Synthetica Empower your dataset with synthetic images.

Choose a file (or use the camera on mobile) → generate clean up and/or synthetic variants. Advanced settings are optional.

Input

JPG / PNG

Outputs:

Selected: clean up + 1 synthetic (2 credits)

Advanced settings

Modality, subtype, augmentation strength, background

Modality

Subtype (optional)

Augmentation strength: Moderate

Mild = subtle · Strong = more variation (still realistic)

Remove photo background

Extra cleanup for clinical/derm photos. Radiology/microscopy won’t be altered.

Tip: best results with a clean, centered ROI.

Results

How to use Synthetica

Prefer YouTube? Watch on YouTube

Log in.
No sign-up needed: your email creates your account automatically.
Enter your email → Send PIN → check your inbox → enter the 4-digit PIN → Verify.

Step 1 — Login
Upload + choose settings.
Upload a JPG/PNG (convert DICOM to PNG first), then set your generation options:
- Number of images (1–4): 1 image = 1 credit = $0.50.
- Modality: use Auto-detect if unsure.
- Subtype (optional): T1/T2/FLAIR, H&E, Gram, dermoscopy, endoscopy, etc.
- Augmentation strength: Mild / Moderate / Strong (CT stays conservative).
- Remove background (toggle): for large clinical/macro images (extremities, body parts) to remove distracting context (bandages, drapes, surrounding covers).
Step 2 — Upload + options
Generate.
Click Generate and wait up to ~1 minute (depends on server load and number of images).

Step 3 — Generate
Download + manage your sessions.
Your synthetic variants appear under Results — download individual images (or ZIP if enabled).
Every run is saved as a session in your account: quick copy/share links, session page, and ZIP access.

Step 4 — Results + sessions

If you run out of quota, click Add credits in the toolbar. 1 credit = $0.50 = 1 image generated.

Tip: Synthetica removes text overlays, but you should still upload de-identified images.

About Synthetica

Synthetica turns a single medical image into dataset-ready outputs in one click: cleanup → synthetic variants → shareable session. Cleanup removes text/identifiers, annotations (arrows, measurements), and non-medical capture context (monitor frames, borders, background). Then Synthetica generates 1–4 realistic variants with modality-aware safety rules (e.g., conservative CT, no flips, no vignette).

Prefer YouTube? Watch on YouTube

Radiology: XR, Angio/DSA, CT, MRI, Ultrasound Microscopy: histology, cytology, slides Macroscopy: dermatology, endoscopy, gross images

Image generation is powered by an advanced generative AI model (currently Gemini-3-pro-image-preview) to create 1–4 plausible variants while preserving the overall visual identity of the source. 1 credit = $0.50 = 1 image generated.

Every run is stored as a session: individual image links, optional ZIP, and a session page you can open/copy/share from your account dashboard.

Examples (before → after cleanup)

Before: macroscopy with ruler/background

Before — ruler + background

After — dataset-ready (context removed)

Before — UI/labels/overlays

After — clean frame, no text

Tip: use only de-identified images (no patient names/IDs).

Synthetica is part of ML in Health Science.

Version

2.0 (09.01.2026): Added Account dashboard — manage all your sessions and generated images, with rename/remove, quick copy/share links, and session ZIP access.
1.0 (27.12.2025): Initial public release.

If something looks off, please include a screenshot + browser/device details when you report it.