Calcium Score made portable

PyCa is a research and teaching tool developed in Python. Standalone and fully portable, it enables coronary and aortic-valve calcium score analysis.

Recommended

PyCa AI · Installer

with AI

Full AI edition, installed on your PC. Fast startup, Start-menu shortcut, no antivirus delay on each launch.

⬇ Download installer

Windows 64-bit · ~198 MB

PyCa AI · Portable

with AI

Full AI edition in a single file, no installation. Run it from a USB stick; no administrator rights required.

⬇ Download portable

Windows 64-bit · ~293 MB

PyCa Lite

no AI

Lightweight edition, manual tools only (Click&Grow, brush). Maximally compatible, smaller file.

⬇ Download Lite

Windows 64-bit · ~178 MB

PyCa requires a free research license, to be activated on first launch.

🔑 Request license 📖 Guide 💬 Feedback

⚠ Research use only · Not validated for clinical use

The coronary calcium score quantifies calcium deposits in the coronary arteries through the Agatston methodology. As confirmed by sub-analyses of the SCOT-HEART trial, a score of zero indicates an excellent long-term prognosis, dramatically reducing the need for invasive examinations.

The aortic-valve calcium score, in a similar fashion, quantifies mechanical calcification of the valve leaflets — a parameter recommended by ESC / AHA guidelines to confirm the clinical severity of aortic stenosis when echocardiographic data are equivocal.

PyCa was built to learn calcium score calculation in a local environment — no dedicated workstation, no cloud, no installation required, even from a small USB stick.

PyCa ships with a small SQLite database to store results and bulk-export them for research analyses.

What PyCa does

Coronary AI analysis

PyCa embeds the SEGMENT-CACS model (Föllmer 2024) converted to ONNX, with GPU support via DirectML, to perform AI-based coronary calcium score analysis.

Slicer-style oblique MPR

Three viewports with 2D resampling in an MPR-like environment; the aortic valve can be visualised on a paraxial plane to better discriminate extra-valvular calcifications.

MESA Data

PyCa ships with the MESA tables (McClelland et al. 2006) to derive a coronary percentile by age · sex · ethnicity.

AVC Cusp Heatmap

Angular distribution of aortic-valve calcium across the 3 leaflets (RCC / LCC / NCC) with ESC/EACTS 2021 thresholds for equivocal stenosis quantification.

Segmentation tools

Click & Grow, brush, eraser, undo/redo, threshold preview. Le maschere di segmentazione possono essere salvati e ricaricate in seguito

PDF report

PDF report che include tutti i dati quantitativi, con score nelle aree di segmentazione e percentili

✨ NEW IN PYCA

Artificial intelligence for calcium scoring

PyCa integrates the multi-task SEGMENT-CACS model (Föllmer 2024, Apache-2.0) to automatically segment LM, LAD, CX and RCA. Accelerated inference on any Windows GPU via DirectML (NVIDIA / AMD / Intel) with automatic CPU fallback. Typical time: ~30 seconds on GPU.

⚠ Experimental feature. Always verify the output and refine manually. PyCa is not a medical device; the AI does not replace the radiologist’s judgement.

Citation

Use of PyCa in scientific publications is free. If you find the results useful, a citation is appreciated:

Suggested citation

Censullo V. — AITeRTC ETS. PyCa — Python Calcium Score Analyzer. 2026.

PyCa · developed by Vittorio Censullo — AITeRTC Scientific Committee
Co-author: AITeRTC ETS · www.aitertc.it · v.censullo@gmail.com

Guide·Request license·Download with AI·Download without AI