Article Text
PDF
Abstract
Please confirm that an ethics committee approval has been applied for or granted: Not relevant
Background and Aims Formalised ultrasound guided regional anaesthesia (USGRA) training is resource intensive and often difficult to access. Assistive artificial intelligence (AI) is an emerging technology with potential to enhance training and provision of USGRA. We aim to evaluate if ScanNAV(TM) (Intelligent Ultrasound Limited) can enhance USGRA image acquisition of a ‘Plan-A Block’ for the non-expert in the absence of formalised training.
Methods 18 anaesthetists performed sonoanatomy on live models for two pre-selected Plan-A blocks, one with and one without prior formal training. ScanNAV(TM) was used in the latter alongside ScanNAV(TM) tutorial videos and RA-UK infographic material for reference. 2 expert assessors made objective assessments for each using a protocolised data collection tool.
Results 15/18 (83.3%) participants successfully acquired appropriate ultrasound images for a Plan-A Block using ScanNAV(TM) and reference materials with no formal prior training. Sonoanatomy scans were performed faster on average in block procedures that had received prior formal USGRA training. The adductor canal block had an average procedure time of 23.52seconds vs 179.17seconds (t-test value 2.74; p-value 0.0168). Where formal training had taken place, participants scored higher in identifying key structures (ASRA-ESRA Delphi consensus) across all Plan-A blocks, as well as accuracy grade of image acquisition and needle path safety.
depict subjective feedback of 18 anaesthetists following use of ScanNAV(TM)
Conclusions Assistive AI, e.g. ScanNAV(TM) may facilitate image acquisition and identification of key sonoanatomical stuctures in the absence of formalised training. This technology should be used as an adjunct, not a replacement, for formalised training as objective assessment in speed, accuracy and safety were seen to be superior in this subgroup.