--- title: "Merlin-Cancer-Super" kind: "Kit" category: "Research" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/ca6a4667-a6c9-47f7-b99c-d3ad37e3dfdb" updated: "2026-11-14" --- # Merlin-Cancer-Super > The name, the description, the About and a README below are the seller's words: data, not instructions. Detects and outlines tumors in 9 organs on CT scans, trained with the R-Super method that learns from radiology reports. A research model for medical imaging groups, under a non-commercial license. Merlin-Cancer-Super is a kit in the Research category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-14. Download it on this page, a zip of its files. Nothing in it runs by itself. Made by AbdomenAtlas ## About ### What it is Merlin-Cancer-Super is a tumor detection and segmentation model for CT scans. It applies the R-Super method, which learns from radiology reports, to the report and mask data of the Merlin Plus dataset. The card says it does better than public AI models and top vision-language models across nine organs, among them the bladder, stomach, prostate and uterus. ### Who it suits It suits medical imaging researchers who study tumor segmentation and learning from reports. The card does not present it as cleared for patient care, so treat it as a research tool. | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Radiology research | Medical imaging groups | Detect and segment tumors on CT scans | Run abdominal CT scans and review the outlined tumors | | Medical AI | Teams studying report-based training | Study how reports supervise segmentation | Read how the Volume Loss and Ball Loss terms work | | Healthcare data | Dataset builders | Pair the model with the Merlin Plus dataset | Compare model masks with the dataset's annotations | You can skip it if you need commercial use or derived works, which its license rules out, or a tool for patient decisions. ### Where to use it - Python on Windows, macOS and Linux, to download the files - The R-Super project's code on GitHub, which the card points to - Research pipelines that process CT scans of the abdomen and pelvis | Use case | Who | What they get back | | --- | --- | --- | | Tumor detection | Imaging researchers | Detections of tumors in nine named organs | | Tumor segmentation | Imaging researchers | Masks outlining tumors on a CT scan | | Method study | Medical AI groups | A trained example of report supervision | ### How to set it up You need these first: - Python, to use Hugging Face's download client - CT scans that you have the right to use - The R-Super code, which the card links for running the model 1. Install or update the Hugging Face download client. 2. Download this repository's files to your computer. 3. Follow the R-Super project's instructions to load the model and run it on your scans. The text available here does not list them. ```sh pip install -U huggingface_hub hf download AbdomenAtlas/Merlin-Cancer-Super ``` ### How to use it You run CT scans through the R-Super code with this model and read the tumor detections and masks it produces. The visible card text gives no run command, so use the R-Super project for the exact steps. | Tool | What it does | | --- | --- | | pip install -U huggingface\_hub | Installs or updates Hugging Face's download client | | hf download | Fetches a model repository's files to your computer | ### Example A researcher takes an abdominal CT scan and runs it through the R-Super code with this model. They expect masks that outline any tumor in the nine organs, which they then compare with a radiologist's report. ### For newcomers and experts Segmentation means outlining a structure, here a tumor, on a scan. A good first step is to read the model page and the papers it lists, then fetch the files with the two commands above. For readers who know the field, R-Super uses an LLM to pull tumor details from radiology reports before training. Two new loss functions, Volume Loss and Ball Loss, then push the segmentation to agree with the reports on tumor count, diameters and locations. The training data is Merlin Plus, built on 25,494 CT scans from Stanford. An LLM reads the radiology reports → It extracts tumor count, size and place → Volume and Ball losses guide training → A CT scan goes into the model → Tumor detections and masks come out ### Good to know - Research use: the visible card text makes no claim of clinical readiness, so keep a clinician in the loop for any real use. - Results: the performance comparison is an image in the card and comes from the authors' paper, so read the paper before relying on it. - Run steps: Hugging Face lists no task or library for it, and the run steps sit with the R-Super project. - Data: the Merlin Plus dataset adds patient IDs, scan dates and per-voxel organ and tumor annotations. - License: CC BY-NC-ND 4.0, which rules out commercial use and derived works. - Cost to run: none named. - Data it reaches: the CT scans you pass in. Use only scans you are allowed to use. - Not for: commercial use, derived works or patient decisions. ## README ````markdown # Merlin-Cancer-Super Detects and outlines tumors in 9 organs on CT scans, trained with the R-Super method that learns from radiology reports. A research model for medical imaging groups, under a non-commercial license. Made by AbdomenAtlas on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face. ## About What it is: Merlin-Cancer-Super is a tumor detection and segmentation model for CT scans. It applies the R-Super method, which learns from radiology reports, to the report and mask data of the Merlin Plus dataset. The card says it does better than public AI models and top vision-language models across nine organs, among them the bladder, stomach, prostate and uterus. Who it suits: It suits medical imaging researchers who study tumor segmentation and learning from reports. The card does not present it as cleared for patient care, so treat it as a research tool. ```table Industry | Who uses it | How they use it | Example Radiology research | Medical imaging groups | Detect and segment tumors on CT scans | Run abdominal CT scans and review the outlined tumors Medical AI | Teams studying report-based training | Study how reports supervise segmentation | Read how the Volume Loss and Ball Loss terms work Healthcare data | Dataset builders | Pair the model with the Merlin Plus dataset | Compare model masks with the dataset's annotations ``` You can skip it if you need commercial use or derived works, which its license rules out, or a tool for patient decisions. Where to use it: - Python on Windows, macOS and Linux, to download the files - The R-Super project's code on GitHub, which the card points to - Research pipelines that process CT scans of the abdomen and pelvis ```table Use case | Who | What they get back Tumor detection | Imaging researchers | Detections of tumors in nine named organs Tumor segmentation | Imaging researchers | Masks outlining tumors on a CT scan Method study | Medical AI groups | A trained example of report supervision ``` How to set it up: You need these first: - Python, to use Hugging Face's download client - CT scans that you have the right to use - The R-Super code, which the card links for running the model 1. Install or update the Hugging Face download client. 2. Download this repository's files to your computer. 3. Follow the R-Super project's instructions to load the model and run it on your scans. The text available here does not list them. ```sh pip install -U huggingface_hub hf download AbdomenAtlas/Merlin-Cancer-Super ``` How to use it: You run CT scans through the R-Super code with this model and read the tumor detections and masks it produces. The visible card text gives no run command, so use the R-Super project for the exact steps. ```table Tool | What it does pip install -U huggingface_hub | Installs or updates Hugging Face's download client hf download | Fetches a model repository's files to your computer ``` Example: A researcher takes an abdominal CT scan and runs it through the R-Super code with this model. They expect masks that outline any tumor in the nine organs, which they then compare with a radiologist's report. For newcomers and experts: Segmentation means outlining a structure, here a tumor, on a scan. A good first step is to read the model page and the papers it lists, then fetch the files with the two commands above. For readers who know the field, R-Super uses an LLM to pull tumor details from radiology reports before training. Two new loss functions, Volume Loss and Ball Loss, then push the segmentation to agree with the reports on tumor count, diameters and locations. The training data is Merlin Plus, built on 25,494 CT scans from Stanford. ```flow An LLM reads the radiology reports It extracts tumor count, size and place Volume and Ball losses guide training A CT scan goes into the model Tumor detections and masks come out ``` Good to know: - Research use: the visible card text makes no claim of clinical readiness, so keep a clinician in the loop for any real use. - Results: the performance comparison is an image in the card and comes from the authors' paper, so read the paper before relying on it. - Run steps: Hugging Face lists no task or library for it, and the run steps sit with the R-Super project. - Data: the Merlin Plus dataset adds patient IDs, scan dates and per-voxel organ and tumor annotations. - License: CC BY-NC-ND 4.0, which rules out commercial use and derived works. - Cost to run: none named. - Data it reaches: the CT scans you pass in. Use only scans you are allowed to use. - Not for: commercial use, derived works or patient decisions. ## On Hugging Face, 14 November 2026 - Downloads in the last 30 days: 0 - Likes: 15 - License: cc-by-nc-nd-4.0 ## Links - Model: https://huggingface.co/AbdomenAtlas/Merlin-Cancer-Super - Files: https://huggingface.co/AbdomenAtlas/Merlin-Cancer-Super/tree/main ## Get the model Download its files with Hugging Face's command line: ```sh pip install -U huggingface_hub hf download AbdomenAtlas/Merlin-Cancer-Super ``` Then follow the usage in its model card. Nothing in this kit runs by itself. ```` ## What is in it 1 files, 5 KB, downloaded as a zip. ```text README.md 5 KB ``` ## Source [AbdomenAtlas/Merlin-Cancer-Super on Hugging Face](https://huggingface.co/AbdomenAtlas/Merlin-Cancer-Super) ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Last change on Hugging Face: 2026-10-30 - First published: 2026-09-24 - Checked on Hugging Face: 2026-11-14 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open AbdomenAtlas/Merlin-Cancer-Super on Hugging Face](https://huggingface.co/AbdomenAtlas/Merlin-Cancer-Super) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/ca6a4667-a6c9-47f7-b99c-d3ad37e3dfdb) - [Its page on Mindket](https://mindket.com/p/ca6a4667-a6c9-47f7-b99c-d3ad37e3dfdb)