Computer vision is a field of artificial intelligence that focuses on training computers to identify and understand information from images or videos. On a basic level, computer vision replicates the human capability of interpreting what we see.

Healthcare is one of the leading industries applying computer vision technology because it generates a vast amount of visual data from various imaging modalities.¹ Doctors and medical researchers are predominately using computer vision on static images, such as X-rays, MRIs, CT scans, ultrasounds, and cellular and molecular structures. Medical image interpretation is popular in specialties such as radiology, pathology, ophthalmology, and dermatology.² This is due to the ability to detect subtle changes or recognize patterns in medical images is critical for early detection, diagnosis, and effective treatment.

The primary focus for computer vision in healthcare has been medical image interpretation, but there is growing interest and research in video interpretation. This is because there are transformative applications in patient monitoring, surgical guidance, and behavioral analysis. Despite transformative potential, there has been slower progress in this space for two main reasons. First, computational techniques and datasets for video analysis evolved more slowly compared to images. Generally, images are easier to handle than video data, which contains multiple frames per second and higher volumes of data. Second, image interpretation has a greater clinical relevance. Nevertheless, advancements in computational power and algorithm development are poised to quicken the pace of video interpretation in healthcare, unlocking its full potential for real-time and procedural applications.

Uses for computer vision in healthcare

• Analyze medical images to diagnose diseases and abnormalities.
• Help surgeons during procedures by providing real-time guidance and feedback.
• Track human activity in healthcare settings, such as hospitals and nursing homes.
• Analyze patient behavior and expressions to assess pain levels, emotional states, and cognitive impairments.
• Track vital signs, such as heart rate and respiratory rate, by analyzing facial and body movements.
• Detect falls and abnormal movements, especially in older and at-risk patients.
• Track disease progression by analyzing changes in visual data over time.