Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

eHealth application and AI-solutions for early detection of cerebral palsy in newborns - based on home video recordings for general movement assessment

Kircheiner Brown, Annemette LU (2025) In Lund University, Faculty of Medicine Doctoral Dissertation Series 65(2025:65).
Abstract
Background
Early detection of cerebral palsy (CP) is essential to initiate timely intervention and
improve long-term outcomes for infants at risk. The General Movements
Assessment (GMA) is a well-established tool for early identification of CP, yet its
implementation in routine clinical practice faces challenges related to access and
scalability.
Aim
This thesis explores the use of GMA as a screening tool for newborns with
detectable risk for CP, focusing on digital innovations, parental experiences,
screening strategies, and future perspectives involving artificial intelligence (AI).
Methods and Results
Through four interconnected studies, the thesis examines: (1) the feasibility... (More)
Background
Early detection of cerebral palsy (CP) is essential to initiate timely intervention and
improve long-term outcomes for infants at risk. The General Movements
Assessment (GMA) is a well-established tool for early identification of CP, yet its
implementation in routine clinical practice faces challenges related to access and
scalability.
Aim
This thesis explores the use of GMA as a screening tool for newborns with
detectable risk for CP, focusing on digital innovations, parental experiences,
screening strategies, and future perspectives involving artificial intelligence (AI).
Methods and Results
Through four interconnected studies, the thesis examines: (1) the feasibility of
performing GMA remotely using smartphone-recorded home videos, (2) parents’
experiences of participating in home-based GMA screening, (3) the potential use of
the Ages and Stages Questionnaire (ASQ) as a pre-screening tool to select infants
born before 32 weeks’ gestation for GMA, and (4) interpretability of the AI-based
In-Motion model.
The findings support the feasibility and clinical utility of remote GMA, with parents
reporting empowerment and increased accessibility. While ASQ is widely used in
developmental monitoring, it was not effective in identifying all infants who should
undergo GMA, limiting its value as a pre-screening tool. The In-Motion model has
earlier demonstrated strong predictive accuracy but relies on movement features that
differ from traditional fidgety movements, raising important questions about
transparency and how to interpret it in a clinical practice.
Conclusion
This thesis supports the use of GMA, particularly via smartphone technology, as a
viable and equitable screening method for CP in at-risk infants. It highlights the
importance of involving parents, optimising resource allocation, and carefully
considering the future role of AI in early detection. Together, these findings
contribute to the development of more accessible, effective, and family-centred
models for early CP screening. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Professor Lorenzen, Jakob, University of Copenhagen
organization
publishing date
type
Thesis
publication status
published
subject
keywords
cerebral palsy, general movements assessment, home video recording, ehealth, Artifical Intelligence, family-centered care, early detection
in
Lund University, Faculty of Medicine Doctoral Dissertation Series
volume
65
issue
2025:65
pages
54 pages
publisher
Lund University, Faculty of Medicine
defense location
GK-salen, Forum Medicum, Sölvegatan 19, Lund. Join by Zoom: https://lu-se.zoom.us/j/68149710295?pwd=AssnMw31gsxtJWyC1ykp3wa4UblAVm.1
defense date
2025-11-13 13:00:00
ISSN
1652-8220
ISBN
978-91-8021-718-7
language
English
LU publication?
yes
id
02848ea3-34b7-4a7d-ba21-c2aa4504845c
date added to LUP
2025-10-20 16:49:35
date last changed
2025-10-27 08:59:03
@phdthesis{02848ea3-34b7-4a7d-ba21-c2aa4504845c,
  abstract     = {{Background<br/>Early detection of cerebral palsy (CP) is essential to initiate timely intervention and<br/>improve long-term outcomes for infants at risk. The General Movements<br/>Assessment (GMA) is a well-established tool for early identification of CP, yet its<br/>implementation in routine clinical practice faces challenges related to access and<br/>scalability.<br/>Aim<br/>This thesis explores the use of GMA as a screening tool for newborns with<br/>detectable risk for CP, focusing on digital innovations, parental experiences,<br/>screening strategies, and future perspectives involving artificial intelligence (AI).<br/>Methods and Results<br/>Through four interconnected studies, the thesis examines: (1) the feasibility of<br/>performing GMA remotely using smartphone-recorded home videos, (2) parents’<br/>experiences of participating in home-based GMA screening, (3) the potential use of<br/>the Ages and Stages Questionnaire (ASQ) as a pre-screening tool to select infants<br/>born before 32 weeks’ gestation for GMA, and (4) interpretability of the AI-based<br/>In-Motion model.<br/>The findings support the feasibility and clinical utility of remote GMA, with parents<br/>reporting empowerment and increased accessibility. While ASQ is widely used in<br/>developmental monitoring, it was not effective in identifying all infants who should<br/>undergo GMA, limiting its value as a pre-screening tool. The In-Motion model has<br/>earlier demonstrated strong predictive accuracy but relies on movement features that<br/>differ from traditional fidgety movements, raising important questions about<br/>transparency and how to interpret it in a clinical practice.<br/>Conclusion<br/>This thesis supports the use of GMA, particularly via smartphone technology, as a<br/>viable and equitable screening method for CP in at-risk infants. It highlights the<br/>importance of involving parents, optimising resource allocation, and carefully<br/>considering the future role of AI in early detection. Together, these findings<br/>contribute to the development of more accessible, effective, and family-centred<br/>models for early CP screening.}},
  author       = {{Kircheiner Brown, Annemette}},
  isbn         = {{978-91-8021-718-7}},
  issn         = {{1652-8220}},
  keywords     = {{cerebral palsy; general movements assessment; home video recording; ehealth; Artifical Intelligence; family-centered care; early detection}},
  language     = {{eng}},
  number       = {{2025:65}},
  publisher    = {{Lund University, Faculty of Medicine}},
  school       = {{Lund University}},
  series       = {{Lund University, Faculty of Medicine Doctoral Dissertation Series}},
  title        = {{eHealth application and AI-solutions for early detection of cerebral palsy in newborns - based on home video recordings for general movement assessment}},
  url          = {{https://lup.lub.lu.se/search/files/230857799/Thesis_final_version.pdf}},
  volume       = {{65}},
  year         = {{2025}},
}