Artificial Intelligence:

Future Technology for Cancer Diagnostics

Categories: Winter 2019
After a patient’s biopsy, a pathologist examines the tissue specimen to see whether any abnormal growth is benign or malignant. However, errors in identifying the tissue’s characteristics can occur. A current area of scientific exploration concerns whether using new technologies, such as machine learning and artificial intelligence (AI), to analyze pathology slides could increase a pathologist’s accuracy and efficiency.

After a patient’s biopsy, a pathologist examines the tissue specimen to see whether any abnormal growth is benign or malignant. However, errors in identifying the tissue’s characteristics can occur. A current area of scientific exploration concerns whether using new technologies, such as machine learning and artificial intelligence (AI), to analyze pathology slides could increase a pathologist’s accuracy and efficiency.

At this time, AI technology is not ready for clinical use. A number of people are working to improve the technology and assess its applicability in scientific studies.

However, the U.S. Food and Drug Administration (FDA) has granted the company Paige.AI a Breakthrough Device Designation. The designation is granted for technologies that have the potential to improve diagnostics or treatment for diseases such as cancer.

Paige.AI is developing machine learning algorithms to review digitized pathology slides and identify characteristics of benign and malignant tissue. Although the applications could be applied to all cancers, the company is starting with prostate cancer.

Subscribe to Quest

"*" indicates required fields

Email

This email address is only for subscribing or unsubscribing to or from Quest. Dr. Catalona cannot respond to questions concerning your Personal Health Information (PHI) that could identify you as an individual. This is a "cyber-security measure" for your protection that will help prevent anyone from legally or illegally accessing your personal health information.

Delivery Method (How do you want to receive your Quest publication?)*
Address*