Op-Ed 1: The Importance of Computer Vision
Introduction:
While computer vision has been around for decades (see Yan LeCun demonstrating LeNet-1 at Bless Labs in 1989), recent advancements in artificial intelligence has catapulted its vast applications into the modern world. Since the days of LeCun, the applications of computer vision have taken the world by storm. Computer vision is now used by governments, corporations, and individuals to source data, synthesize complex information, and inform crucial decision-making processes.
As an overarching definition, computer vision is a field of artificial intelligence that allows computers to read and interpret visual information. The primary goals of computer vision include image recognition, object tracking, image segmentation, 3D scene reconstruction, pose estimation, and object detection. Computer vision has made significant advancements in recent years, attributable to the development of deep learning networks. Specifically, convolutional neural networks (CNNs) have significantly improved the accuracy of computer vision systems.
Due to these advancements in technology and innovation, the future of computer vision is incredibly promising. In this Op-Ed, I want to examine the future application of computer vision in the fields of healthcare and the environment; specifically I would like to better understand how computer vision will help reduce the cost of healthcare and curb climate change.
Healthcare:
In an industry that is rife with overspend, mis-utilization, and varying quality of care, computer vision is poised to bring transformative changes to the healthcare industry. In the coming years, computer vision and artificial intelligence will help to standardize diagnoses and optimize surgical procedures. These advancements will have a material impact on the quality of patient care, appropriation of resources, and overall medical spend.
With respect to diagnoses, medical image analysis and telemedicine are two sub-segments that will benefit from computer vision. Computer vision will continue to enhance the analysis of medical images, including X-rays, MRIs, CT scans, and histopathology slides. Advanced algorithms will aid radiologists and pathologists in detecting and diagnosing diseases with higher accuracy. Similar applications will be implemented in the field of telemedicine. As telemedicine grows, it is increasingly important to make it as effective as possible. Particularly for skin conditions, wounds, or rashes, physicians can use computer vision to more accurately diagnose and treat ailments. Technological advances will allow physicians to leverage computer vision for more accurate patient assessment and treatment.
Advancements in computer vision will also impact and aid physicians in the surgery room. During surgical procedures, computer vision can provide real-time assistance to surgeons. It can enhance surgical navigation, offer augmented reality displays, and provide guidance for precise incisions, tissue removal, and suturing. Surgical robots and medical robots will increasingly use computer vision to perform delicate and complex procedures. These robots can assist in minimally invasive surgeries, allowing for greater precision and reduced invasiveness. This can lead to improved surgical outcomes and reduced complications.
As the United States’ geriatric population grows (and outnumbers the young, tax-paying population) a cost burden looms overhead. Healthcare professionals will need to figure out a way to reduce the spend per CMS beneficiary. One way mitigate this imminent expense is to improve the quality of care and make the patient journey efficient. To accomplish this optimization of care, technologies like computer vision play like a key role. Applications such as diagnosis optimization and surgical assistance are two ways in which computer vision will improve the quality of care and contribute to crucial cost reductions.
Environment:
In addition to the technological innovations required by the healthcare industry, curbing climate change is an objective that will require significant help from technology. Inevitably, computer vision technologies will play an important role.
With respect to forest/land management, computer vision technologies (which are fed data from satellite images) are currently tracking changes in forest cover to identify illegal logging activities. Authorities can use this information to alert authorities and curb deforestation. Analyses from satellite images also inform stakeholders on areas that are suitable for reforestation and the type of trees should be planted. Similar analysis will be utilized in agriculture and crop management. Today, computer vision is used by the agricultural industry to track crop health and growth, detect diseases, and inform decision about fertilization cycles. In addition to crop management, computer vision will be used to monitor soil and erosion. These efforts will allow farmers to better practice sustainable farming to mitigate the consequences of erosion.
On the most macro of levels, computer vision will help us observe Earth. By analyzing satellite images of and videos of the earth, computer vision will allow stakeholders (e.g., governments, corporations, regulatory bodies) to track large changes in the ecosystem (e.g., ice melt, land mass movements). These types of observations will inform stakeholders on the quality of the Earth’s ecosystem, and enable predictive modeling — which should yield optimal legislation and decision-making.
Conclusion:
Whether Yan LeCun (circa 1989) could have ever predicted where computer vision would be today — I’m not sure. Nevertheless, I am thankful for this technological advancement. Without computer vision technologies, I do not know if humankind could successfully undertake momentous initiatives such as reforming the US Healthcare system or curbing climate change. The application of computer vision in healthcare and the environment is so important that I believe computer vision is one of the most significant technologies born from artificial intelligence.
