We are in the midst of a technological renaissance, with artificial intelligence (AI) leading the way. Among the many promising tools within AI, computer vision arguably stands out the most for its ability to create incredible innovations, especially in the medical field. However, before exploring how computer vision can transform healthcare, we need to clearly understand what AI is and the key subfields that uphold it.
Central Technology
First, it’s important we take a step back and view computer vision from the broader hierarchy of AI. This structure starts with the foundation of AI at its base and works its way up through machine learning before culminating with computer vision.
The current infatuation with artificial intelligence has led to the term getting thrown around endlessly with a lack of context. At its core, AI is the technology that allows computers to rival the intelligence of humans.
Although commonly lumped together, AI and machine learning have distinct meanings. In fact, machine learning is a branch of AI that focuses on using massive amounts of data to allow machines to gain knowledge and learn from it, rather than relying on pre-programmed responses or other human intervention. Mimicking neurons in the human brain, machine learning models are constantly trained through neural networks that grant them the ability to analyze and extract information.
Computer vision is by far the most specialized field among all the technologies mentioned. Simply put, computer vision builds on top of machine learning models to allow computers to process and “see” visual data like images and videos. This allows computers to not only understand text-based problems, but also solve real issues that require “sight”, like surgery or anti-theft video tracking. Computer vision is such a practical branch of AI because, after all, a picture is worth a thousand words.
The Application of Computer Vision for Medical Imaging
Currently, computer vision is actively playing a role in several medical fields, including radiology and pathology. In regard to radiology, computer vision is being used to interpret results from X-rays, ultrasound and MRIs. In particular, machines are using computer vision to analyze images provided by these technologies for efficiency and accuracy. Similarly, in the field of pathology, computer vision is being used in systems to analyze tissues in the human body to detect a variety of diseases, such as cancer. As we continue to refine these machines, they are becoming increasingly useful for critical diagnostics that require nearly flawless accuracy and execution.
Below is an example that demonstrates how beneficial computer vision can be for human health. Take a moment to consider the context of brain tumors— many of which grow at a rapid rate and need quick medical intervention to be removed immediately. Healthcare professionals can use the assistance of computer vision to precisely segment and identify the tumor, reducing turnaround time for such a critical surgery. This approach increases both resource and time efficiency, relative to the alternative of purely manual analysis.
Brain tumor segmentation and identification using artificial intelligence and computer vision (Image courtesy of Nahiyan Habib Khan)
Future Implications
The future is bright for computer vision with no signs of halting progression. As more medical data becomes available and statistical noise gradually reduces, AI will reach a point at which the diagnoses will be nearly perfect. Moreover, with the combination of augmented reality and computer vision, more complex healthcare issues can be addressed and resolved. For instance, with overcrowding in hospitals being a prominent issue in our current society, online consulting is a promising proposal that can be implemented through merging computer vision and augmented reality. In non-urgent medical situations, patients would be able to attend online consultations, in which doctors would communicate with them using the ever-advancing technologies. To make matters even better, while also partaking in online appointments, systems using computer vision could provide real-time interpretation to the medical professionals, making their job all the easier. Bringing this consideration into light demonstrates how medical copiloting with computer vision may just be currently sparking the next great medical revolution.
