One area where artificial intelligence and machine learning are set to make an impact is with the detection of cancer. This is a big step-change, considering that for over 150 years the classic approach to pathology has involved a sample of tissue being examined under an optical microscope, with the pathologist drawing upon his or her experiences to determine whether the tissue sample is cancerous.
New developments are changing this. For example, digital whole slide imaging enables the pathologist to capture an entire tissue sample on a slide. The image can then be digitally captured and analyzed with the help of specialized digital pathology software. Such software, according to Dr. David West Jr., who runs an imaging company called Proscia, allows laboratories to send images between each other, so that different opinions can be obtained. The digitalization also enables the software to perform part of the analysis. Here image analysis algorithms can engage in immunohistochemistry quantification. This process, whereby the algorithm can help to interpret the material, improves standardization and the consistency of decision making. It also speeds up analysis.
The use of artificial intelligence and machine learning means that the deep neural networks of the diagnostic machine can be trained to recognize broad or specific patterns on a whole slide image. The software can also interpret features in the tissue and make prediction (such as metastasis and recurrence). Where cancer is present, this can also be classified by the software (in terms of staging, grading, and differential diagnosis). This works on the basis of the digital image being interpreted by the machine in terms of numbers and the pattern of numerical variations is examined.
As technology has advanced many barriers to adoption have been removed, such as concerns with limiting technology, image quality, problems with scanning all materials digital slide storage, and ergonomics.
Through multiple screening the artificial intelligence becomes better and the ability to spot predictive biomarkers becomes better, based on precise measurements of histological patterns. An example of such a system is a platform provided by Aperio ePathology (there are many others). These systems offer further advantages by allowing workflow is integrated into the institution’s overall operational environment.
Based on the success of digital pathology platforms, the U.S. Food and Drug Administration (FDA) gave approval for such systems to be used with the U.S. in May 2017. With this, Dr. who is the Director of the Office of In Vitro Diagnostics and Radiological Health in the FDA’s Center for Devices and Radiological Health, stated: “Because the system digitizes slides that would otherwise be stored in physical files, it also provides a streamlined slide storage and retrieval system that may ultimately help make critical health information available to pathologists, other health care professionals and patients faster.”
