Prospective evaluation of novice-acquired three-dimensional point-of-care ultrasound for carotid stenosis

Artificial intelligence and point-of-care ultrasound: Benefits, limitations, and implications for the future

The utilization of artificial intelligence (AI) in medical imaging has become a rapidly growing field as a means to address contemporary demands and challenges of healthcare. Among the emerging applications of AI is point-of-care ultrasound (POCUS), in which the combination of these two technologies has garnered recent attention in research and clinical settings. In this Controversies paper, we will discuss the benefits, limitations, and future considerations of AI in POCUS for patients, clinicians, and healthcare systems.

Accuracy and Feasibility of Three-Dimensional Ultrasound Testing in Eye Clinic and Emergency Department Patients with Vision Complaints

BACKGROUND

Ocular emergencies comprise 2-3% of emergency department (ED) visits, with retinal detachment requiring emergency surgery. Two-dimensional ultrasound is a rapid bedside tool but is highly operator dependent.

OBJECTIVE

We determined three-dimensional ultrasound (3DUS) feasibility, acceptability, and usability in eye pathology detection using the ophthalmologist examination as reference standard.

METHODS

We performed a prospective, blinded cohort study of a 3DUS-enabling device in 30 eye clinic and ED patients with visual symptoms and calculated 3DUS performance characteristics. Two expert readers interpreted the 3DUS images for pathology. All participants completed surveys.

RESULTS

3DUS sensitivity was 0.81, specificity 0.73, positive predictive value 0.54, negative predictive value 0.91, and likelihood ratio (LR)+/LR- 3.03 and 0.26, respectively. Novice and expert sonographers had "substantial" agreement in correct diagnosis of abnormal vs. normal (κ = 0.68, 95% confidence interval 0.48-0.88). Most patients indicated that 3DUS is fast, comfortable, helps them understand their problem, and improves provider interaction/care, and all sonographers agreed; 4/5 sonographers felt confident performing ultrasound. Expert readers correctly identified an abnormal eye in 83/120 scans (76%) and correct diagnosis in 72/120 scans (65%), with no statistical difference between novice (79%; 69%) and expert (72%; 61%) sonographers (p = 0.39, p = 0.55), suggesting reduced operator dependence. Reader diagnosis confidence and image quality varied widely. Image acquisition times were fast for novice (mean 225 ± 83 s) and expert (201 ± 51) sonographers, with fast expert reader interpretation times (225 ± 136).

CONCLUSIONS

A 3DUS-enabling device demonstrates a sensitivity of 0.81 and specificity of 0.73 for disease detection, fast image acquisition, and may reduce operator dependence for detecting emergent retinal pathologies. Further technological development is needed to improve diagnostic accuracy in identifying and characterizing retinal pathology.