Author: Amal El Batouti

Status: Project Concept


Maxillary growth and dentoalveolar development are impaired in cleft lip and palate patients after surgical closure of the palate. Growth impairment is mainly due to manipulation of mucoperiosteum where wound contraction and scar fibers account for traction of permanent teeth toward it, resulting in hypoplastic tendency toward posterior and anterior crossbite. Maxillary expanders have been routinely used for transverse expansion. Relapse and arch form instability affect the treatment outcome.
To develop a machine learning algorithm that would promote orthodontic treatment and achieve a higher level of efficiency in: 1. maxillary expansion for the well defined and exact width required of maxilla.2. symmetric maxillary arch form. 3. repositioning of retroclined teeth and accompanied collapsed alveolar ridge. 4. maintaining a stable occlusion.
Expansion through a midpalatal automated screw incorporated in Hyrax appliance to exert force that stimulates bone formation in the cleft via distraction osteogenesis.The scarred soft tissue would respond to the application of forces allowing the maxilla to be widened. The machine learning algorithm would control the width required of the maxilla by utilizing the following anticipated feature sets: 1. the progress of the activated screw in interval visits, 2. the amount of bone formation, 3. the expansion needed according to the case (unilateral or bilateral cleft), 4. severity of the case (mild,moderate or severe) and 5. the age sequence at which the treatment had been performed starting from. Additional features identified through the training set will additionally be evaluated. Based on computerized cast analysis in addition to analysis of anteroposterior cephalogram, orthopedic protraction with maxillary device (Delaire mask) would start immediately after maxillary expansion to apply a tensile force on the circumaxillary sutures, and thereby, stimulate bone apposition in the suture areas. Additional inputs can be applied to the algorithm based on analysis of CT taken before and after protraction. These scans also have the benefit of providing 3D models superimposed on the anterior cranial base will analyze the amount of displacement in the frontal, vertical and lateral directions of the nasomaxillary complex and approximate it to the natural growth direction of the maxilla with the opportunity to leverage augmented and/or virtual reality to allow the orthodontist to visualize the abnormality and treatment progress.
– Machine learning to define arch form pattern and describe the specific arch shape to each patient depending on scanned images of maxillary and mandibular casts superimposed with the Pentamorphic arch system. Preservation of dental arch shape during growth is an indicator of teeth equilibrium between tongue and circumoral muscle forces. Alveolar bone grafting at (9-11 years) in cleft side promotes influence of canine eruption that contribute to long term stability of maxillary arch form.
It is reasonable to assume that applying learning machine to guide the treatment plan of patients with cleft lip and palate would improve the accuracy and inform the best course of treatment, positively guiding the maxillary and dentoalveolar arch toward normal growth and more stability would enhance the treatment to be completed at age of maturation, preventing the need of any further surgical interference, then reducing possibilities of prolonged treatment.

Dr.Amal El Batouti , MS.Orthodontics Alexandria university.Egypt