A survey of blood oxygen saturation assessment from video

The combination of technical advances and the emergence of artificial intelligence in the medical field has given rise to new oxygen saturation (SpO 2 ) measurement processes. In recent years, camera-based estimation of SpO 2 has attracted increasing interest. This technique offers an optical way to acquire SpO 2 in a very practical way. This review provides an overview of the imaging photoplethysmography-based techniques developed to achieve SpO 2 prediction from video. A total of 81 articles have been assessed and are included in this review. The results obtained using artificial intelligence-based methods in particular are very promising in view of traditional methods. This suggests that AI-based methods will have an important future role to play in medical devices. The prospects are now focused on efforts to improve the robustness and performance of the models, and in fine, to stamp these medical devices in accordance with current standards (CE or FDA) to meet the demands of the medical profession and patient safety.