The study of preterm infants neuro-motor status can be performed by analyzing infants spontaneous movements. Nowadays, available automatic methods for assessing infants motion patterns are still limited. We present a novel pipeline for the characterization of infants spontaneous movements, which given RGB videos leverages on network analysis and NLP. First, we describe a body configuration for each frame considering landmark points on infants bodies as nodes of a network and connecting them depending on their proximity. Each configuration can be described by means of attributed graphettes. We identify each attributed graphette by a string, thus allowing to study videos as texts, i.e. sequences of strings. This allows us exploiting NLP methods as topic modelling to obtain interpretable representations. We analyze topics to describe both global and local differences in infants with normal and abnormal motion patters. We find encouraging correspondences between our results and evaluations performed by expert physicians.

Attributed Graphettes-Based Preterm Infants Motion Analysis

Garbarino, Davide;Moro, Matteo;Tacchino, Chiara;Casadio, Maura;Odone, Francesca;Barla, Annalisa
2022

Abstract

The study of preterm infants neuro-motor status can be performed by analyzing infants spontaneous movements. Nowadays, available automatic methods for assessing infants motion patterns are still limited. We present a novel pipeline for the characterization of infants spontaneous movements, which given RGB videos leverages on network analysis and NLP. First, we describe a body configuration for each frame considering landmark points on infants bodies as nodes of a network and connecting them depending on their proximity. Each configuration can be described by means of attributed graphettes. We identify each attributed graphette by a string, thus allowing to study videos as texts, i.e. sequences of strings. This allows us exploiting NLP methods as topic modelling to obtain interpretable representations. We analyze topics to describe both global and local differences in infants with normal and abnormal motion patters. We find encouraging correspondences between our results and evaluations performed by expert physicians.
978-3-030-93408-8
978-3-030-93409-5
File in questo prodotto:
File Dimensione Formato  
garbarino2022-postprint.pdf

accesso chiuso

Tipologia: Documento in versione editoriale
Dimensione 1.7 MB
Formato Adobe PDF
1.7 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11567/1065560
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact