Deep Learning in Healthcare

 

Healthcare organizations of all sizes, types, and specialties are becoming increasingly interested in how artificial intelligence can support better patient care while reducing costs and improving efficiency. In a relatively short period of time, the availability and sophistication of AI have exploded, leaving providers, payers, and other stakeholders with a dizzying array of tools, technologies, and strategies to choose from. Learning just the language has been a major challenge for many organizations. There are subtle but important differences between key terms such as AI, Machine Learning, Deep Learning, and Semantic Computing. Understanding how data is ingested, analyzed, and returned to the end-user can have a major impact on expectations for accuracy and reliability, not necessarily affecting any investment required to shape an organization's data assets. to mention In order to choose efficiently and effectively between vendor products or to hire the right data science staff to develop algorithms in-house, healthcare organizations must be confident they have a firm grasp on the various flavors of artificial intelligence and How can they apply to specific use cases. Deep learning is a good place to start. This branch of artificial intelligence has become very rapidly transformative for healthcare, providing the ability to analyze data with never-before-seen speed and accuracy. But what exactly is deep learning, how does it differ from other machine learning strategies, and how can healthcare organizations leverage deep learning technologies to solve some of the most pressing problems?

What are the use cases for deep learning in health care?

Many of the industry's headlines for deep learning are currently related to small-scale pilots or research projects in their pre-commercial stages. However, deep learning is continually finding its way into innovative tools that have high-value applications in real-world clinical environments. Some of the most promising use cases include innovative patient-facing applications as well as some surprisingly established strategies for improving the health IT user experience. 

Imaging Analytics and Diagnostics 

One type of deep learning, known as a convolutional neural network (CNN), is particularly suitable for analyzing images, such as MRI results or X-rays. CNN's are designed with the assumption that they will process images, according to computer science experts at Stanford University, allowing networks to operate more efficiently and handle larger images. As a result, some CNNs are exceeding – or even surpassing – the accuracy of human diagnostics when identifying important features in clinical imaging studies. In June 2018, a study in the Annals of Oncology showed that a convolutional neural network trained to analyze dermatological images identified melanoma with ten percent higher specificity than human physicians. Even when human physicians were equipped with patients' background information, such as age, gender, and body size of the suspected characteristic, CNN outperformed dermatologists by about 7 percent. "Our data clearly show that the CNN algorithm can be a suitable tool to assist clinicians in detecting melanoma regardless of their individual level of experience and training," said the team of researchers from several German academic institutions. In addition to being highly accurate, deep learning tools are fast. Researchers at Mount Sinai Icahn School of Medicine have developed a deep neural network capable of diagnosing important neurological conditions, such as stroke and brain hemorrhage, up to 150 times faster than a human radiologist.

What is the future of deep learning in health care?

As complex as these pilots and projects may be, they represent only the beginning of the role of deep learning in healthcare analytics. The excitement and interest about deep learning are everywhere, capturing the imaginations of regulators and rule-makers, private companies, care providers, and even patients. The Office of the National Coordinator (ONC) is an organization that has particularly high hopes for deep learning, and it is already applauding some developers for achieving remarkable results. In a recent report on the state of AI in healthcare settings, the agency noted that some deep learning algorithms have already produced "transformative" results. "There has been a significant demonstration of the potential usefulness of deep learning-based artificial intelligence approaches for use in medical diagnostics," the report said. “Where good training sets represent the highest level of medical expertise, the application of deep learning algorithms in clinical settings offers the potential to deliver consistently high-quality results.” The report highlights early successes in diabetic retinal screening and classification of skin cancer as two areas where deep learning is already changing the status quo. On the clinical side, imaging analytics is likely to be a focal point for the foreseeable future, due to the fact that deep learning has already begun on many high-value applications.