Artificial Intelligence in Three-Dimensional Complete-Physique Images for Skin Most cancers Surveillance

Abstract

Artificial intelligence is commonly utilized in research that detect pores and skin most cancers, however most work makes use of pictures of chosen lesions. These pictures are helpful for classification. However, they don’t absolutely match actual screening conditions. In lesion surveillance, clinicians want to examine many lesions throughout the entire physique and monitor whether or not lesions change over time. Three-dimensional total-body images is ready to seize a wider pores and skin floor and may help lesion choice, lesion triage, danger evaluation, and follow-up comparability. This strategy offers synthetic intelligence extra alternative to research the affected person past one chosen lesion. This Mini Review discusses current proof on synthetic intelligence-assisted three-dimensional total-body images for pores and skin most cancers surveillance. In present purposes, this strategy can be utilized in automated triage and lesion detection, multimodal danger prediction, phenotype extraction, and longitudinal monitoring. The reviewed papers present early progress, however the proof stays restricted. Skin tone reporting, workflow integration, dataset transparency, false-positive and false-negative hurt, value, and fairness stay key adoption points. The goal is to give attention to the shift from selected-lesion classification to whole-body imaging. Clinical worth shouldn’t be judged solely by lesion-level accuracy, but additionally by whether or not synthetic intelligence-assisted three-dimensional total-body images can enhance patient-level surveillance throughout the entire pores and skin floor and over time. Stronger potential, longitudinal, workflow, value, and fairness proof remains to be wanted earlier than routine scientific adoption in observe.