Affiliations 

  • 1 School of Dental Sciences, Health Campus, Universiti Sains Malaysia, Kelantan
  • 2 Faculty of Ocean Engineering Technology and Informatics, Universiti Malaysia Terengganu, Terengganu, Malaysia
  • 3 College of Dentistry, Jouf University, Sakaka, Saudi Arabia
J Craniofac Surg, 2021 Jun 01;32(4):1500-1503.
PMID: 33852515 DOI: 10.1097/SCS.0000000000007435

Abstract

ABSTRACT: Oral and maxillofacial fractures are the most common injuries among multiple trauma. About 5% to 10% of trauma patients having facial fractures. The objectives of this case study are to focus the most common mid-face fractures types' and to determine the relationship of the midface fracture in maxillofacial trauma among the patient who attended the outpatient clinic in a Hospital Universiti Sains Malaysia. In this research paper, an advanced statistical tool was chosen through the multilayer perceptron neural network methodology (MLPNN). Multilayer perceptron neural network methodology was applied to determine the most associated predictor important toward maxillary bone injury. Through the predictor important classification analysis, the relationship of each bone will be determined, and sorting according to their contribution. After sorting the most associated predictor important toward maxillary bone injury, the validation process will be applied through the value of training, testing, and validation. The input variables of MLPNN were zygomatic complex fracture, orbital wall fracture, nasal bone fracture, frontal bone fracture, and zygomatic arch fracture. The performance of MLPNN having high accuracy with 82.2%. As a conclusion, the zygomatic complex fracture is the most common fracture trauma among the patient, having the most important association toward maxillary bone fracture. This finding has the highest potential for further statistical modeling for education purposes and the decision-maker among the surgeon.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.