AI Image Enhancement Boosts Carotid Plaque Detection in Primary Care

Researchers from the Affiliated Changsha Central Hospital and Macao Polytechnic University developed a super-resolution model known as Hyper-CycleGAN. Designed to process static images exported from handheld ultrasound devices, the tool sharpens boundaries between arterial plaque and blood channels, making small or faint blockages easier to identify. In a study of 450 adults, the AI-enhanced images identified 94.8% of plaques, outperforming standard handheld imaging which captured 87.6%. Beyond simple detection, the technology improved the assessment of vessel narrowing from moderate to excellent agreement with reference standards.

While the model increased the detection of potentially unstable plaques from 47.4% to 63.2%, authors emphasize that the system serves as a triage support tool rather than a diagnostic replacement for clinical decision-making. In an accompanying editorial, Daria Szkwarko and her colleagues noted that while the technology successfully brings specialized screening into community clinics, its clinical value remains tied to existing infrastructure. They warned that without robust referral pathways and quality assurance protocols, early detection may not translate into improved patient outcomes.