Artificial Intelligence: An Automated New World for Labs & Pathology
From - Laboratory Industry Report Artificial intelligence (AI) and machine learning are making significant inroads in the diagnostics realm. Here are four ways… . . . read more
Artificial intelligence (AI) and machine learning are making significant inroads in the diagnostics realm. Here are four ways the technology is changing the field and the business of pathology.
1. Diagnostics
The traditional method of reviewing tissue slides is time-consuming and subjective. Two pathologists assessing the same slide will only agree approximately 60% percent of the time, studies show. By contrast, deep learning tools can identify abnormalities with as much as 100% accuracy.
2. Analysis
Because it relies on algorithms, AI continually builds a library of knowledge. This ever-growing library is accessed each time the tools are used. As a result, pathologists, even those with extensive experience, have access to far more information than they otherwise would. This library also serves as a resource for long-term analysis, and new recommendations and protocols.
3. Detection
AI tools allow for faster review, and therefore earlier detection. Likewise, greater accuracy leads to fewer false positives. These advantages have important implications for patient health. They also have the potential to reduce costs. In addition, AI can be used to enhance radiology tools, in some instances eliminating the need for a tissue sample.
4. Treatment
Early detection and greater accuracy help clinicians identify the onset of disease, and better plan for treatment. Among the treatment options for which a patient may qualify are clinical trial enrollment.
Practical Impact & Applications
In terms of practical impact, applications of AI in the diagnostic field include:
- Use by researchers in a recent study to predict patient survival from pathology images in lung cancer;
- Development of new computational pathology software leveraging deep learning AI to aid in prostate and ovarian tumor tissue identification;
- Use of computer-aided diagnosis (CAD) powered by AI to assess diminutive colorectal polyps with 98% accuracy, according to a recent study;
- Development by Johns Hopkins researchers of a deep learning AI system to diagnose pancreatic cancer early.
Subscribe to view Essential
Start a Free Trial for immediate access to this article