Asset Type: Posters, Imaging Core Lab, Additional  Therapeutic Areas

Automated Multi-Anatomy Lesion Segmentation and Classification in PSMA-PET Imaging

This poster, presented at SNMMI 2026, presents an AI-driven workflow for automated lesion segmentation and classification in PSMA-PET imaging. Using a deep learning model trained on more than 1,600 expertly segmented scans, the approach automatically identifies, delineates, and classifies lesions by anatomy, achieving strong accuracy while reducing reader time by more than 50%.

Why download this poster?

Learn how AI can automate lesion segmentation and classification in PSMA-PET imaging workflows.
Explore a machine learning approach trained on 1,630 expertly segmented images.
Review performance results demonstrating high accuracy (Dice score 0.88, Precision 0.91).
Discover how automated processing can reduce image review time by more than 50% while improving reproducibility.
Gain insight into emerging capabilities for lesion-level quantification and longitudinal lesion tracking in prostate cancer trials.

Want the Full Poster?

This field is for validation purposes and should be left unchanged.

First Name(Required)

Last Name(Required)

Business Email(Required)

Phone Number

Institute or Company(Required)

Service Line Interest(Required)

Therapeutic Area(Required)

I would like to subscribe to Perceptive marketing communications. By subscribing, you consent to receive emails from Perceptive. We respect your privacy, and you can manage your communication preferences anytime. You can find the Perceptive privacy policy.

Automated Multi-Anatomy Lesion Segmentation and Classification in PSMA-PET Imaging