Quantification of head and trunk control for children with neuromotor and neuromuscular disorders

Head and trunk control (trunk control) is mandatory for effective performance of everyday functional activities such as use of vision, sitting or any upper limb activity. Children with cerebral palsy (CP), spinal muscular atrophy (SMA) and inherited neuromuscular disorders (NMD) experience impairments in trunk control which are a fundamental component of their condition.

Children with CP show a direct relationship between poor trunk control (orienting the head, sitting balance) and compromised function including communication by eye gaze, use of both hands for learning, play and social development, mobility and urinary incontinence. Children with more severe CP regress in their functional abilities from age 7 years. Accurate diagnosis of trunk control can enable individualised physiotherapy, when combined with treatment as usual, to improve general motor function even in severe CP (preprint). An accurate, objective tool is needed to provide evidence encompassing multiple therapies and centres.

Infants with SMA1 have similar difficulty acquiring trunk control as do children with severe CP. Using Nusinersen, these children now survive, but this drug is expensive. An accurate, objective test is needed to monitor effectiveness of this and alternative drugs.

The increasing muscle weakness in children with NMD (SMA2, congenital myopathy, congenital muscular dystrophy, Duchenne muscular dystrophy) leads to deterioration in trunk control with subsequent scoliosis. Objective measurement sensitive enough to detect changes in trunk control, pre-scoliosis, would enable appropriate treatment such as physiotherapy at the optimal time.

Our hypothesis is that segmental trunk control can be measured objectively and efficiently, classified to seven segmental levels. This objective measurement would promote benefits in terms of new understanding of the mechanisms of trunk control, monitoring treatment and providing evidence to inform practice for all therapy strategies that aim to improve postural control and associated motor function. All current assessments suitable for use in a clinic are validated but rely on clinical judgement that must be maintained over time and between therapists, and many cannot be used for children unable to sit by themselves. They also consider the trunk as a single unit, thus losing the accurate and detailed assessment required for specialised therapy. We anticipate this objective assessment tool could lead to interventions to enhance trunk control, and thus improvement of functional skills, for thousands of children worldwide including those with greater severity of CP.

Our project will i) develop live imaging analysis technology to automate the assessment of trunk control in sitting, ii) provide a cost effective and accessible tool usable within any clinic without increasing assessment time and iii) publish an online interactive database disseminating a public standard, a reference and a training resource for measuring trunk control, increasing understanding and enhancing expertise. Assessments of trunk control of children in sitting (400 CP, 20 SMA1, 40 typically developing (TD), 40 NMD) will be recorded using 3D cameras available for the home market. This dataset will include examples of all trunk control problems, body morphology and height that the clinical tool would be required to meet. Machine learning methods will be used to train and validate the automated diagnosis. The tool will be validated against expert clinical assessment and deployed as a laptop, software and two cameras, suitable for use by clinical staff in a routine clinical environment. To prepare power calculations necessary to plan clinical trials, we will collect a longitudinal dataset of 20 children with CP and 5 infants with SMA 1 during a twelve-month period.

This imaging analysis tool will be trained to provide detailed, objective information of the segmental trunk control status of children with cerebral palsy (CP), and spinal muscular atrophy (SMA). Two core processes in the identification of trunk control will be automated: i) measurement of alignment of head and trunk segments to the vertical and ii) identification of when the hands and arms are free from contact with the body and any external surface.

A cross sectional dataset (400 CP, 20 SMA1, 40 TD) and small exploratory dataset (40 NMD) will be acquired with each child tested using the Segmental Assessment of Trunk Control (SATCo). This subjective test will be recorded using two "RGB-D" cameras each giving 3D information as raw video and depth images. The CP sample (Manchester, Liverpool and Cheshire community services) and SMA sample (Oswestry Hospital and linked centres) would include the range of trunk control problems and body morphology of these populations. Two experienced SATCo clinicians, verified by a third, will label the video identifying i) segmental alignment i.e. when the child loses postural alignment; ii) upper limb support i.e. when the upper limbs contact the trunk or bench; and iii) trunk control status. Using contemporary machine learning methods, the raw RGB-D images and associated labels will train and test our networks through a cross validation process. (For motor function c.f. Case for Support).

Exploitation of results includes: a) publication of a substantial public database of (400+20) cases representing trunk control impairment in children with CP and SMA1. This disseminates a public reference and training resource with data, software and instructions enabling inexperienced clinicians to reach the highest standard of clinical assessment of trunk control; b) acquisition of a 6-12-month longitudinal dataset of (20 CP, 5 SMA) children within our clinical network to provide power calculations necessary to plan clinical trials.

This research will principally benefit children with cerebral palsy (CP) and with spinal muscular atrophy (SMA), their families and communities, and the health and social services which support them. Other neuromuscular disorders (NMD) benefiting include congenital myopathy, congenital muscular dystrophy, SMA2 and Duchenne muscular dystrophy. The clinical tool will apply to all therapy strategies, from drug therapy to the physical therapies, based on the need to monitor and improve postural control which is recognised as having a direct link to functional ability.

Recent research shows a direct relationship between poor trunk control (orienting the head, sitting balance, mobility) and compromised function including urinary incontinence, communication by eye gaze, use of both hands for learning, play and social development in children with CP. The use of this new tool will benefit children with CP by increased functional abilities resulting from more effective (physio)therapy treatment based on a greater understanding of the mechanisms of trunk control problems. Children with SMA will benefit from more accurate monitoring of trials of new drugs targeting the genetic defects responsible for these conditions. Children with inherited neuromuscular conditions (NMD) will benefit from treatments more directly informed by accurate measurement of their trunk control.

Our main objective will facilitate clinical research programmes, to involve clinically-based (physio)therapists in regular practice. The tool to measure trunk control will not increase assessment time or adversely impact on routine clinical activities. This potential to increase clinical physiotherapy involvement in research has been long awaited. In accordance with three of the 'Childhood Disability Top 10' research priorities (1,2 and 7) published by the CSP/James Lind Alliance, this tool will provide the detail of trunk impairment to enable optimal treatment. This tool will allow, for the first time, objective assessment facilitating: i) a greater understanding of the means by which therapy for trunk control problems work so that progression of these therapies is based on sound criteria 1st priority; ii) clear evidence of whether therapy to improve trunk control is delivered effectively 1st prior.; iii) a means of relating segmental trunk control status to functional skills, wider aspects such as social interaction 2nd prior., and urinary incontinence 7th prior..

Paediatricians, paediatric (physio)therapists and others concerned with the medical care and long-term welfare of these children will benefit from this new objective evidence base. The publication of a database of trunk control status in children with CP would commence a major teaching resource on trunk control for use worldwide with other clinicians adding to this database, again with multi-professional benefit. Training of (physio)therapists to use this system, and dissemination of the tool within the local and national clinical network and, subsequently worldwide will increase professional skill sets.

Funders (NHS, Charity and/or private individuals) will benefit financially. There is potential saving in regular equipment costs following increase in trunk control in children. For example, the cost difference between complex fully supportive seating and seating that is appropriate for a child with some extent of trunk control is a reduction of at least £500 (data from therapy equipment manufacturer). An increase in active trunk control can delay or eliminate the need for expensive spinal bracing or the need for surgery to stabilize the spine in the event of increasing spinal deformity; this is relatively common with greater severity of CP. Further cost savings are in reduced inaccurate diagnosis leading to inappropriate treatment. The beneficiaries of this research are many but all lead back to the primary beneficiaries, the children with neuromotor disability and with neuromuscular disease.