compression that can alter detection.
In addition, it is advisable that each radiograph contains a maximum of five implants. An excessive number may reduce the system’s effectiveness in identifying each one accurately.
By following these recommendations, the system will be able to analyze radiographs more reliably and provide more precise results.
In both the AI Identification and Collaborate sections, an Acquisition Guide is available at the bottom, summarizing all the recommendations to achieve an accurate analysis.
