Globally, almost every third person suffers from malocclusion, yet only a tiny fraction of those in need ever receive orthodontic treatment. The time-intensive nature of orthodontic planning results in prohibitively expensive therapy while limiting the number of patients a single practitioner can treat. Fortunately, the last two decades have seen multiple promising technologies reshaping modern orthodontics, making care more accessible, efficient, and safer.
One up-and-coming technology in orthodontics is Cone Beam Computed Tomography (CBCT). CBCT helps assess root fractures and resorption, condylar shapes, alveolar boundary conditions, tooth angulation, and potential root collisions. Despite higher radiation doses, its use can be justified in cases where it significantly impacts treatment options. According to a study from Haney et al. [1], up to 25% of treatment plans for impacted canines were readjusted once CBCT data was included in planning. To be able to add root and bones to treatment plans, companies like Align, Spark, LightForce, Angel Align, SureSmile, SoftSmile, ArchForm, Arcad, and Onyxceph have started integrating options to include CBCT data into their software. Today, around 15% of digital treatment plans leverage CBCT data.
Despite CBCT’s advantages compared to two-dimensional radiographs, the third dimension also introduces complexity. With practitioners already spending significant time on treatment planning, any additional time spent assessing and segmenting structures in imaging can prove critical.
Segmenting CBCT images poses several challenges. The lower contrast and higher noise levels compared to conventional CT scans can obscure details, making it difficult to differentiate between tissues. Additionally, artifacts from metal structures, the partial volume effect, and the complexity of small anatomical structures further complicate the process. These difficulties necessitate advanced image processing techniques and often manual intervention, which can be time-consuming. Overcoming these challenges is essential for fully leveraging the detailed data CBCT provides and enhancing treatment planning in orthodontics.
“Up to 25% of treatment plans for impacted canines were readjusted once CBCT data was included in planning.”
AI for CBCTs and Beyond
Given the higher radiation dosage compared to 2D radiographs, CBCT data should be leveraged to its full potential, including segmenting the entire scan. Relu, a Belgian software enterprise proficient in dental AI, has pioneered algorithms capable of segmenting CBCT images in under 10 minutes. This innovation enables significant time savings for orthodontic professionals globally. Fast and precise segmentation ensures that the data obtained from CBCTs has the maximum possible impact on the quality of treatment.
Traditionally, dentists are trained and most proficient in analyzing 2D imaging. Relu recognizes this, automating the generation of cephalometric views and panoramic radiographs from CBCT, enabling swift 2D evaluations and a smooth transition to 3D imaging. This innovation provides the best of both worlds: a quick overview in 2D while maintaining the expansive data of the CBCT.
However, the innovations go beyond CBCT; the automation ranges from registering intra-oral and facial scans onto the CBCT model to closing the gingiva in virtual extractions and automatically highlighting the tooth axis and mesial and distal points. This is only the beginning, as more innovations in orthodontics are on the horizon, with further automation already in development.
Relu’s mission is to ensure that the declining number of dental technicians can keep up with the demand for digital dental treatments like clear aligners, thereby providing the best possible care to each patient globally. By leveraging automation and optimizing CBCT technology, the cost of dental treatment can be reduced while maintaining a high quality of care.
SOURCES
(1) Haney E, Gansky SA, Lee JS, Johnson E, Maki K, Miller AJ, Huang JC. Comparative analysis of traditional radiographs and cone-beam computed tomography
volumetric images in the diagnosis and treatment planning of maxillary impacted canines. Am J Orthod Dentofacial Orthop. 2010 May;137(5):590-7. doi: 10.1016/j.
ajodo.2008.06.035. PMID: 20451777.
2) Conrad studied dentistry for four yearsbefore pivoting to software development. He completed his degree in Business InformationSystems in Goettingen, Germany. Last year, he joined Relu to pursue his passion for ArtificialIntelligence in health care.