Most pathologists we speak to dreamed of becoming Doctors to help patients - not to do paperwork. They also want to ensure their patients get an accurate diagnosis.
As a pathologist using Franklin.ai Digital or Glass, you're always in control. Only now, you also have an A.I. assistant, helping identify small foci of cancer and speeding up your reporting workflow.
Developed with pathologists like you.
Franklin.ai has been developed hand-in-hand with a global cohort of leading pathologists. After 100's of hours of interviews, they have told us an adjunct tool like this reduces the chances something will be missed, and increases their confidence in diagnoses.
Finding cancer
Accuracy you can rely on.
Across 1,735 slides that were reviewed by a panel of 12 subspecialists, our A.I. model achieved near perfect agreement or substantial agreement in significant clinical findings in Prostate Core Biopsies (PCB) and Transurethral Resection of the Prostate (TURP) cases.
Automation
Let Franklin.ai do the tedious tasks.
Automated quantification with grading and involvement for each specimen. Pre-populated synoptic reports designed to save you time.
A comprehensive product
Broader and deeper ontology.
Detects 45 clinical findings on prostate core biopsy and TURP chips. This includes Gleason graded malignancies, prognostic factors and benign findings to help pre-fill more synoptic reports.
An adjunct tool
You’re always in control.
Our tool is built as a second pair of eyes and an adjunct tool - we are never aiming to replace the expertise of a human pathologist. Disagree with the A.I. results? An easy-to-use 'diagreement workflow' allows you to edit the findings, and automatically have the results aggregated up into your final report.
"We have all been experiencing increased workloads since COVID. The promise of having A.I. screen my cases and prefill my reports so I can get home to cook dinner for my kids is super exciting."
Today, we're launching with Prostate tissue support. We are looking at adding multiple additional tissue types in the near future.
What kind of case types does Franklin Digital support?
Prostate CORE Biopsy and TURP chips; Radical Prostatectomy is also on our product roadmap.
If the model doesn't identify cancer, how certain are you it's not there?
Our model is very good as a negative test. For example, our negative predictive value (NPV) for Gleason Pattern 5 is 98.6%.
What ontology do you support?
We are validating against a wide range of 45+ malignant and benign findings.
Will the A.I. bias me in my clinical judgement?
As the pathologist, you’re always in control.
We’ve designed the product so you can reference the AI whenever you want.You can keep it on your second monitor so it’s outside your vision when you are performing primary diagnosis on your main monitor using a Whole Slide Image Viewer. The more you click, the more information and specifics we show you about findings.
As you interact with the tool, more information will be revealed to you, ensuring you’re in control of what you see and when.
By clicking the submit button you confirm that you have read our Privacy Policy and understand that we will collect use, and disclose your personal information in accordance with our Privacy Policy, the Privacy Act 1988 (Cth) and other Australian privacy laws and regulations, where applicable. By submitting your details you agree to receive electronic marketing messages from us about other and future products and services (including future webinars, newsletters and publications). You can unsubscribe at any time.
By clicking the submit button you confirm that you have read our Privacy Policy and understand that we will collect use, and disclose your personal information in accordance with our Privacy Policy, the Privacy Act 1988 (Cth) and other Australian privacy laws and regulations, where applicable. By submitting your details you agree to receive electronic marketing messages from us about other and future products and services (including future webinars, newsletters and publications). You can unsubscribe at any time.
