PTOLEMI
99.72% Time Savings
PTOLEMI (Python Tensorflow Object Locator Examining Microfluidic Instruments) is a machine learning system that automates cancer cell analysis in microfluidic assays. Developed in collaboration with Dr. Doron Gerber at Bar-Ilan University, the system processes what would take a technician 3 work weeks in under 18 minutes—a 360-fold improvement in analysis time.
- Time Savings: 99.72%
- Speed Increase: 360x
- Images/Assay: 6,400
- Analysis Time: <18 min
Key Features
ML Image Analysis
Cutting-edge AI algorithms rapidly discriminate between cell viability and distinct cell types within biopsy samples, automating traditionally manual microscopy analysis.
360x Faster
A single assay produces 6,400 images requiring analysis. Manual review takes ~107 hours; PTOLEMI completes it in under 18 minutes.
Published Research
Peer-reviewed paper published in Diagnostics journal (2023) documenting methodology, training process, and validation results.
The Research
PTOLEMI was developed to address a critical bottleneck in personalized cancer treatment: the manual analysis of microfluidic assay images. Each patient's assay generates 256 cultivation chamber images across 25 time points—6,400 images total that previously required nearly 3 weeks of technician time to analyze.
Working under Professor Doron Gerber at Bar-Ilan University, I developed an AI system that could classify different cancer cell types based on microscope images. The system leverages cutting-edge machine learning to process large datasets rapidly and with high accuracy.
The result is a diagnostic process that's automated, swift, precise, and resource-efficient—well-suited for point-of-care requirements where timely treatment decisions can significantly impact patient outcomes.