Morphological Information Extraction in medical imaging using deep learning interpretability: application case on craniofacial dysmorphism

Research Square

Abstract

Abstract Recent advances in the interpretability of convolutional neural networks (CNNs) have allowed applications in imaging as a novel method for visual feature extraction. We used this approach to investigate the impact of changing occlusal forces on craniofacial architecture in class II retrognathism (C2Rm) pathology. Better understanding the points of impact of C2Rm on the entire skull is a major challenge in the diagnosis, treatment, and management of this dysmorphism, but also allows to be part of the debate on changes in the shape of the skull during human evolution. To address these challenges, we introduced a methodology for extracting morphological information with an interpretable CNN (MIE-ICNN) that combines: 1) a pre-processing step to train a deep CNN, 2) an interpretability step using the Score-CAM method to explain the areas of the radiographs used by the model to identify the pathological class, and 3) a calculation step to generate a global activation map, which represents an average activation map of all classifications. The main conclusions of this study are as follows: I) the proposed technique made it possible to find the anatomical areas affected by C2Rm and already identified in the literature (i.e. the cranial base and the vertebrae), to confirm the involvement of a controversial area (the frontal sinus), and to identify a new structure (the parietal bone) as a biomarker. In addition, the study of the involvement of anatomical zones according to the severity of the dysmorphism (by the same methodology), has made it possible to propose hypotheses on the evolution of the forms of the skulls during human evolution (highlighting, in particular, the role of mastication). This new information enriches the anthropological debate around C2Rm and human evolution, and medical knowledge on the etiology of C2Rm. Finally, through this application on craniospinal architecture, the proposed methodological pipeline appears as a new toolbox to extract and visualize complex morphological systems and will certainly allow new advances in the fields of health and medical imaging.

Authors 8

1. Masrour Makaremi

   Université de Bordeaux · Inserm · Bordeaux Population Health

   Affiliation as printed

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, Bordeaux University

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, University of Bordeaux, Bordeaux, France

   Handicap Activity Cognition Health, INSERM U1219, Bordeaux university, Talence, France

2. Benoit Marcy

   Université de Toulouse

   Affiliation as printed

   Toulouse University

   Toulouse university, Toulouse, France

3. Ikram Chraibi Kaadoud

   IMT Atlantique

   Affiliation as printed

   LUSSI Department, IMT Atlantique, F-29238 Brest

   LUSSI Department, IMT Atlantique, F-29238 Brest, France

4. Alireza Vafaei Sadr Aachen Pathology Institute

   University of Geneva · Universitätsklinikum Aachen

   Affiliation as printed

   Pathology Institute, Uniklinik RWTH Aachen

   Pathology Institute, Uniklinik RWTH Aachen, Germany

   Theoretical physics department, University of Geneva, Geneva, Switzerland

5. Ali Mohammad‐Djafari

   Supélec · CentraleSupélec

   Affiliation as printed

   LSS, Central Supelec, Paris

   LSS, Centrale Supelec, Paris, France

6. Salomé Sadoun

   Université de Bordeaux

   Affiliation as printed

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, Bordeaux University

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, University of Bordeaux, Bordeaux, France

7. François DE BRONDEAU

   Université de Bordeaux

   Affiliation as printed

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, Bordeaux University

   Dentofacial Orthopedics Department, UFR des Sciences Odontologiques, University of Bordeaux, Bordeaux, France

8. Bernard N'KAOUA

   Inserm · Bordeaux Population Health

   Affiliation as printed

   Handicap Activity Cognition Health, INSERM U1219, Bordeaux university, Talence

   Handicap Activity Cognition Health, INSERM U1219, Bordeaux university, Talence, France