Predicting energy requirements in traumatic brain injury using 3D body shape app and indirect calorimetry.

Patients with traumatic brain injury (TBI) have dynamic and complex nutritional needs that are often not met whilst they are being treated in intensive care. Malnutrition is estimated to occur in 40-50% of critically ill patients (Sriram et al, 2010) and is associated with increased morbidity, length of stay, infections, organ failure, prolonged mechanical ventilation and mortality (Moisey et al, 2013). However, several studies have shown that overfeeding critically ill patients can cause significant harm such as increased hyperglycemia, hepatic steatosis and mortality (Moonen et al, 2021). Despite the adverse outcomes associated with under and overfeeding in critically ill patients, optimisation of nutrition is currently prevented by the lack of resources to assess changes in energy requirements and body composition assessments to monitor the response to nutrition (Wischmeyer et al, 2023). The objective of this research project is to develop and implement a 3D body shape app into a critical care setting to provide a simple and accessible resource for the prediction of body composition in ventilated TBI patients. We have brought together a team of experts in critical care including Prof Puthucheary from Queens University London, critical care consultants and dieticians from Cambridge and metabolic and body composition experts from NIHR Cambridge Clinical Research Facility (CRF) and University of Cambridge Engineering. The 3D body shape app uses computer vision and machine learning algorithms to construct 3D body meshes from a smart phone image to predict body composition. A recent collaboration between the University of Cambridge department of Engineering, MRC Epidemiology Unit and NIHR CRF has validated 3D body meshes from 17,461 DXA silhouettes and 119 smartphone 3D meshes in healthy adults (Qiao et al, 2024). For this work, an Addenbrookes Hospital PPI panel were approached for their comments and suggestions on the 3D body shape app. Their contributions were included in the protocol design and information leaflets, addressing image and data confidentiality queries, lay language and clear instructions of use. Energy requirements in TBI are currently estimated using previously published prediction equations but their use in critical care has repeatedly shown inaccuracies compared to indirect Calorimetry (IC) (Tatucu-Babet, 2016, Waele et al 2021). Watson et al has previously described the importance of body composition in the prediction of energy requirements in patients where energy expenditure or body composition is altered by disease or treatment (Watson et al, 2014, 2019, 2023). Lean body mass (LBM) accounts for 70-80% of resting energy expenditure (REE) (Bosy-Westphal et al, 2009), however, in patients with critical illness they present with significant and rapid reductions in LBM which in turn triggers modifications of REE and requirements (Berger et al, 2019). Clinically accredited calorimeters are expensive and require expertise in the data collection, analysis and ongoing maintenance all of which are barriers to widespread adoption. Improved accuracy in predicting the variation in energy requirements by monitoring the response in body composition, as we propose to do by developing the 3D body shape app, is crucial for progression of managing nutrition in TBI patients.