Don't fear it, embrace it: 3 ways machine learning can enhance your role
Machine learning is not science fiction—it’s impacting your business today. Yet, when you think about integrating this technology into your lab, you may have some concerns and questions: Can this technology do my job better than I can? Will my role be diminished? Will it eventually replace me as a pathologist?
These questions are common. But, the reality today is that machine learning cannot replace you or do your job better than you. Rather, it can support you in your role, to help you elevate the value you bring to your lab—and your institution—every day.
That’s why LabLeaders is dedicated to transforming your uncertainty into confidence. Here are our top three ways you can leverage machine learning to do more, in less time, with fewer resources.
Here are our top three tips to show how you can turn genetic testing from cost-prohibitive to cost-efficient.
Machine learning is a critical piece of technology on the rise in labs
With the help of machine learning, pathologists can minimize subjective bias, maximize time and accuracy and improve clinical outcomes
Embrace machine learning as a tool to help you step outside the lab and impact care on a bigger level
Contributing Lab Leader
President and Head of Oncology
Human Longevity Inc.
Associate Professor of Pathology and Laboratory Medicine
University of Kentucky Medical Center
1. Speed up the analysis process
One major advantage of machine learning is its ability to reduce the amount of routine and time-consuming tasks you need to perform. Take image analysis, for example. Looking through numerous slides for disease-related features can monopolize the time you could be using on higher-value tasks.
Machines quickly narrow down areas of interest and flag what you need to look at. This not only gives you a lot of your time back, it accelerates how quickly you can provide an accurate diagnosis and empowers you to take on a larger role in the treatment process.
QUICK TIP: putting your time to better use
Machine learning can supplement your clinical workflow so you can invest more time in:
- Developing knowledge-sharing programs that will elevate other pathologists
- Building relationships with physicians to get the right test orders for the right patients
- Becoming a physician extender to meet with patients to explain their diagnosis
- Breaking down silos and becoming an integral part of a care team
2. Move from qualitative to quantitative
Too often, multiple pathologists look at the exact same image. Each may see different areas of interest. The result: a potential clash of opinions. This can leave a much-needed diagnosis hanging in the balance while an agreed-upon conclusion is being made. The good news: Machine learning may help to minimize this subjective bias.
When machines pinpoint exactly which abnormalities are important to analyze, the chances of human variables getting in the way will significantly decrease. And, as algorithms become more sophisticated, you may be able to simply review a single area of interest. This can further improve the value you and your lab provide by:
Streamlining your workflow
Decreasing test results turnaround time
Delivering more accurate results
Driving optimized care plans
Informing a preventive approach to care
Decreasing costs while increasing revenue
There’s a machine for that
Approximately 80% of 1 million prostate biopsies in the United States are benign, suggesting that prostate pathologists are spending 80% of their time looking through benign tissue.2
3. Enhancing algorithms enhances you
Machine learning in pathology is still in the early stages of how it can be used. However, innovations in software platforms are opening unique opportunities.
Today’s most forward-thinking lab leaders aren’t just relying on existing technology, they’re taking an active role in advancing it. They’re taking the role of machines beyond simple image analysis to begin studying diseases in more depth.
Here’s where to start:
- Identify software platforms: Learn where the technologies are coming from, get in touch with the development teams and share your vision.
- Improve algorithms: Lend your unique expertise to inform new, evidence-based algorithms.
- Customize the software: Whether the software be mass distributed or tailor made for your purposes, see if you can get your hands on these upgraded machine tools.
Today, you can have a role in enhancing machines’ performance, thereby enhancing their ability to support you.
New software platforms can:
- Improve analysis time
- Increase management and access to data
- Strengthen collaboration for you and your lab
- Much, much more
Redefine your perception—redefine your value
The future of machine learning is bright—but what the future holds for the role of pathologists is even brighter. Machine learning exists not to take your role, but to elevate the valuable care you already provide.
Make the case to your management team for why you need machine learning:
- Machines don’t experience burnout when reviewing hundreds of images a day
- Machines perform more routine tasks so you can focus on the human element of care
- New algorithms shorten the diagnostic latency period
- Innovative software platforms enable increased sharing of valuable data needed to make critical care decisions
With these benefits, you are free to step outside the lab and become an integral part of care teams like never before. Always remember,
“A machine will never equal you. It will not outperform a well-trained pathologist, but it can help.””
- Kenneth Bloom, M.D.
President and Head of Oncology
Human Longevity Inc.
1. Harvard and Beth Israel Deaconess Researchers Use Machine Learning Software Plus Human Intelligence to Improve Accuracy and Speed of Cancer Diagnoses. Dark Daily Web site. https://www.darkdaily.com/harvard-and-beth-israel-deaconess-researchers-use-machine-learning-software-plus-human-intelligence-to-improve-accuracy-and-speed-of-cancer-diagnoses-1116#axzz4gnRhlfgi. Accessed May 25, 2017. 2. Gurcan MN, Boucheron LE, Can A, Madabhushi A, Rajpoot NM, Yener B. Histopathological image analysis: a review. IEEE Rev Biomed Eng. 2009;2:147-171.
Deep Learning as a Tool for Increased Accuracy and Efficiency of Histopathological Diagnosis
By Geert Litjens, Clara I. Sánchez, Nadya Timofeeva, Meyke Hermsen, Iris Nagtegaal, Iringo Kovacs, Christina Hulsbergen - van de Kaa, Peter Bult, Bram van Ginneken, & Jeroen van der Laak
Discover how deep learning is helping to reduce the workflow for pathologists while increasing the accuracy of cancer diagnosis.
Image Analysis and Machine Learning in Digital Pathology: Challenges and Opportunities
By Anant Madabhushi and George Lee
Learn about the current challenges facing machine learning and some promising opportunities it may bring to pathology.
Proscia introduces first Cloud-based Software Platform for Digital Pathology
By Peeyush Shrivastava
Explore how this cloud-based software platform is doing more than just improving workflow for pathologists.