As a promised follow-up from our last blog post, 10 (Plus One) Healthcare Trends on the Lab’s Radar for 2020, one more topic deserves attention on our laboratory hot topic list: Artificial Intelligence (AI). AI has been a buzzworthy term for a while now, as we keep hearing it is the next big thing. So much hype has led us to be speculative about its tangible contributions. However, expect AI to continue gaining traction as we move into 2020 because practical applications for AI are on the rise.
Global Access to Data & Productivity Gains
Digitalization of healthcare data is making it possible to transform unstructured data into shareable, actionable data. AI takes the rules-based decision support that we use today to the next level with its ability to recognize patterns beyond Boolean logic. AI’s ability to process and analyze huge amounts of data rapidly makes it an invaluable tool for healthcare.
Here are three laboratory-specific examples: AI can monitor predictive maintenance data to reduce system and analyzer downtime; 1 machine-learning is being tested to sort, prioritize, and route collection tubes; 1 and AI algorithms can be used to analyze and compare patient data trends across geographical locations and healthcare facilities and to aggregate results from various departments (lab, imaging, pathology, genetic, etc.) to help providers find the best treatment pathways faster. 1
Microbiology—AI Reading Cultures
AI and its interpretive algorithms are being used in microbiology to help with high workloads by assisting with negative cultures. This AI assistance then allows technologists to spend their time on more complex cultures. So far, AI has found success in screening cultures using chromogenic agar, as well as standard media for urine cultures, and for rapid disk diffusion antimicrobial susceptibility testing. 2 AI’s role in microbiology will help microbiologists gain productivity by lending their expertise where it is needed most.
AI Aids in Cancer Diagnoses
An AI algorithm or deep convolutional neural network has been proven able to “learn” the characteristics of the most common types of brain tumor and help predict diagnosis. 3 This methodology, developed at Michigan Medicine, is called stimulated Raman histology, which entails rapidly generated images of tumor tissue at the bedside to provide an accurate diagnostic prediction within minutes at the point of care. 3
In addition, AI has been proven to correctly identify cancers from images at an accuracy equivalent to expert radiologists. 4 AI’s rapid ability to scan images is being used in breast cancer screening, and while In its beginning stages, its potential to aid laboratory professionals and providers in diagnosis is huge.
The Future of AI
Healthcare is increasingly developing and using sophisticated AI solutions to derive meaning from and contextually classify all types of content, including structured and unstructured data. AI solutions can quickly read and understand huge stores of data without fatigue or distraction. AI and its algorithms can enable healthcare workers to gain rapid insights into patient data that are not easily gleaned without these tools.
Although in its beginning stages, AI is gaining traction in healthcare. This is projected to be an exciting year as the evolution of AI continues. Its abilities to support healthcare workers will continue to be tested and, as with any new healthcare technology, ethical decisions are needed around its use as we move forward. However, there is no doubt that AI implementation will continue, and that its future is full of exciting possibilities.
Orchard Software is committed to keeping you informed and being a trusted resource you can turn to for industry-related education. As always, we welcome your feedback. Follow us on Twitter at @OrchardSoftware, and feel free to respond to this post by emailing us at email@example.com.