AI’s Potential in Helping Clinicians Identify Dental Implants

Artificial Intelligence has enabled healthcare specialists to achieve goals that were nigh impossible only a few years ago. A recent study by Stony Brook University’s SUNY Empire Innovation Professor Haibin Ling and the Director of the Renaissance School of Medicine at SBU Jie Yang shows that AI can now identify the attributes of various dental implant systems with an accuracy of over 95%—an application with great implications for the orthopedic industry.

The demand for dental implants has been steadily increasing ever since they were introduced in the 1970s. However, patient mobility from different dental offices and the lack of past available records have made it difficult for clinicians to identify existing implants before operating on their new patients, resulting in unnecessary expenses and complications.

The team at SBU attempted to solve this problem by having AI study radiographs of close to 800 dental implants manufactured by three different companies (BioHorizons, Straumann, and Nobel Biocare). They trained the model to look at various attributes of 75% of these implant systems, including their interface, their taper, as well as the type of their bone tissue, among other things. Once the model was trained, its accuracy was tested on the remaining 25% of the dental implants.

The results of their efforts were promising, showing an average accuracy greater than 95%. Professor Haibin Ling, who has been in the AI industry for over two decades, commented, “The idea is to take this a step further—we want to expand our dataset to include more radiographs, more companies, so that periodontists can identify their patients’ dental histories more easily. If we continue to achieve as high an accuracy, this will be a big step for the dental implant industry.”

Professor Jie Yang adds, “There is potential for identifying different kinds of non-biological implants using AI with accuracy rather than visual guesswork. The variety, speed, and correctness of this method of analysis are time-saving for many clinicians who specialize in placing and restoring implants.”

 

Ankita Nagpal
Communications Assistant