Laboratory of Advanced Technology and Intelligent Systems (LATIS)  — Sousse University · AI in Medical Imaging
Research → Industry · Medical AI

From the research lab to production-ready medical AI.

PTM bridges academic research and industry. We industrialize published mammography algorithms into a deployable medical SaaS, a standardized Python processing stack and developer tools for DICOM — engineered for reliability, integration and scale.

See how it works → Launch Medical SaaS
SaaSDeployed web platform
4Published methods in production
PyPIOpen-source library
DICOMNative workflows
Core Technology

Detecting microcalcifications — the earliest sign of breast cancer.

Microcalcifications are tiny calcium deposits, often well under a millimetre, with low contrast against dense breast tissue. Our detection engine combines proven image-analysis methods with attention-guided deep learning, validated on public benchmarks and packaged for use in real imaging workflows.

Mathematical morphologyMulti-scale structuring elements, Top-Hat and structural similarity maps.
Superpixel analysisLocal, adaptive thresholding over homogeneous regions.
Collaborative classificationGraph-based knowledge propagation across detectors.
Deep learningCNN segmentation guided by image-processing priors.
Attention & transformersSteering models toward clinically relevant regions.
Benchmarked performanceStandard metrics on public datasets, tracked across every model release.
From Lab to Product

From a mammogram region of interest to a segmentation map.

Every published method is industrialized into the same production pipeline that powers the medical SaaS.

  1. DICOM ingestion & standardization
  2. Image enhancement & computational priors
  3. AI detection & segmentation
  4. Quantitative evaluation
Mammogram region of interest
Region of interest
AI microcalcification segmentation output
AI segmentation output
The Product

One medical SaaS for microcalcification detection.

Medical Saas brings every PTM component — viewers, AI models, inference and study management — into a single web platform.

Flagship Web platform DICOM-native

Medical Saas

An end-to-end SaaS for AI-assisted microcalcification detection in mammography. Upload DICOM studies, review them in a dedicated mammography viewer, run the detection models, and manage studies and results from one ERP-style workspace — with full technical documentation for integration and evaluation.

Launch Medical SaaS → Read the documentation
mcd.ptm.tn · mcddocs.ptm.tn

What's inside

DCM
DICOM processingIngestion, standardization and preprocessing of mammograms.
VIEW
Mammography viewerPurpose-built review of images and AI overlays.
AI
Models & inferenceClassical and deep-learning detectors behind one service.
ERP
Study & workflow managementPatients, studies, results and users in one place.
Open Source

Open, production-grade building blocks.

The components behind our SaaS are released openly, so research teams and companies can integrate them directly into their own medical imaging products.

PY

medical-image-std

Python library · open source

Standardized loading, preprocessing and processing of medical images for research and production workflows — the same foundation running inside the medical SaaS.

pip install medical-image-std
GitHub → PyPI →
DOC

Technical documentation

Architecture · models · usage

Reference documentation for the Medical Saas platform: how the pipeline is built, which models it runs, and how to integrate it — written for engineering teams and partners evaluating the system.

mcddocs.ptm.tn →
IDE

DICOM Viewer for JetBrains IDEs

JetBrains Marketplace plugin

Open and inspect DICOM medical images directly inside PyCharm, IntelliJ IDEA and other JetBrains IDEs — no context switching while building medical imaging code.

Get the plugin →
Collaborate with PTM

Build the next stage with us.

We are looking for academic, clinical and industry partners to validate, scale and deploy medical imaging AI.

Research collaborators

Universities, hospitals, radiologists
  • Joint R&D projects and studies
  • Clinical validation of detection models
  • Annotated mammography datasets

Infrastructure sponsors

Cloud, GPU and compute providers
  • GPU compute for training and benchmarking
  • Production hosting for the medical SaaS
  • Visible credit across the platform and releases

Open-source ecosystems

Medical imaging frameworks and communities
  • Integration of our models and tools
  • Shared, reproducible benchmarks
  • Co-maintained open tooling
Start a collaboration →