AI (Artificial Intelligence), Ultrasound-Guided Nerve Blocks
Conditions
Brief summary
This study intends to have experts with over 10 years of experience in ultrasound-guided regional anesthesia acquire 1,000 target images. These images will be annotated by both an AI model and human experts, and the model's annotations will be evaluated by a panel of 10 specialists. The primary outcome is the accuracy of identifying the target anatomical structures in the annotated images. Secondary outcomes include the number of true positive, false positive, true negative, and false negative structures, the time required for annotation, and the potential application of the AI model in assisting the performance of ultrasound-guided regional blocks.
Interventions
DEVICEAI Annotation
Use the AI model to assist in annotating the target images.
OTHERmanual annotation
Use the manual to annotate the target images.
Sponsors
Shanghai Jiao Tong University Affiliated Sixth People's Hospital
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
DOUBLE (Subject, Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
18 Years to 70 Years
Healthy volunteers
No
Inclusion criteria
* Aged 18-70 years * BMI between 19-23 kg·m-² * ASA physical status classification I-II * Normal orientation and able to cooperate with physical examination.
Exclusion criteria
* Patient refusal * Infection, trauma, skin defect, deformity, or other abnormalities at the intended nerve block site * Pregnancy
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| The accuracy of identifying the target anatomical structures in the annotated images. | From enrollment to the end of treatment at 4 months |
Secondary
| Measure | Time frame |
|---|---|
| The number of false positive (FP) structures | From enrollment to the end of treatment at 4 months |
| The number of true negative (TN) structures | From enrollment to the end of treatment at 4 months |
| The number of true positive (TP) structures | From enrollment to the end of treatment at 4 months |
| The annotation time (time from the start to the end of annotation) | From enrollment to the end of treatment at 4 months |
| An evaluation of the AI model if applied to assist in performing ultrasound-guided regional blocks: a) its impact on block failure rate (increase, no effect, decrease); b) its impact on the incidence of adverse events (increase, no effect, decrease) | From enrollment to the end of treatment at 4 months |
| The number of false negative (FN) structures | From enrollment to the end of treatment at 4 months |
Countries
China
Outcome results
None listed