Head Injury
Conditions
Brief summary
This study aims to test a new AI-powered portable MRI scanner that can quickly identify whether a brain scan is normal or abnormal. Currently, standard MRI scans are expensive and have long waiting times. Our goal is to see if a smaller, cheaper, and more accessible MRI scanner-combined with artificial intelligence (AI)-can help doctors identify abnormalities faster and improve patient care. We will invite patients from King's College Hospital (KCH) who are already having a standard MRI scan. They will be asked to have an extra scan using the portable MRI, which takes about 60 minutes. The AI tool will then analyse these scans and compare its results to those of expert radiologists. By the end of the study, we hope to prove whether portable MRI with AI can be used in hospitals and GP clinics, making brain scans more accessible, reducing wait times, and helping doctors prioritise urgent cases. This study is funded by the Medical Research Council (MRC) and has been approved by UK research ethics committees.
Interventions
DEVICEPortable, ultra-low-field MRI scanner
This study evaluates a portable, ultra-low-field MRI scanner (the Hyperfine Swoop) combined with artificial intelligence (AI) to detect brain abnormalities. Patients undergoing a standard brain MRI scan will be invited to have an additional portable MRI scan within 30 days of their clinical scan. The portable MRI scan will take approximately 60 minutes, using multiple imaging sequences, including T2-weighted scans. The AI system will then analyse the portable MRI images and categorise them as normal or abnormal. The results will be compared with expert neuroradiologist reports from standard MRI scans to validate accuracy. This intervention aims to assess whether portable MRI with AI can provide a low-cost, accessible alternative to standard MRI, potentially improving triage and reducing waiting times for patients requiring urgent brain imaging.
Sponsors
King's College London
King's College Hospital NHS Trust
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
SCREENING
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
Adults ≥18 years old. Undergoing standard brain MRI including T2-weighted sequences.
Exclusion criteria
Contraindications to MRI (e.g. pacemaker, pregnancy). Poor quality MRI scans without a neuroradiology report.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Accuracy of AI toll for triaging scans as normal or abnormal | 36 months | Ai Triage accuracy compared with consultant neuroradiologists assessment. |
Secondary
| Measure | Time frame |
|---|---|
| Generalisability of AI tool (evaluated on external dataset). | 36 months |
| Patient acceptability of portable MRI (survey/interviews) | 36 months |
| Feasibility of integrating portable MRI in clinical pathways. | 36 months |
Contacts
Primary ContactFrantisek Vasa, PhD
Backup ContactGiusi Manfredi, PhD
Outcome results
None listed