A dental implant is a screw-like device that is surgically placed in the jawbone to provide a foundation for artificial teeth. This involves precise removal of bone using drills, which is often risky because of proximity to delicate structures such as the maxillary sinus, orofacial nerves, and blood vessels. Mistakes in the drilling path may result in permanent nerve damage, life threatening hemorrhage, or injuries to adjacent teeth. This research project aims to reduce errors in the process by developing an objective and sensor-based method to assist practitioners in conducting the drilling process.
Our method will analyze the sounds generated during implant drillings to monitor the process and recognize different bone tissues, providing real-time feedback on whether the practitioner is taking the correct line. Proof of concept exists in that drilling sounds have already been used in similar applications to discriminate between tooth materials.
To collect the data, we will drill sample jawbones (pig or cow) as we would in typical implant surgeries. We will record the sounds produced by drilling bone tissues under different conditions such as direction, feed rate, speed, and applied forces. Advanced signal processing methods such as machine learning will analyze the data to allow us to discriminate between different bone tissues.
We will optimize the resulting algorithm to produce an aid for practitioners that will improve the safety and precision of their dental implant surgeries. Future work could include further customizing the algorithm to extend its use to other medical procedures that involve bone drilling such as orthopedics, spine, and ear surgeries.