DermAI™ is a first-of-its-kind deep learning technology, designed to intelligently manage dermatopathology cases to drive efficiency, improve quality, and expand revenue-generating opportunities for labs. DermAI’s cutting-edge computational backbone assesses and pre-classifies whole slide images of skin tissue and leverages these classifications to optimize workflows through Proscia’s digital pathology platform, Concentriq™. This is a potential game-changer for dermatopathology and pathology more broadly.
How It Works
DermAI runs automatically in the background on whole slide images stored in our Concentriq digital pathology platform. Once run, it produces an output that includes a classification into one of four categories as well as confidence in this classification. These categories include: 1) Basaloid Patterns; 2) Squamous Patterns; 3) Melanocytic Patterns; and 4) Other. This fourth category captures cases that have been confidently identified as not belonging to the prior three. A fifth category exists to capture slides that do not meet the minimum confidence threshold for categories 1-4. The confidence in classification predictions allows for user-defined thresholding, enabling each implementation of DermAI to be configured to meet the requirements of each laboratory. Once generated, these outputs of DermAIare used to supercharge workflows within the Concentriq platform.
Why It Matters
Dermatopathology is one of the fastest growing specialties of pathology as well as one of the most hard-pressed to meet growing demand. This situation is being driven by the confluence of two factors: the increase in dermatology business resulting from a growing elderly population and the decreased availability of relevant diagnosticians as a large generation of dermatopathologists reaches retirement. Dermatopathology labs face a dual pressure of maintaining a high quality practice while driving efficiency in order to handle an increased caseload. Additionally, the lab’s focus on quality must be balanced with the push for efficiency resulting these trends.
DermAI helps labs strike this balance by enabling three efficiency, quality, and case-volume boosting applications that fit right into the existing processes that dermatopathology labs use to get work done: intelligent workload balancing and prioritization, automated QA and 100% AI review, and enhanced technical component reporting.
Intelligent workload balancing and case prioritization
Intelligent workload balancing and prioritization helps the lab to sort, triage, and otherwise prioritize cases. It optimizes how cases are distributed to the various dermatopathologists in a lab – by subject matter expertise, by order of cases to examine, or even to create continuity so investigators maximize their focus on high-impact cases. In turn, the lab can make the best use of specialist time, which is often one of its largest fixed costs. For example, a general pathology lab that sees a variety of case types will want to optimize a dermpath specialist’s time and ensure general pathologists are not spending time screening cases that will need to be reassigned anyways.
For labs that primarily deal with dermatopathology cases, prioritization based on DermAI outputs means that cases can be arranged to maximize effectiveness and efficiency. It can also help the lab to optimize operations by identifying cases that are most likely to need additional work (like stains or re-cuts) earlier in the day, so that orders are sent to the lab for quick turn-around. Finally, sorting cases can minimize context switching and the mental fatigue associated with it, allowing pathologists to get “in-the-zone.”
Automated QA and 100% AI re-review
This application analyzes a lab’s entire caseload, providing an AI-based interpretation for every case. Running in the background before or after pathologist review, DermAI provides an automated second layer of quality review across the lab, making existing quality programs easier, faster, and more robust.
By scanning slides after pathologist review, DermAI serves as an extra set of eyes. It can reassess all cases or a subset of cases like those that are highest priority or those with instances of discordance. DermAI could be used to institute a policy of “100% Digital Peer Reviews,” whereas most labs today hit a maximum of 5% re-reviews. For labs that choose to assess just a subset of cases, Concentriq can be configured to automatically select cases to be sent into a digital peer review/QA worklist. Targeting the highest impact cases can help to ensure that the lab is optimizing time spent on re-review while, at the same time, improving quality.
Enhanced technical component reporting
Technical component (labwork) and professional component (diagnostic work) are billable services offered by anatomic pathology labs. DermAI allows a dermatopathology lab to provide additional insight into the labwork it delivers. This added value may inform when to split up these two services and when to perform them both in-house. After slides are prepared, they are traditionally sent back directly to the dermatologist. As applied in this context, DermAI is run before sending them back, providing the additional AI assessment results that are delivered back to the originating dermatologist and, optionally, can be offered to that dermatologist either at an additional fee or as a differentiator to drive new business. In an industry that’s constantly concerned about margins due to cuts in reimbursement rates, it’s important for labs to think about ways to safeguard business vitality.
DermAI by Proscia is a first-of its kind AI application that is designed to intelligently manage dermatopathology cases to drive efficiency, improve quality, and expand revenue-generating opportunities for labs; however, this initial offering of DermAI serves as just the first touchpoint in what will be a series of AI modules that leverage intelligent software to change the way we look at cancer. From expansions within DermAI and translation to other specialties, this technology is automating and augmenting an increasing portion of the pathology workflow to drive informed decision-making, efficiency, and quality across the lab.
For an inside look at DermAI and how it could work in your lab, register for our demonstration webinar on Wednesday, June 26, 2019 at 12PM EDT.