Research
Print page Print page
Switch language
The Capital Region of Denmark - a part of Copenhagen University Hospital
Published

Prediction of Patient Demographics using 3D Craniofacial Scans and Multi-view CNNs

Research output: Contribution to journalJournal articleResearchpeer-review

  1. Automatic Segmentation to Cluster Patterns of Breathing in Sleep Apnea

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Polysomnographic Plethysmography Excursions are Reduced in Obese Elderly Men

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Prediction of severe adverse event from vital signs for post-operative patients

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. Semi-Supervised Analysis of the Electrocardiogram Using Deep Generative Models

    Research output: Contribution to journalJournal articleResearchpeer-review

  5. Upper Airway Classification in Sleep Endoscopy Examinations using Convolutional Recurrent Neural Networks

    Research output: Contribution to journalJournal articleResearchpeer-review

  1. Arousal characteristics in patients with Parkinson's disease and isolated rapid eye movement sleep behavior disorder

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Brain tumours in children and adolescents may affect the circadian rhythm and quality of life

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Brain tumours result in sleep disorders in children and adolescents

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. European guideline and expert statements on the management of narcolepsy in adults and children

    Research output: Contribution to journalReviewResearchpeer-review

  5. Automatic Segmentation to Cluster Patterns of Breathing in Sleep Apnea

    Research output: Contribution to journalJournal articleResearchpeer-review

  • Umaer Hanif
  • Rasmus R Paulsen
  • Eileen B Leary
  • Emmanuel Mignot
  • Poul Jennum
  • Helge B D Sorensen
View graph of relations

3D data is becoming increasingly popular and accessible for computer vision tasks. A popular format for 3D data is the mesh format, which can depict a 3D surface accurately and cost-effectively by connecting points in the (x, y, z) plane, known as vertices, into triangles that can be combined to approximate geometrical surfaces. However, mesh objects are not suitable for standard deep learning techniques due to their non-euclidean structure. We present an algorithm which predicts the sex, age, and body mass index of a subject based on a 3D scan of their face and neck. This algorithm relies on an automatic pre-processing technique, which renders and captures the 3D scan from eight different angles around the x-axis in the form of 2D images and depth maps. Subsequently, the generated data is used to train three convolutional neural networks, each with a ResNet18 architecture, to learn a mapping between the set of 16 images per subject (eight 2D images and eight depth maps from different angles) and their demographics. For age and body mass index, we achieved a mean absolute error of 7.77 years and 4.04 kg/m2 on the respective test sets, while Pearson correlation coefficients of 0.76 and 0.80 were obtained, respectively. The prediction of sex yielded an accuracy of 93%. The developed framework serves as a proof of concept for prediction of more clinically relevant variables based on 3D craniofacial scans stored in mesh objects.

Original languageEnglish
JournalProceedings of the International Conference of the IEEE Engineering in Medicine and Biology Society
Volume2020
Pages (from-to)1950-1953
Number of pages4
ISSN2375-7477
DOIs
Publication statusPublished - Jul 2020

    Research areas

  • Algorithms, Demography, Head, Humans, Neural Networks, Computer

ID: 61826508