1
|
Letzkus L, Pulido JV, Adeyemo A, Baek S, Zanelli S. Machine learning approaches to evaluate infants' general movements in the writhing stage-a pilot study. Sci Rep 2024; 14:4522. [PMID: 38402234 PMCID: PMC10894291 DOI: 10.1038/s41598-024-54297-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/26/2023] [Accepted: 02/11/2024] [Indexed: 02/26/2024] Open
Abstract
The goals of this study are to describe machine learning techniques employing computer-vision movement algorithms to automatically evaluate infants' general movements (GMs) in the writhing stage. This is a retrospective study of infants admitted 07/2019 to 11/2021 to a level IV neonatal intensive care unit (NICU). Infant GMs, classified by certified expert, were analyzed in two-steps (1) determination of anatomic key point location using a NICU-trained pose estimation model [accuracy determined using object key point similarity (OKS)]; (2) development of a preliminary movement model to distinguish normal versus cramped-synchronized (CS) GMs using cosine similarity and autocorrelation of major joints. GMs were analyzed using 85 videos from 74 infants; gestational age at birth 28.9 ± 4.1 weeks and postmenstrual age (PMA) at time of video 35.9 ± 4.6 weeks The NICU-trained pose estimation model was more accurate (0.91 ± 0.008 OKS) than a generic model (0.83 ± 0.032 OKS, p < 0.001). Autocorrelation values in the lower limbs were significantly different between normal (5 videos) and CS GMs (5 videos, p < 0.05). These data indicate that automated pose estimation of anatomical key points is feasible in NICU patients and that a NICU-trained model can distinguish between normal and CS GMs. These preliminary data indicate that machine learning techniques may represent a promising tool for earlier CP risk assessment in the writhing stage and prior to hospital discharge.
Collapse
Affiliation(s)
- Lisa Letzkus
- Department of Pediatrics, University of Virginia Children's Hospital, PO Box 800828, Charlottesville, VA, 22908, USA.
| | - J Vince Pulido
- Applied Physics Laboratory, Johns Hopkins University, Laurel, MD, USA
| | - Abiodun Adeyemo
- Department of Pediatrics, University of Virginia Children's Hospital, PO Box 800828, Charlottesville, VA, 22908, USA
| | - Stephen Baek
- School of Data Science, University of Virginia, Charlottesville, VA, USA
| | - Santina Zanelli
- Department of Pediatrics, University of Virginia Children's Hospital, PO Box 800828, Charlottesville, VA, 22908, USA
| |
Collapse
|
2
|
Fortune A, Perkins E, Paize F, Palanisami B, Gladstone M. Managing mothers' and fathers' uncertainty during their journey through early neurodevelopmental follow-up for their high-risk infants-A qualitative account. Child Care Health Dev 2024; 50:e13168. [PMID: 37737651 DOI: 10.1111/cch.13168] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 06/27/2022] [Revised: 06/30/2023] [Accepted: 08/01/2023] [Indexed: 09/23/2023]
Abstract
BACKGROUND Early diagnosis of cerebral palsy is possible by 5 months corrected age for 'at-risk' infants, using diagnostic tools such as the Hammersmith Infant Neurological Examination (HINE), Prechtl's General Movements Assessment (GMA) and magnetic resonance imaging (MRI). This is an uncertain and stressful time for parents where provision of appropriate information and support is essential. AIM To explore parents' views and experiences in relation to the new early neurodevelopmental follow-up of 'at-risk' infants. METHODS Thirteen in-depth one-to-one qualitative interviews were conducted by the primary researcher, with eight parents (six mothers and two fathers) of 'at-risk' infants eligible for a follow-up clinic where the GMA and HINE were performed at 12-week corrected age. Interviews used a pre-piloted topic guide and took place before and after the clinic. Interviews were audio-recorded, transcribed verbatim and analysed using inductive coding and thematic analysis using the framework approach. FINDINGS Seven themes were identified: (1) attempting to manage uncertainty, (2) taking priority, (3) trusting professionals, (4) independence in the parent role, (5) feeling understood, (6) patterns of care and (7) individuality. Parents reported experiencing uncertainty about their current situation and future. Adequate preparation for and timing of information are vital. When uncertainty is poorly managed, parents' wellbeing suffers. Individual parents' perspectives and infants' developmental trajectories differ, and information should be tailored specifically for this. CONCLUSION A parent's understanding of the journey through neurodevelopmental care for their high risk infants is initially very limited. Implementing a counselling service for parents to access psychological support and digital reminder system for clinic appointments, as well as providing more tailored information through trusted professionals, could all improve future parents' experiences.
Collapse
Affiliation(s)
- Alice Fortune
- Department Women and Children's Health, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, UK
| | - Elizabeth Perkins
- Department of Primary Care and Mental Health, Institute of Population Health, Liverpool, UK
| | - Fauzia Paize
- Liverpool Women's NHS Foundation Trust, Liverpool, UK
| | | | - Melissa Gladstone
- Department Women and Children's Health, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, UK
| |
Collapse
|
3
|
Hurrion EM, Badawi N, Boyd RN, Morgan C, Gibbons K, Hennig S, Koorts P, Chauhan M, Bowling F, Flenady V, Kumar S, Dawson PA. SuPreme Study: a protocol to study the neuroprotective potential of sulfate among very/extremely preterm infants. BMJ Open 2023; 13:e076130. [PMID: 37451710 PMCID: PMC10351292 DOI: 10.1136/bmjopen-2023-076130] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/29/2023] [Accepted: 06/30/2023] [Indexed: 07/18/2023] Open
Abstract
INTRODUCTION Antenatal maternal magnesium sulfate (MgSO4) administration is a proven efficacious neuroprotective treatment reducing the risk of cerebral palsy (CP) among infants born preterm. Identification of the neuroprotective component with target plasma concentrations could lead to neonatal treatment with greater efficacy and accessibility. METHODS AND ANALYSIS This is a prospective observational cohort study, in three tertiary Australian centres. Participants are preterm infants, irrespective of antenatal MgSO4 exposure, born in 2013-2020 at 24+0 to 31+6 weeks gestation, and followed up to 2 years corrected age (CA) (to September 2023). 1595 participants are required (allowing for 17% deaths/loss to follow-up) to detect a clinically significant reduction (30% relative risk reduction) in CP when sulfate concentration at 7 days of age is 1 SD above the mean.A blood sample is collected on day 7 of age for plasma sulfate and magnesium measurement. In a subset of participants multiple blood and urine samples are collected for pharmacokinetic studies, between days 1-28, and in a further subset mother/infant blood is screened for genetic variants of sulfate transporter genes.The primary outcome is CP. Surviving infants are assessed for high risk of CP at 12-14 weeks CA according to Prechtl's Method to assess General Movements. Follow-up at 2 years CA includes assessments for CP, cognitive, language and motor development, and social/behavioural difficulties.Multivariate analyses will examine the association between day 7 plasma sulfate/magnesium concentrations with adverse neurodevelopmental outcomes. A population pharmacokinetic model for sulfate in the preterm infant will be created using non-linear mixed-effects modelling. ETHICS AND DISSEMINATION The study has been approved by Mater Misericordiae Ltd Human Research Ethics Committee (HREC/14/MHS/188). Results will be disseminated in peer-reviewed journal publications, and provided to the funding bodies. Using consumer input, a summary will be prepared for participants and consumer groups.
Collapse
Affiliation(s)
- Elizabeth M Hurrion
- Department of Newborn Services, Mater Mothers' Hospital, Brisbane, Queensland, Australia
- Mater Research Institute The University of Queensland, South Brisbane, Queensland, Australia
| | - Nadia Badawi
- Cerebral Palsy Alliance, The University of Sydney, Sydney, New South Wales, Australia
| | - Roslyn N Boyd
- Queensland Cerebral Palsy and Rehabilitation Research Centre, The University of Queensland, Saint Lucia, Queensland, Australia
| | - Catherine Morgan
- Cerebral Palsy Alliance, The University of Sydney, Sydney, New South Wales, Australia
| | - Kristen Gibbons
- Child Health Research Centre, Mater Research Institute The University of Queensland, South Brisbane, Queensland, Australia
| | - Stefanie Hennig
- School of Clinical Sciences, Queensland University of Technology Faculty of Health, Kelvin Grove, Queensland, Australia
- Integrated Drug Development, Certara Strategic Consulting, Certara LP, Princeton, New Jersey, USA
| | - Pieter Koorts
- Grantley Stable Neonatal Unit, Royal Brisbane and Women's Hospital, Herston, Queensland, Australia
| | - Manbir Chauhan
- Department of Newborn Care, Gold Coast University Hospital, Southport, Queensland, Australia
| | - Francis Bowling
- Walter and Eliza Hall Institute of Medical Research, Melbourne, Victoria, Australia
| | - Vicki Flenady
- Mater Research Institute The University of Queensland, South Brisbane, Queensland, Australia
| | - Sailesh Kumar
- Mater Research Institute The University of Queensland, South Brisbane, Queensland, Australia
| | - Paul A Dawson
- Mater Research Institute The University of Queensland, South Brisbane, Queensland, Australia
| |
Collapse
|
4
|
Aker K, Thomas N, Adde L, Koshy B, Martinez-Biarge M, Nakken I, Padankatti CS, Støen R. Prediction of outcome from MRI and general movements assessment after hypoxic-ischaemic encephalopathy in low-income and middle-income countries: data from a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 2022; 107:32-38. [PMID: 34112719 PMCID: PMC8685634 DOI: 10.1136/archdischild-2020-321309] [Citation(s) in RCA: 7] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/01/2020] [Revised: 03/26/2021] [Accepted: 04/20/2021] [Indexed: 11/12/2022]
Abstract
OBJECTIVE To evaluate the accuracy of neonatal MRI and general movements assessment (GMA) in predicting neurodevelopmental outcomes in infants with hypoxic-ischaemic encephalopathy (HIE). DESIGN Secondary analyses of a randomised controlled trial (RCT). SETTING Tertiary neonatal intensive care unit in India. METHODS Fifty infants with HIE were included in an RCT of therapeutic hypothermia (25 cooled and 25 non-cooled). All infants underwent brain MRI at day 5, GMA at 10-15 weeks and outcome assessments including Bayley Scales of Infant and Toddler Development, third edition, at 18 months. Associations between patterns of brain injury, presence/absence of fidgety movements (FMs) and outcomes were assessed. RESULTS Seventeen of 47 (36%) had adverse outcome (5 (21%) cooled vs 12 (52%) non-cooled, p=0.025). Eight infants died (four before an MRI, another three before GMA). Two developed severe cerebral palsy and seven had Bayley-III motor/cognitive composite score <85. Twelve (26%) had moderately/severely abnormal MRI and nine (23%) had absent FMs. The positive predictive value (95% CI) of an adverse outcome was 89% (53% to 98%) for moderate/severe basal ganglia and thalami (BGT) injury, 83% (56% to 95%) for absent/equivocal signal in the posterior limb of the internal capsule (PLIC) and 67% (38% to 87%) for absent FMs. Negative predictive values (95% CI) were 85% (74% to 92%) for normal/mild BGT injury, 90% (78% to 96%) for normal PLIC and 86% (74% to 93%) for present FMs. CONCLUSIONS Neonatal MRI and GMA predicted outcomes with high accuracy in infants with HIE. The GMA is a feasible low-cost method which can be used alone or complementary to MRI in low-resource settings to prognosticate and direct follow-up. TRIAL REGISTRATION NUMBER CTRI/2013/05/003693.
Collapse
Affiliation(s)
- Karoline Aker
- Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Faculty of Medicine and Health Sciences, Trondheim, Norway
- Department of Paediatrics, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
| | - Niranjan Thomas
- Department of Neonatology, Christian Medical College Vellore, Vellore, Tamil Nadu, India
| | - Lars Adde
- Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Faculty of Medicine and Health Sciences, Trondheim, Norway
- Clinic of Clinical Services, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
| | - Beena Koshy
- Department of Developmental Paediatrics, Christian Medical College Vellore, Vellore, Tamil Nadu, India
| | | | - Ingeborg Nakken
- Norwegian Advisory Unit for Functional MRI, Department of Radiology, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
| | - Caroline S Padankatti
- Department of Developmental Paediatrics, Christian Medical College Vellore, Vellore, Tamil Nadu, India
| | - Ragnhild Støen
- Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Faculty of Medicine and Health Sciences, Trondheim, Norway
- Department of Paediatrics, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
| |
Collapse
|
5
|
Age of Diagnosis, Fidelity and Acceptability of an Early Diagnosis Clinic for Cerebral Palsy: A Single Site Implementation Study. Brain Sci 2021; 11:brainsci11081074. [PMID: 34439692 PMCID: PMC8391606 DOI: 10.3390/brainsci11081074] [Citation(s) in RCA: 19] [Impact Index Per Article: 6.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/01/2021] [Revised: 08/06/2021] [Accepted: 08/11/2021] [Indexed: 11/24/2022] Open
Abstract
Cerebral palsy (CP) diagnosis is historically late, at between 12 and 24 months. We aimed to determine diagnosis age, fidelity to recommended tests and acceptability to parents and referrers of an early diagnosis clinic to implement a recent evidence-based clinical guideline for the early diagnosis of CP. A prospective observational case series of infants <12 months with detectable risks for CP attending our clinic was completed with data analysed cross-sectionally. Infants had a high risk of CP diagnosis at a mean age of 4.4 (standard deviation [SD] 2.3) months and CP diagnosis at 8.5 [4.1] months. Of the 109 infants seen, 57% had a diagnosis of CP or high risk of CP, showing high specificity to our inclusion criteria. Parent and referrer acceptability of the clinic was high. Paediatricians had the highest rate of referral (39%) followed by allied health (31%), primary carer (14%) and other health workers (16%). Fidelity to the guideline was also high. All infants referred <5 mths had the General Movements Assessment (GMA) and all except one had the Hammersmith Infant Neurological Examination (HINE) administered. N = 92 (84%) of infants seen had neuroimaging, including n = 53 (49%) who had magnetic resonance imaging (MRI), showing recommended tests are feasible. Referral to CP-specific interventions was at 4.7 [3.0] months, sometimes before referral to clinic. Clinicians can be confident CP can be diagnosed well under 12 months using recommended tools. This clinic model is acceptable to parents and referrers and supports access to CP-specific early interventions when they are likely to be most effective.
Collapse
|
6
|
Seesahai J, Luther M, Church PT, Maddalena P, Asztalos E, Rotter T, Banihani R. The assessment of general movements in term and late-preterm infants diagnosed with neonatal encephalopathy, as a predictive tool of cerebral palsy by 2 years of age-a scoping review. Syst Rev 2021; 10:226. [PMID: 34384482 PMCID: PMC8359053 DOI: 10.1186/s13643-021-01765-8] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 08/14/2020] [Accepted: 07/20/2021] [Indexed: 11/10/2022] Open
Abstract
BACKGROUND The General Movements Assessment is a non-invasive and cost-effective tool with demonstrated reliability for identifying infants at risk for cerebral palsy. Early detection of cerebral palsy allows for the implementation of early intervention and is associated with better functional outcomes. No review to date has summarized the utility of the General Movements Assessment to predict cerebral palsy in term and late-preterm infants diagnosed with neonatal encephalopathy. METHODS We conducted a scoping review involving infants born greater than or equal to 34 weeks gestational age to identify all available evidence and delineate research gaps. We extracted data on sensitivity, specificity, and positive and negative predictive values and described the strengths and limitations of the results. We searched five databases (MEDLINE, Embase, PsychINFO, Scopus, and CINAHL) and the General Movements Trust website. Two reviewers conducted all screening and data extraction independently. The articles were categorized according to key findings, and a critical appraisal was performed. RESULTS Only three studies, a cohort and two case series, met all of the inclusion criteria. The total number of participants was 118. None of the final eligible studies included late-preterm neonates. All three studies reported on sensitivity, specificity, and positive predictive and negative predictive values. An abnormal General Movement Assessment at 3-5 months has a high specificity (84.6-98%) for cerebral palsy with a similarly high negative predictive value (84.6-98%) when it was normal. Absent fidgety movements, in particular, are highly specific (96%) for moderate to severe cerebral palsy and carry a high negative predictive value (98%) when normal. In the time period between term and 4-5 months post-term, any cramped synchronized movements had results of 100% sensitivity and variable results for specificity, positive predictive value, and negative predictive value. CONCLUSIONS A normal General Movements Assessment at 3 months in a term high-risk infant is likely associated with a low risk for moderate/severe cerebral palsy. The finding of cramped synchronized General Movements is a strong predictor for the diagnosis of cerebral palsy by 2 years of age in the term population with neonatal encephalopathy. The deficit of high-quality research limits the applicability, and so the General Movements Assessment should not be used in isolation when assessing this population. SYSTEMATIC REVIEW REGISTRATION Title registration with Joanna Briggs Institute. URL: http://joannabriggswebdev.org/research/registered_titles.aspx .
Collapse
Affiliation(s)
- Judy Seesahai
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
| | - Maureen Luther
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
| | - Paige Terrien Church
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
- Department of Pediatrics, Division of Neonatal-Perinatal Medicine, University of Toronto, Toronto, Canada
| | - Patricia Maddalena
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
| | - Elizabeth Asztalos
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
- Department of Pediatrics, Division of Neonatal-Perinatal Medicine, University of Toronto, Toronto, Canada
| | | | - Rudaina Banihani
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada.
- Department of Pediatrics, Division of Neonatal-Perinatal Medicine, University of Toronto, Toronto, Canada.
- Newborn & Developmental Paediatrics, Sunnybrook Health Science Centre, 2075, Bayview Ave., Toronto, ON, M4N 3M5, Canada.
| |
Collapse
|