Unstructured
Data Decoded

Unified suite of AI products to leverage unstructured clinical data at scale

Decode the context
Medical records contain gaps because they are written by clinicians for clinicians. Mendel transcends the mere words in the text to capture clinical context behind them.

Structure at Patient Level, Not Document

Mendel’s platform is the only system that can consolidate patient data across documents into a coherent patient journey.

Optical Character Recognition

Proprietary OCR built for clinical documents that captures  more correct clinical terms on the hardest-to-read pages compared to the standard generic OCR.

Classification

Mendel can classify clinical documents according to their content to catalog source documents like doctor notes and pathology reports.

We Deliver the World’s Most Advanced Clinical NLP

We started Mendel knowing using current ontologies and training existing machine learning wouldn’t get us anywhere. We invested in developing a novel approach built for medicine by training a team of physicians to think like AI scientists, and AI scientists to think like physicians.

Cancer Type
Stage
Treatment
Mendel
87.50%
75.40%
56.00%
GPT-4
84.30%
59.40%
26.70%
Spark NLP
43.84%
15.92%
21.40%
Related publications

The Mendel Difference

Learn about the technology
Patient-level understanding

Most solutions read isolated documents, like the pathology report. Mendel deciphers the full medical record, including notes, pathology reports, imaging, lab reports, and more, to create a full patient journey.

01
Clinical context

Clinicians write medical documentation for other clinicians assuming they will have clinical context. Mendel’s clinical AI goes beyond parsing basic medical vocabulary to accurately interpret and infer what’s not specifically mentioned.

02
Transparency & quality at the variable level

Generic AI measures only general system accuracy instead of providing granular visibility into each variable. This can be deceiving because some variables (e.g. medications) are much easier to capture than others (e.g. outcomes), artificially inflating the overall accuracy measurement.

03

FAQs

What is the quality of your structuring process compared to human abstraction?
For typical cohort selection criteria, Mendel can match or exceed the quality of human abstraction with up to 30x abstraction speed/efficiency.
What therapeutic areas do you cover?
We are focused on oncology and immunology. If you’re interested in other areas, let’s chat.
Can the system be deployed in my environment?
This is on our roadmap but not currently available for the full clinical NLP pipeline. We can deploy our de-identification product in your environment so we never see PHI.
Do you map the outputs against standard ontologies?
Yes, we map to standard ontologies.

Contact an Expert

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.