Static MEDTRACE workstation: export, recorded responses and assets (part 5)

README.md

@@ -1,12 +1,1321 @@
----
-title: Medtrace
-emoji: 📊
-colorFrom: pink
-colorTo: purple
-sdk: static
-pinned: false
-license: mit
-short_description: Brain tumor analysis & 3D evolution.
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+---
+title: MEDTRACE
+colorFrom: indigo
+colorTo: blue
+sdk: static
+app_file: index.html
+pinned: true
+license: mit
+short_description: Longitudinal AI for brain-MRI disease evolution
+tags:
+  - medical-imaging
+  - mri
+  - brain-tumour
+  - segmentation
+  - longitudinal
+  - cornerstone3d
+  - vtk
+---
+
+<div align="center">
+
+<img src="GIF/Brain.gif" alt="MEDTRACE: longitudinal brain MRI analysis" width="460" />
+
+# MEDTRACE
+
+### Longitudinal AI for brain-MRI disease evolution
+
+**An interactive, evidence-linked map of how a brain tumour changes over time.**
+
+<br />
+
+![Stage](https://img.shields.io/badge/build-stages%200--5%20complete-0f766e?style=for-the-badge)
+![Tests](https://img.shields.io/badge/checks-345%20unit%20%2B%20351%20browser-155e75?style=for-the-badge)
+![Dice](https://img.shields.io/badge/test%20Dice-0.894%20mean-1e40af?style=for-the-badge)
+![Patients](https://img.shields.io/badge/real%20patients-209-4c1d95?style=for-the-badge)
+
+![Python](https://img.shields.io/badge/Python-3.11%2B-3776AB?logo=python&logoColor=white)
+![FastAPI](https://img.shields.io/badge/FastAPI-modular%20monolith-009688?logo=fastapi&logoColor=white)
+![PyTorch](https://img.shields.io/badge/PyTorch-2.10-EE4C2C?logo=pytorch&logoColor=white)
+![MONAI](https://img.shields.io/badge/MONAI-1.6-00A0B0)
+![Next.js](https://img.shields.io/badge/Next.js-15.5-000000?logo=nextdotjs&logoColor=white)
+![React](https://img.shields.io/badge/React-19.1-61DAFB?logo=react&logoColor=black)
+![Cornerstone3D](https://img.shields.io/badge/Cornerstone3D-5.7-0ea5e9)
+![vtk.js](https://img.shields.io/badge/vtk.js-36.4-6366f1)
+![PostgreSQL](https://img.shields.io/badge/PostgreSQL-17-4169E1?logo=postgresql&logoColor=white)
+![Docker](https://img.shields.io/badge/Docker%20Compose-5%20services-2496ED?logo=docker&logoColor=white)
+
+</div>
+
+> [!WARNING]
+> **Research prototype. Not a medical device.**
+> MEDTRACE is not intended for diagnosis, treatment planning, or any clinical decision. It has
+> not been clinically validated. It reports **measured change only**, never a diagnosis, grade,
+> progression judgement, treatment recommendation, or prognosis. All clinical decisions remain
+> with qualified healthcare professionals. Built exclusively on public, de-identified research
+> data.
+
+
+> [!NOTE]
+> **What this hosted demo serves.** 28 glioblastoma patients from
+> [RHUH-GBM](https://doi.org/10.7937/4545-c905), 3 timepoints each, under CC BY 4.0. The figures
+> below describe the full local build across LUMIERE and BraTS 2023, which are covered by data use
+> agreements that grant use but not redistribution, so they are not published here.
+>
+> There is no server. This is a Static Space: the interface and its recorded responses are served
+> from this repository, imaging is range-fetched from the
+> [dataset repository](https://huggingface.co/datasets/AIOmarRehan/medtrace-rhuh-gbm-derived), and
+> both renderers run on your GPU. Nothing is computed on request.
+>
+> [Model](https://huggingface.co/AIOmarRehan/medtrace-brats-segresnet) ·
+> [Dataset](https://huggingface.co/datasets/AIOmarRehan/medtrace-rhuh-gbm-derived) ·
+> [Code](https://github.com/AIOmarRehan/medtrace)
+
+---
+
+<div align="center">
+
+<table>
+<tr>
+<td align="center"><b>209</b><br />real patients</td>
+<td align="center"><b>874</b><br />MRI studies</td>
+<td align="center"><b>3,431</b><br />image series</td>
+<td align="center"><b>1,153</b><br />disease observations</td>
+</tr>
+<tr>
+<td align="center"><b>772</b><br />tracked lesions</td>
+<td align="center"><b>6,499</b><br />measurements</td>
+<td align="center"><b>599</b><br />atlas volumes</td>
+<td align="center"><b>59.9</b><br />3D fps measured</td>
+</tr>
+</table>
+
+</div>
+
+---
+
+## Contents
+
+<table>
+<tr>
+<td valign="top" width="33%">
+
+**Understanding it**
+- [The clinical problem](#the-clinical-problem)
+- [What MEDTRACE answers](#what-medtrace-answers)
+- [The workstation](#the-workstation)
+- [Analysis pipeline](#analysis-pipeline)
+
+</td>
+<td valign="top" width="33%">
+
+**How it works**
+- [Domain model](#domain-model-observations-not-images)
+- [AI segmentation](#ai-segmentation)
+- [Training notebook](#the-training-notebook)
+- [Validation & results](#validation--results)
+- [Measurement & change](#measurement-matching-and-change)
+- [3D disease evolution](#3d-disease-evolution)
+- [Evidence engine](#the-evidence-engine)
+
+</td>
+<td valign="top" width="33%">
+
+**Running it**
+- [Architecture](#architecture)
+- [Technology stack](#technology-stack)
+- [Datasets](#datasets)
+- [Verification](#verification)
+- [Getting started](#getting-started)
+
+</td>
+</tr>
+</table>
+
+---
+
+## The clinical problem
+
+A patient with a brain tumour is imaged repeatedly: before surgery, after surgery, during
+radiotherapy and chemotherapy, then at follow-up for years. **Clinical decisions are made by
+comparing these examinations, not by reading any one of them.**
+
+Today that comparison is largely manual:
+
+```
+radiologist opens prior study ──► reads previous report ──► scrolls both studies side by side
+       ──► re-measures the lesion by hand ──► mentally reconstructs the patient's history
+```
+
+Most medical-imaging AI does not help, because it analyses **one scan at one moment**.
+Longitudinal comparison is a recognised underdeveloped area of medical imaging AI, and it is
+exactly where the clinical decision actually happens.
+
+MEDTRACE exists to answer one question:
+
+<div align="center">
+
+### *What changed in this patient between examinations?*
+
+</div>
+
+---
+
+## What MEDTRACE answers
+
+Every screen exists to serve one of five questions. Nothing else is in scope.
+
+| Question | How MEDTRACE answers it |
+|---|---|
+| **What changed?** | Quantified, with units and direction: volume, diameter, surface area, growth rate |
+| **Where did it change?** | Highlighted in the image, per lesion, in 2D and 3D |
+| **When did it change?** | Located on the disease timeline, with the interval in days |
+| **How confident are we?** | Per pipeline stage, with the reason in plain language |
+| **Why do you believe that?** | The evidence, one click away |
+
+**The signature interaction is the time scrubber.** Dragging it moves imaging, segmentation,
+measurements, clinical events, the 3D surface and the evidence panel together.
+
+> [!NOTE]
+> An observation without evidence **cannot exist** in MEDTRACE. This is enforced in the service
+> layer, not by convention. It is why the system can always answer "why".
+
+---
+
+## The workstation
+
+<div align="center">
+
+<img src="docs/screenshots/01-workstation.png" alt="MEDTRACE clinical workstation" width="100%" />
+
+*The clinical workstation: patient timeline, synchronised prior/current comparison, measured
+findings, and the AI evidence strip.*
+
+</div>
+
+<table>
+<tr>
+<td width="50%" valign="top">
+
+### 2D comparison
+
+<img src="docs/screenshots/02-2d-comparison.png" alt="Synchronised prior and current study" width="100%" />
+
+Prior and current study side by side, **linked by fractional depth** rather than world
+coordinates, because unregistered studies differ by tens of millimetres and copying a camera
+between them can place it outside the other volume entirely.
+
+<img src="GIF/gif-04-slice-sync.gif" alt="Scrolling one pane moves the other with it" width="100%" />
+
+*Scroll one pane and the other follows.*
+
+</td>
+<td width="50%" valign="top">
+
+### Segmentation overlay
+
+<img src="docs/screenshots/03-segmentation-overlay.png" alt="Tumour mask overlaid on native image" width="100%" />
+
+The tumour mask drawn in the **series' own voxel grid**, reoriented from the NIfTI affines with
+no resampling. Outline by default, fill on demand.
+
+</td>
+</tr>
+</table>
+
+---
+
+## Analysis pipeline
+
+Deliberately a pipeline of specialised stages, not one large model. **Every stage emits its
+prediction, its confidence, its quality flags and its model version.**
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","textColor":"#1e293b","nodeTextColor":"#1e293b","lineColor":"#475569","edgeLabelBackground":"#ffffff"},"flowchart":{"nodeSpacing":34,"rankSpacing":50,"padding":10}}}%%
+flowchart LR
+    A[("MRI studies")] --> B["Quality control"]
+    B --> C["Segmentation"]
+    C --> D["Lesion detection"]
+    D --> E["Registration"]
+    E --> F["Lesion matching"]
+    F --> G["Change detection"]
+    G --> H[("DiseaseObservation")]
+    H --> I["Disease timeline"]
+    H --> J["3D evolution map"]
+    H --> K["Evidence engine"]
+
+    classDef src fill:#e0f2fe,stroke:#0284c7,stroke-width:1px,color:#0c4a6e
+    classDef stage fill:#ffffff,stroke:#64748b,stroke-width:1px,color:#1e293b
+    classDef core fill:#ccfbf1,stroke:#0d9488,stroke-width:2px,color:#134e4a
+    classDef view fill:#e0e7ff,stroke:#4f46e5,stroke-width:1px,color:#312e81
+
+    class A src
+    class B,C,D,E,F,G stage
+    class H core
+    class I,J,K view
+```
+
+<details>
+<summary><b>What each stage actually does</b></summary>
+
+<br />
+
+| Stage | Implementation | Output |
+|---|---|---|
+| **Quality control** | Rule-based validation of sequences, voxel spacing, orientation, protocol drift between timepoints | `QualityFlag[]` per study |
+| **Segmentation** | MONAI `SegResNet`, 18.8 M parameters, 4 sequences in → 3 overlapping compartments out | Tumour compartment masks |
+| **Lesion detection** | Connected-component extraction per compartment, with a measurability floor | `Lesion` candidates |
+| **Registration** | SimpleITK rigid then affine; atlas-space masks need none, native-space ones do | Transform + score |
+| **Lesion matching** | Similarity scoring over overlap (Dice), centroid distance, proximity and volume similarity | `Lesion` identity across time |
+| **Change detection** | Volumetric and morphological comparison with uncertainty propagation | Absolute + relative change, growth rate |
+| **Trajectory** | Longitudinal feature vectors per lesion across all timepoints | Per-lesion history |
+| **Evidence engine** | Structured findings → validated answer → linked evidence | `AIObservation` + `EvidenceItem[]` |
+
+</details>
+
+---
+
+## Domain model: observations, not images
+
+The core domain object is **`DiseaseObservation`**, not `MRI`. The timeline, the analytics, the
+AI answers and the audit trail are all *views over observations*.
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","textColor":"#1e293b","nodeTextColor":"#1e293b","lineColor":"#475569","edgeLabelBackground":"#ffffff"}}}%%
+erDiagram
+    PATIENT ||--o{ STUDY : "timepoints"
+    PATIENT ||--o{ LESION : "identity"
+    PATIENT ||--o{ CLINICAL_EVENT : "treatment"
+    PATIENT ||--o{ AI_OBSERVATION : "answers"
+    STUDY ||--o{ SERIES : "sequences"
+    STUDY ||--o| STUDY_QUALITY : "validation"
+    STUDY ||--o{ SEGMENTATION : "masks"
+    STUDY ||--o{ LESION_OBSERVATION : "observed in"
+    LESION ||--o{ LESION_OBSERVATION : "observed at"
+    LESION_OBSERVATION ||--o{ MEASUREMENT : "quantified by"
+    LESION_OBSERVATION ||--o{ EVIDENCE_ITEM : "supports"
+    AI_OBSERVATION ||--o{ EVIDENCE_ITEM : "must cite"
+    MODEL_RUN ||--o{ LESION_OBSERVATION : "produced"
+    MODEL_RUN ||--o{ AI_OBSERVATION : "produced"
+```
+
+A **`Lesion` belongs to a patient, not a study**, and that is what makes a per-lesion trajectory
+possible, and it is precisely what a conventional viewer or segmentation tool does not provide.
+
+<details>
+<summary><b>The 14-table schema</b></summary>
+
+<br />
+
+`patients` · `studies` · `series` · `study_quality` · `registrations` · `segmentations` ·
+`lesions` · `lesion_observations` · `measurements` · `clinical_events` · `model_runs` ·
+`ai_observations` · `evidence_items` · `audit_events`
+
+Managed with SQLAlchemy + Alembic. PostgreSQL stores **metadata only**. Imaging stays on the
+filesystem and is served through a path-allowlisted endpoint, never by a database blob.
+
+</details>
+
+---
+
+## AI segmentation
+
+A 3D `SegResNet` trained from scratch on BraTS 2023 GLI, on a Kaggle Tesla T4.
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","clusterBkg":"#f1f5f9","clusterBorder":"#94a3b8","textColor":"#1e293b","nodeTextColor":"#1e293b","lineColor":"#475569","edgeLabelBackground":"#ffffff"},"flowchart":{"nodeSpacing":34,"rankSpacing":58,"padding":10}}}%%
+flowchart LR
+    subgraph IN ["4 co-registered sequences"]
+        direction TB
+        A1["T1c"]
+        A2["T1n"]
+        A3["T2-FLAIR"]
+        A4["T2w"]
+    end
+
+    B["Normalise"]
+    C["Crop"]
+    D["SegResNet 3D"]
+    E["Sliding window"]
+
+    subgraph OUT ["3 overlapping compartments"]
+        direction TB
+        F1["TC"]
+        F2["WT"]
+        F3["ET"]
+    end
+
+    G["Threshold"]
+    H["BraTS labels"]
+
+    A1 --> B
+    A2 --> B
+    A3 --> B
+    A4 --> B
+    B --> C --> D --> E
+    E --> F1
+    E --> F2
+    E --> F3
+    F1 --> G
+    F2 --> G
+    F3 --> G
+    G --> H
+
+    classDef seq fill:#e0f2fe,stroke:#0284c7,stroke-width:1px,color:#0c4a6e
+    classDef step fill:#ffffff,stroke:#64748b,stroke-width:1px,color:#1e293b
+    classDef model fill:#ccfbf1,stroke:#0d9488,stroke-width:2px,color:#134e4a
+    classDef out fill:#ede9fe,stroke:#7c3aed,stroke-width:1px,color:#4c1d95
+
+    class A1,A2,A3,A4 seq
+    class B,C,E,G,H step
+    class D model
+    class F1,F2,F3 out
+```
+
+| Step | What happens |
+|---|---|
+| **T1c · T1n · T2-FLAIR · T2w** | Four sequences, in **this exact channel order** |
+| **Normalise** | Per case, per channel, zero mean unit variance over **non-zero voxels only** |
+| **Crop** | To the non-zero bounding box of the summed channels, 4-voxel margin |
+| **SegResNet 3D** | MONAI `SegResNet`, 18,798,627 parameters, 32 init filters, blocks down `[1,2,2,4]` |
+| **Sliding window** | 128³ patches, 0.5 overlap, gaussian blending |
+| **TC · WT · ET** | Independent sigmoid per channel, so the three compartments **overlap** rather than compete |
+| **Threshold** | 0.5, plus an enhancing-tumour floor of 200 voxels |
+| **BraTS labels** | Written WT→2, then TC→1, then ET→3, in that order |
+
+> [!IMPORTANT]
+> **Channel order is not recoverable from the weights.** `[t1c, t1n, t2f, t2w]` is part of the
+> model contract, recorded in `ml/artifacts/model_card.json`. Wrong order → wrong output → *no
+> error*. This is the kind of silent failure the model card exists to prevent.
+
+### Held-out test performance
+
+**186 cases from 169 subjects never seen in training or tuning.** Splits are computed at
+**subject level**, because BraTS contains 1,251 cases from only 1,133 subjects, so a random split over
+cases would leak the same patient into train and test.
+
+| Compartment | Dice (mean) | Dice (median) | HD95 (median) | Sensitivity | Precision | Dice (mean) on a 0 to 1 scale |
+|---|---:|---:|---:|---:|---:|:---|
+| **Whole tumour** | **0.9216** | 0.9489 | 2.45 mm | 0.9248 | 0.9236 | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱ |
+| **Tumour core** | **0.9078** | 0.9563 | 2.00 mm | 0.9166 | 0.9170 | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱ |
+| **Enhancing tumour** | **0.8520** | 0.8984 | 1.41 mm | 0.8833 | 0.8487 | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱▱ |
+| **Mean of the three** | **0.8938** | - | - | - | - | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱ |
+
+> [!TIP]
+> **Test scored higher than validation** (0.8938 vs 0.8899 mean Dice). Since both epoch
+> selection and post-processing tuning used the validation split, the validation figures are
+> optimistic by construction, so the test figures are the honest ones, and they did not degrade.
+
+<details>
+<summary><b>Training configuration and honest limitations</b></summary>
+
+<br />
+
+| Setting | Value |
+|---|---|
+| Architecture | `monai.networks.nets.SegResNet`, 3D, 32 init filters, blocks down `[1,2,2,4]` |
+| Loss | `DiceFocalLoss(sigmoid=True, squared_pred=True, batch=True)` |
+| Optimiser | AdamW, lr 2e-4, wd 1e-5, `CosineAnnealingLR` |
+| Precision | AMP float16 on Tesla T4 |
+| Patch sampling | 128³, 80% centred on whole tumour, 20% uniform |
+| Augmentation | Random axis flips, intensity scale ±10%, intensity shift ±10% |
+| Epochs | 37 of 60 completed (host RAM exhausted); **epoch 32 selected** on validation mean Dice |
+| Seed | `20260813` for both split and training |
+
+**Stated limitations** (from the model card):
+
+- Trained on **pre-operative** adult glioma only. Post-treatment appearances, such as resection
+  cavities, radiation change, are not represented.
+- Requires all four sequences. Behaviour with a missing sequence is untested.
+- Assumes BraTS preprocessing: skull-stripped, co-registered, 1 mm isotropic.
+- Measures **agreement with one annotation protocol on one dataset**. That is not a measure of
+  clinical accuracy.
+
+</details>
+
+---
+
+## The training notebook
+
+The model was trained in a single notebook on a Kaggle Tesla T4, and it is in the repository with
+its outputs intact: [`ml/notebooks/brats_segmentation_training_output.ipynb`](ml/notebooks/brats_segmentation_training_output.ipynb).
+GitHub renders it, so every number below can be traced to the cell that printed it. The clean
+unexecuted version is [`brats_segmentation_training.ipynb`](ml/notebooks/brats_segmentation_training.ipynb).
+
+Nineteen numbered stages, from configuration through to verifying the exported weights actually
+load. Twenty-two code cells, twenty-one of them executed.
+
+### What the data looked like before any model existed
+
+<div align="center">
+
+<img src="docs/notebook/01-tumour-volume-distribution.png" alt="Tumour volume distributions across the BraTS 2023 GLI training split" width="100%" />
+
+</div>
+
+Across 1,251 cases the whole tumour has a median volume of 89.3 cm3 and a range of 2.8 to 361.8
+cm3. The compartments are far smaller: enhancing tumour has a median of 17.3 cm3, and its minimum
+is **zero**, which is why the export applies a 200 voxel floor rather than reporting a
+one-voxel enhancing region as a finding.
+
+The single most consequential line the notebook printed:
+
+```
+tumour occupies 1.07% of all voxels
+-> uniform random patches would be almost pure background; sampling must be biased
+```
+
+That measurement is the reason patch sampling is 80% centred on the whole tumour and 20% uniform.
+It was not a hyperparameter guess.
+
+<div align="center">
+
+<img src="docs/notebook/02-sequences-and-labels.png" alt="The four co-registered sequences with the reference labels overlaid" width="100%" />
+
+</div>
+
+The four sequences for one case with the reference labels on T1C. Looking at the actual images is
+how the channel order was confirmed, and channel order is not recoverable from the weights: get it
+wrong and the model produces a plausible, wrong answer with no error anywhere.
+
+### Training
+
+<div align="center">
+
+<img src="docs/notebook/03-training-curve.png" alt="Training loss and validation Dice per epoch, with the selected epoch marked" width="100%" />
+
+</div>
+
+Loss on the left, per-region validation Dice on the right, with the selected epoch marked. Thirty
+seven of a planned sixty epochs completed before host RAM was exhausted, and **epoch 32 was
+selected on mean validation Dice at 0.8867**. The curve is what justifies stopping there rather
+than at the last epoch: validation had flattened well before the run ended.
+
+### The test set, run once
+
+<div align="center">
+
+<img src="docs/notebook/04-test-dice-distribution.png" alt="Test Dice per region, and Dice against tumour size" width="100%" />
+
+</div>
+
+186 cases from 169 subjects, held out at subject level. Per-region Dice on the left, Dice against
+tumour size on the right, and that right-hand panel is the honest one: agreement collapses on the
+smallest tumours, where a few voxels of disagreement dominate the metric.
+
+| Split comparison | Validation | Test | Gap |
+|---|---:|---:|---:|
+| Mean Dice | 0.8899 | **0.8938** | **-0.0039** |
+
+The test score is marginally *higher* than validation. There is no overfitting to report, and the
+generalisation gap is smaller than the run-to-run noise.
+
+| Region | Cases below 0.5 Dice | Cases with empty ground truth |
+|---|---:|---:|
+| Whole tumour | 2 of 186 | 0 |
+| Tumour core | 5 of 186 | 1 |
+| Enhancing tumour | 8 of 186 | 5 |
+
+### The cases it got wrong
+
+Most projects show the best cases. The notebook prints the worst four, because those are the ones
+that say something.
+
+<table>
+<tr>
+<td width="50%" valign="top">
+
+<img src="docs/notebook/05-worst-test-case-1.png" alt="Worst test case by whole-tumour Dice" width="100%" />
+
+`BraTS-GLI-00675-001`, whole tumour Dice **0.000**, yet tumour core and enhancing tumour both
+**1.000**. A whole tumour score of zero alongside perfect compartments is a labelling edge case,
+not a model that cannot see the tumour.
+
+</td>
+<td width="50%" valign="top">
+
+<img src="docs/notebook/06-worst-test-case-2.png" alt="Second worst test case by whole-tumour Dice" width="100%" />
+
+`BraTS-GLI-00493-000`, whole tumour Dice **0.184** on 34,451 labelled voxels, while core reaches
+0.924 and enhancing 0.883. The oedema boundary is the disagreement, which is the least
+reproducible boundary between human annotators too.
+
+</td>
+</tr>
+</table>
+
+> [!NOTE]
+> The worst case is where the enhancing floor and the quality gate earn their place. A study that
+> produces a result like this is flagged rather than reported as a confident measurement, and
+> nothing from this model reaches the measurement pipeline at all. See
+> [Validation & results](#validation--results) for why.
+
+Regenerate the figures from the notebook at any time:
+
+```bash
+make export-notebook-figures
+```
+
+## Validation & results
+
+### The decision not to ship the model into the measurement pipeline
+
+The trained model scores 0.894 mean Dice on held-out BraTS. It is **deliberately not** used for
+MEDTRACE's measurements, and the reason is measurement, not caution.
+
+Run against DeepBraTumIA on **12 randomly chosen real LUMIERE studies**
+(`ml/scripts/compare_on_lumiere.py`):
+
+| Region | Median Dice | Studies below 0.5 | Median Dice on a 0 to 1 scale |
+|---|---:|---:|:---|
+| Whole tumour | **0.923** | 0 of 12 | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱ |
+| Tumour core | 0.816 | 1 of 12 | ▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▱▱▱▱ |
+| **Enhancing tumour** | **0.486** | **6 of 12** | ▰▰▰▰▰▰▰▰▰▰▱▱▱▱▱▱▱▱▱▱ |
+
+Splitting by how much enhancement is actually present shows this is not a uniform weakness:
+
+| DeepBraTumIA enhancing volume | n | Median ET Dice | Our volume vs theirs |
+|---|---:|---:|---:|
+| Bulky, ≥ 5 cm³ | 5 | **0.861** | 1.03× |
+| Small, < 5 cm³ | 7 | **0.193** | ~4× |
+
+**The cause was predicted before the comparison was run.** BraTS is pre-operative glioma, where
+enhancing tumour is a thick contrast-avid ring. LUMIERE is post-treatment: resection margins,
+radiation change and post-surgical enhancement all enhance, and none of it appears in BraTS. The
+model has never been shown a brain that has been operated on.
+
+> [!CAUTION]
+> Enhancing tumour is **the compartment MEDTRACE measures and reports change on**. Swapping the
+> pipeline over would inflate every enhancing volume, worst on exactly the small lesions where a
+> change of a few tenths of a cm³ decides whether progression is reported. So the pipeline keeps
+> DeepBraTumIA's masks, and this comparison is documented as *agreement between two automated
+> tools*, not as accuracy.
+
+---
+
+## Measurement, matching and change
+
+### Measurement chosen by measurement
+
+Surface area is computed by marching cubes over a **signed distance field**, not over the binary
+mask directly. The estimator was selected by comparison against analytic shapes: it degrades far
+more gracefully with anisotropic voxels, and this data spans **0.36 mm to 6.0 mm** slice spacing.
+
+The reported surface area **is** the area of the mesh shipped to the 3D viewer, verified to
+within 0.009% across 1,153 meshes, so the number in the findings panel and the surface on screen
+cannot disagree.
+
+### Cross-time lesion matching
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","textColor":"#1e293b","nodeTextColor":"#1e293b","lineColor":"#475569","edgeLabelBackground":"#ffffff"},"flowchart":{"nodeSpacing":40,"rankSpacing":55,"padding":10}}}%%
+flowchart LR
+    P["Prior lesions"] --> S{"Score each pair"}
+    C["Current lesions"] --> S
+
+    S --> O["Overlap Dice"]
+    S --> D["Centroid distance"]
+    S --> X["Proximity"]
+    S --> V["Volume similarity"]
+
+    O --> M["Optimal assignment"]
+    D --> M
+    X --> M
+    V --> M
+
+    M --> R1["Matched"]
+    M --> R2["New lesion"]
+    M --> R3["Disappeared"]
+    M --> R4["Uncertain"]
+
+    classDef inp fill:#e0f2fe,stroke:#0284c7,stroke-width:1px,color:#0c4a6e
+    classDef comp fill:#ffffff,stroke:#64748b,stroke-width:1px,color:#1e293b
+    classDef dec fill:#ccfbf1,stroke:#0d9488,stroke-width:2px,color:#134e4a
+    classDef good fill:#dcfce7,stroke:#16a34a,stroke-width:1px,color:#14532d
+    classDef warn fill:#ffedd5,stroke:#ea580c,stroke-width:2px,color:#7c2d12
+
+    class P,C inp
+    class O,D,X,V comp
+    class S,M dec
+    class R1,R2,R3 good
+    class R4 warn
+```
+
+| Outcome | Meaning |
+|---|---|
+| **Matched** | Keeps the same `Lesion` id, so the trajectory continues |
+| **New lesion** | An unmatched current lesion, so a new `Lesion` identity is created |
+| **Disappeared** | An unmatched prior lesion, recorded as an absence, not silently dropped |
+| **Uncertain** | `MATCH_UNCERTAIN`, `WEAK_SCORE` or `AMBIGUOUS_ALTERNATIVE`. The finding is de-emphasised and the 3D surface stays grey |
+
+A doubtful correspondence is **never presented as a confident one**. `MATCH_UNCERTAIN`,
+`WEAK_SCORE` and `AMBIGUOUS_ALTERNATIVE` de-emphasise the finding in the panel and force the 3D
+surface to render grey rather than in a change colour, because a colour that says "growing" is
+a claim, and it must not be made when the lesion it is compared against may be a different
+lesion.
+
+### Change with propagated uncertainty
+
+<div align="center">
+
+<img src="docs/screenshots/04-findings-panel.png" alt="Findings panel with measured change and confidence" width="80%" />
+
+*Measured change per lesion, with confidence and the reason it is reduced.*
+
+</div>
+
+Confidence is reported **per pipeline stage**, and an answer's confidence is the **minimum**
+across the observations it rests on, never an average. A confident segmentation combined with an
+uncertain registration produces an uncertain change measurement, and averaging would let the
+reliable measurement hide the unreliable one.
+
+Eleven quality flags feed this, each translated into plain language: *"the two studies used
+different acquisition protocols"*, *"slice thickness was large enough to affect volume
+measurement"*, *"another lesion scored almost as well as this correspondence"*.
+
+---
+
+## 3D disease evolution
+
+<div align="center">
+
+<img src="GIF/gif-01-timeline-3d.gif" alt="Dragging the time scrubber updates the 3D disease map" width="90%" />
+
+*Dragging the time scrubber: the lesion surface, the measurements and the evidence move together.*
+
+<br />
+
+<img src="GIF/gif-02-3d-rotate.gif" alt="Rotating the volume-rendered head" width="90%" />
+
+*Rotation follows the pointer, and the head stays solid from every angle.*
+
+</div>
+
+### The head is volume-rendered, and that was a hard-won decision
+
+The brain context was originally a **surface** extracted from the skull-strip mask. It was
+reported as having holes five separate times. Each round found something real: open edges from a
+crop that borrowed its margin from the source volume, front-face culling that erased deep
+concavities, a camera whose view-up was parallel to its view direction, lesions left unlit by a
+single headlight. Each round fixed it, measured the rendered image as clean, and the report still stood.
+
+> [!NOTE]
+> A surface leaves only two options, and **each has a failure mode invisible to a software
+> rasteriser**. Translucent, and the result depends on blending order and multisample resolve,
+> which vary by driver. Opaque with the near wall culled, and it is a hollow bowl that hides the
+> anatomy it exists to show. The checks ran under SwiftShader; the defect lived on the GPU.
+
+The head is now the patient's own **skull-stripped contrast-enhanced T1, ray-cast as a volume**.
+There is no surface to close, no winding, no culling, no blending order, so *a gap in the anatomy is
+not expressible*. And it is the real anatomy at full 1 mm resolution rather than a smoothed
+approximation of its outer boundary.
+
+| Measurement | Measured on the Intel Iris Plus iGPU |
+|---|---|
+| Median frame | **16.7 ms → 59.9 fps** *(16.7 ms is the vsync interval, so the renderer is not the limit)* |
+| 95th percentile frame | 16.8 ms |
+| First frame after load | 4.0 s |
+| Main thread after a 60-step drag | 2 ms |
+
+Benchmarked on real hardware rather than in software, because software rendering is exactly how
+five rounds of a rendering defect stayed invisible. Run it yourself: `make benchmark-3d`.
+
+<table>
+<tr>
+<td width="50%" valign="top">
+
+<img src="docs/screenshots/05-3d-evolution.png" alt="3D disease evolution view" width="100%" />
+
+**Coloured by measured change.** Red > +25%, blue < −25%, green stable, **grey for a baseline or
+a doubtful match**. The prior timepoint is drawn as a wireframe.
+
+</td>
+<td width="50%" valign="top">
+
+<img src="docs/screenshots/06-3d-lesion-selected.png" alt="Single lesion selected in 3D" width="100%" />
+
+Selecting a lesion highlights it simultaneously in the 2D panes, the findings list and the 3D
+view. Lesion meshes are **never smoothed**, because that mesh is the source of the reported surface area.
+
+</td>
+</tr>
+</table>
+
+<details>
+<summary><b>Why the lesions stay as surfaces while the head is a volume</b></summary>
+
+<br />
+
+The lesions are the *measured* objects. Each needs its own colour for its own change, and a
+surface is the honest way to draw a boundary that came from a mask. They are extracted at full
+resolution (median 71 KB, max 715 KB per mesh) and shipped as binary PLY with per-vertex normals.
+
+`ml/scripts/check_mesh_integrity.py` verifies the shipped bytes rather than synthetic spheres:
+unit-length normals, outward orientation, triangle winding consistent with them, zero open edges,
+zero non-manifold edges, one connected component.
+
+</details>
+
+---
+
+## The evidence engine
+
+Five fixed clinical questions, answered from measurements already in the database, validated
+before delivery, and stored with the evidence that supports them.
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","actorFontSize":"15px","noteFontSize":"14px","messageFontSize":"14px","textColor":"#1e293b","actorTextColor":"#1e293b","noteTextColor":"#1e293b","signalTextColor":"#1e293b","actorBkg":"#e0f2fe","actorBorder":"#0284c7","noteBkgColor":"#fef3c7","noteBorderColor":"#d97706","labelBoxBkgColor":"#e0f2fe","labelTextColor":"#1e293b"}}}%%
+sequenceDiagram
+    autonumber
+    actor U as Clinician
+    participant F as findings
+    participant A as answers
+    participant L as LLM
+    participant S as SafetyGuard
+    participant D as Database
+
+    U->>F: Ask one of five questions
+    F->>D: Read recorded measurements
+    D-->>F: Observations, confidence, flags
+    F->>F: Assemble structured findings
+    Note over F: No evidence means no observation
+    F->>A: Compose deterministic answer
+    A-->>S: Ground truth
+    F->>L: Same findings, ask for prose
+    L-->>S: Draft, or nothing at all
+    S->>S: Forbidden claim? Invented number?
+    S-->>U: Answer, evidence, confidence
+    S->>D: Persist and audit
+```
+
+> [!IMPORTANT]
+> **The model never sees pixels, never computes a number, and never has the last word.**
+> `answers.py` composes the answer from findings alone with no model involved, and that sentence is
+> the ground truth. A language model may make it *more readable*; it may not make it *different*.
+
+<div align="center">
+
+<img src="docs/screenshots/07-evidence-panel.png" alt="AI evidence panel with an answer and evidence chips" width="90%" />
+
+<br />
+
+<img src="GIF/gif-03-show-me-why.gif" alt="From an answer to its evidence to the study it came from" width="90%" />
+
+*Every answer carries its evidence, and every piece of evidence navigates to the study it came from.*
+
+</div>
+
+An actual answer, generated from real measurements:
+
+> No measured enhancing volume changed by more than 25% between week-019-2 and week-033, 98 days
+> apart. 1 lesion changed by less: L06 measures 4.2 mm³, decreased by 5.7%. 5 lesions had no
+> prior to compare against… **Confidence 0.64; slice thickness was large enough to affect volume
+> measurement.**
+
+### SafetyGuard
+
+Implements a fixed table of permitted and forbidden statements, and nothing beyond it. What a medical
+tool may and may not state is not an engineering decision.
+
+| Check | Outcome |
+|---|---|
+| Diagnosis, tumour type or grade | Draft discarded → measurement delivered |
+| Progression, response, improvement, recurrence | Draft discarded → measurement delivered |
+| Treatment recommendation | Draft discarded → measurement delivered |
+| Prognosis or survival | Draft discarded → measurement delivered |
+| Clinical urgency | Draft discarded → measurement delivered |
+| A number not present in the structured findings | Draft discarded → measurement delivered |
+| A lesion or study not in the evidence set | Draft discarded → measurement delivered |
+| **No linked evidence** | **Blocked, nothing delivered** |
+| Confidence below 0.6 | Delivered, marked low, reason in plain language |
+| Quality flags on the inputs | Delivered, flags surfaced alongside |
+
+The regular expressions are deliberately coarse and are **not** treated as a semantic filter. A
+pattern cannot understand a sentence, so the guarantee comes from the deterministic fallback, not
+from the cleverness of the patterns. A blocked draft is **kept**, because a block is a signal
+about the pipeline rather than just a filtered string.
+
+<details>
+<summary><b>A worked example</b></summary>
+
+<br />
+
+```
+LLM draft:   "The tumour is malignant and has progressed."
+SafetyGuard: MODIFIED, diagnostic or grading claim; progression or response judgement
+Delivered:   "The segmented enhancing volume increased from 12.2 cm³ to 19.7 cm³
+              (+61.5%) between 2025-06-12 and 2025-09-04."
+```
+
+Two false positives were found by measurement and fixed: a timepoint called `week-012` parses as
+the number −12, and a lesion called `L01` as 1, so identifiers are stripped before the numeric
+scan. Rounding is not fabrication. 19.7 written as 20 is accepted, 42 is not.
+
+</details>
+
+### Reproducibility and audit
+
+`findings_hash` digests the question and the entire findings payload. Asking the same question
+about unchanged findings returns the **stored** answer rather than generating a second one, so a
+past statement stays reconstructable. Every generation writes an `AuditEvent` carrying the model
+version, the input hash, the output hash and the safety outcome, with the patient referenced by
+UUID and no findings in the detail.
+
+---
+
+## Architecture
+
+```mermaid
+%%{init: {"theme":"base","themeVariables":{"fontSize":"15px","clusterBkg":"#f8fafc","clusterBorder":"#94a3b8","textColor":"#1e293b","nodeTextColor":"#1e293b","lineColor":"#475569","edgeLabelBackground":"#ffffff"},"flowchart":{"nodeSpacing":40,"rankSpacing":62,"padding":12}}}%%
+flowchart TB
+    subgraph BROWSER ["Browser"]
+        W["Next.js · React"]
+        CS["Cornerstone3D"]
+        VTK["vtk.js"]
+    end
+
+    subgraph APILAYER ["FastAPI modular monolith"]
+        R["Routers"]
+        AN["Analysis"]
+        EV["Evidence engine"]
+        IG["Ingestion"]
+    end
+
+    subgraph MLLAYER ["ml, separate package"]
+        MM["medtrace_ml"]
+        ART["artifacts"]
+    end
+
+    subgraph SVC ["Docker Compose"]
+        PG[("PostgreSQL")]
+        MIO[("MinIO")]
+        ORT[("Orthanc")]
+        RD[("Redis")]
+    end
+
+    FS[("Filesystem")]
+
+    W --- CS
+    W --- VTK
+    W -->|"REST · PLY · NIfTI"| R
+    R --> AN
+    R --> EV
+    R --> IG
+    AN --> MM
+    EV --> MM
+    IG --> MM
+    MM --- ART
+    AN --> PG
+    IG --> PG
+    EV --> PG
+    IG --> ORT
+    AN --> MIO
+    R --> RD
+    R -->|"path allowlist"| FS
+
+    classDef ui fill:#e0f2fe,stroke:#0284c7,stroke-width:1px,color:#0c4a6e
+    classDef api fill:#ccfbf1,stroke:#0d9488,stroke-width:1px,color:#134e4a
+    classDef ml fill:#e0e7ff,stroke:#4f46e5,stroke-width:1px,color:#312e81
+    classDef svc fill:#f3e8ff,stroke:#9333ea,stroke-width:1px,color:#581c87
+    classDef fs fill:#fef3c7,stroke:#d97706,stroke-width:1px,color:#78350f
+
+    class W,CS,VTK ui
+    class R,AN,EV,IG api
+    class MM,ART ml
+    class PG,MIO,ORT,RD svc
+    class FS fs
+```
+
+<table>
+<tr><td valign="top" width="25%">
+
+**Browser**
+
+- `Next.js · React`: Zustand, TanStack Query
+- `Cornerstone3D`: 2D volumes + labelmaps
+- `vtk.js`: 3D volume ray-cast
+
+</td><td valign="top" width="25%">
+
+**FastAPI**
+
+- `Routers`: patients, studies, timeline, files, evidence
+- `Analysis`: pipeline orchestration
+- `Evidence engine`: findings, safety, audit
+- `Ingestion`: BraTS, LUMIERE, quality control
+
+</td><td valign="top" width="25%">
+
+**ml, a separate package**
+
+- `medtrace_ml`: measure, matching, change, mesh, volume, labels
+- `artifacts`: weights, TorchScript, model card
+
+</td><td valign="top" width="25%">
+
+**Services**
+
+- `PostgreSQL`: metadata only
+- `MinIO`: object storage
+- `Orthanc`: DICOM
+- `Redis`: cache
+- `Filesystem`: `data/raw`, `data/derived`
+
+</td></tr>
+</table>
+
+**ML code stays out of application code.** The application depends on model *contracts*, never on
+training code. Two Python packages, `medtrace-api` and `medtrace-ml`, with the API importing the
+latter but never the reverse.
+
+<div align="center">
+
+<img src="docs/screenshots/10-api-docs.png" alt="MEDTRACE OpenAPI documentation" width="90%" />
+
+*Every endpoint is typed end to end: Pydantic v2 on the server, generated TypeScript contracts in
+the browser.*
+
+</div>
+
+<details>
+<summary><b>Repository layout</b></summary>
+
+<br />
+
+```
+medtrace/
+├── apps/
+│   ├── api/                  FastAPI modular monolith
+│   │   ├── alembic/versions/ schema migrations
+│   │   ├── medtrace/
+│   │   │   ├── analysis/     the measurement pipeline
+│   │   │   ├── domain/       models.py, enums.py
+│   │   │   ├── evidence/     findings · answers · llm · safety · service
+│   │   │   ├── ingestion/    BraTS and LUMIERE readers, quality control
+│   │   │   └── routers/
+│   │   ├── scripts/          one-off data operations
+│   │   └── tests/
+│   └── web/                  Next.js clinical workstation
+│       ├── scripts/          browser verification and the 3D benchmark
+│       └── src/{app,components,lib,store}
+├── ml/
+│   ├── medtrace_ml/          measure · lesions · matching · change · trajectory
+│   │                         registration · mesh · volume · labels
+│   ├── notebooks/            training notebook, its generator and its checks
+│   ├── scripts/              dataset preparation, model and mesh verification
+│   ├── artifacts/            trained weights, TorchScript, model card
+│   └── reports/              evaluation and agreement CSVs
+├── packages/                 reserved for shared contracts
+├── infrastructure/           MLflow image, database init
+├── data/
+│   ├── raw/                  the datasets
+│   ├── derived/              generated meshes and atlas volumes
+│   └── kaggle/               upload staging, manifest and splits kept
+└── docs/screenshots/         interface captures used in this README
+```
+
+</details>
+
+---
+
+## Technology stack
+
+<table>
+<tr><td valign="top" width="50%">
+
+**Frontend**
+
+| Layer | Choice |
+|---|---|
+| Framework | Next.js 15.5 · React 19.1 |
+| Language | TypeScript 5.9 (strict) |
+| Styling | Tailwind CSS 4 |
+| Client state | Zustand 5 |
+| Server state | TanStack Query 5 |
+| 2D viewer | Cornerstone3D 5.7 |
+| 3D renderer | vtk.js 36.4 |
+
+**Backend**
+
+| Layer | Choice |
+|---|---|
+| API | FastAPI · Pydantic v2 |
+| ORM | SQLAlchemy 2 · Alembic |
+| Database | PostgreSQL 17 |
+| Object storage | MinIO |
+| DICOM | Orthanc 24.10 |
+| Cache | Redis 7 |
+
+</td><td valign="top" width="50%">
+
+**Machine learning**
+
+| Layer | Choice |
+|---|---|
+| Framework | PyTorch 2.10 (cu128) |
+| Medical DL | MONAI 1.6 |
+| Registration | SimpleITK |
+| Imaging I/O | nibabel · NumPy · SciPy |
+| Meshing | scikit-image marching cubes |
+| Training | Kaggle Tesla T4 |
+| LLM | MedGemma (text-only) behind a provider adapter |
+
+**Engineering**
+
+| Layer | Choice |
+|---|---|
+| Infrastructure | Docker Compose, 5 services |
+| Testing | Pytest · Vitest · Playwright |
+| Linting | Ruff · ESLint |
+| CI | GitHub Actions |
+
+</td></tr>
+</table>
+
+> [!NOTE]
+> **On the LLM:** the default configuration is `MEDTRACE_LLM_PROVIDER=none`, and that is a
+> *working* configuration rather than a disabled one. MedGemma-27B does not fit on the
+> development GPU, and the five clinical questions must be answerable regardless, so the
+> deterministic composer answers them. Point `MEDTRACE_LLM_BASE_URL` at a vLLM, Ollama or
+> llama.cpp endpoint and it will rephrase; if that endpoint is unreachable you lose wording, not
+> correctness.
+
+---
+
+## Datasets
+
+Public, de-identified research data only. No real hospital patient data enters this repository
+under any circumstances.
+
+| Dataset | Role | Scale ingested |
+|---|---|---|
+| **BraTS 2023 GLI** | Segmentation training and held-out evaluation | 1,251 cases from 1,133 subjects · 118 with two labelled timepoints |
+| **LUMIERE** | The longitudinal dataset: timeline, matching, trajectory | 91 patients · 638 timepoints · 2,487 series |
+
+LUMIERE's acquisition metadata records **three field strengths, 21 scanner models and slice
+thickness from 0.8 mm to 6.0 mm**, so quality control and confidence reporting are built against
+measured heterogeneity rather than imagined inputs.
+
+<details>
+<summary><b>Data we hold and deliberately will not use</b></summary>
+
+<br />
+
+LUMIERE ships **survival time in weeks, IDH status and MGMT methylation** for all 91 patients.
+That makes outcome and molecular prediction technically possible with the data already on disk.
+
+We are not building it. Prognosis is forbidden by the safety policy this project holds itself to, and a
+measurement system must be trustworthy before prediction built on top of it means anything. This
+is recorded as a deliberate decision rather than an oversight, so the temptation is resolved once
+instead of repeatedly.
+
+LUMIERE's **expert RANO ratings for 616 timepoints** are likewise used only as a reference
+standard for evaluation. Never a training target, never surfaced as a MEDTRACE output.
+
+</details>
+
+---
+
+## Verification
+
+Nothing here is asserted without measurement. **696 automated checks.**
+
+| Suite | Checks | What it proves |
+|---|---:|---|
+| `pytest` API | **155** | Contracts, ingestion, quality control, pipeline output, SafetyGuard |
+| `pytest` ML | **168** | Measurement, matching, change, meshing, volume windowing, label mappings |
+| `vitest` web | **17** | Camera conventions, hole detection geometry |
+| `verify-study-linkage` | **189** | Every pane displays the study it claims, at every timepoint |
+| `verify-evidence` | **42** | Answers carry evidence, cite only measured numbers, and navigate |
+| `verify-evolution` | **37** | The 3D view draws the right surfaces, in one coordinate frame |
+| `verify-workstation` | **19** | The clinical shell, the scrubber, and the intended-use notice |
+| `verify-brain-shell` | **15** | No holes in the rendered head, at five viewing angles |
+| `verify-overlay` | **15** | The 2D tumour overlay draws the right mask |
+| `verify-slice-sync` | **14** | The two panes really scroll together, and stop when unlinked |
+| `verify-findings` | **8** | The UI shows real measured findings, not seeded numbers |
+
+<details>
+<summary><b>Why the browser checks exist at all</b></summary>
+
+<br />
+
+Because instrumentation has been wrong more than once, and each time it was wrong in a way that
+*passed*:
+
+- A hole detector that counted the gaps between the legend's **text glyphs**, because an element
+  screenshot captures whatever is drawn over the element.
+- The same check running on the patient the app opens on, which has a single 96-pixel lesion, and
+  and passing with zero holes while the defect was obvious on a patient with twelve.
+- A watertightness test that only ever ran at stride 1, while the shipped meshes used stride 2.
+- A "fit view" check that measured the brain instead of the lesion.
+
+Expected values are taken **from the API**, not from the page. The hole detector now lives in its
+own module with its own unit tests, because three wrong versions of a check is enough.
+
+</details>
+
+---
+
+## Getting started
+
+### Prerequisites
+
+| Requirement | Needed |
+|---|---|
+| Docker Desktop | Running, with the Linux engine |
+| Python | ≥ 3.11 *(resolved against 3.14)* |
+| Node.js | ≥ 20 *(tested on 24)* |
+| Disk | ~50 GB for both datasets |
+
+### First-time setup
+
+```bash
+cp .env.example .env
+
+make up          # PostgreSQL, Redis, MinIO, Orthanc, MLflow
+make install     # API venv + ML venv + web dependencies
+make migrate     # create the schema
+make seed        # synthetic demonstration patient (labelled synthetic in the UI)
+```
+
+Then ingest and analyse. This is the slow part:
+
+```bash
+make ingest-lumiere    # 91 patients, 638 timepoints
+make ingest-brats      # 118 subjects with two labelled timepoints
+make analyse           # the full measurement pipeline  (~20 min)
+make link-masks        # reorient each series' mask for the 2D overlay
+```
+
+### Running it
+
+Three terminals, in this order:
+
+```bash
+# 1. services
+docker compose up -d
+
+# 2. API
+cd apps/api
+.venv/Scripts/python.exe -m uvicorn medtrace.main:app --host 127.0.0.1 --port 8000
+
+# 3. web (production build, see the note below)
+cd apps/web
+npm run start
+```
+
+| Service | Address |
+|---|---|
+| **Workstation** | **http://localhost:3000** |
+| API documentation | http://localhost:8000/docs |
+| Health check | http://localhost:8000/health |
+| Orthanc | http://localhost:8042 |
+| MinIO console | http://localhost:9101 |
+
+**Stop cleanly:** `Ctrl+C` in the web and API terminals, then `docker compose stop`.
+
+> [!TIP]
+> Use `npm run start`, not `npm run dev`. Volume rendering is the heaviest part of MEDTRACE, and
+> development mode's HMR, source maps and double-invoked effects consume memory the renderer
+> needs. Never run `npm run build` while a dev server is running, because they share `.next`.
+>
+> The viewer caps its cache at **500 MB and six resident volumes**, shows its footprint in the
+> toolbar, and offers a `reset viewer` control if the graphics context is ever lost.
+
+> [!WARNING]
+> **No authentication.** The API is unauthenticated and binds to `127.0.0.1` only, with
+> development credentials from `.env`. It is for local use. Do not expose it.
+
+<details>
+<summary><b>Optional commands</b></summary>
+
+<br />
+
+```bash
+# Tests
+cd apps/api && .venv/Scripts/python.exe -m pytest -q      # 155
+cd ml       && ../apps/api/.venv/Scripts/python.exe -m pytest -q   # 168
+cd apps/web && npm run test                                # 22 unit
+
+# Browser verification (needs both servers running)
+make verify                    # every browser suite
+make verify-evidence           # the AI evidence engine
+make verify-brain-shell        # no holes in the rendered head
+make benchmark-3d              # 3D frame rate on the real GPU
+
+# Quality
+make lint                      # Ruff + ESLint
+make typecheck                 # tsc --noEmit
+
+# Data and model inspection
+make inspect p=Patient-006     # per-lesion trajectories
+make quality-report            # what quality control found
+make check-mesh-integrity      # normals, winding, watertightness of shipped meshes
+make check-volume-pockets      # no enclosed pockets of air inside the 3D display volumes
+make verify-meshes             # every stored mesh against its observation
+make check-readme-diagrams     # render this README's diagrams, check no label overflows
+```
+
+Ports are shifted off the defaults, **PostgreSQL on 5433 and MinIO console on 9101**, so the stack
+does not collide with locally installed services.
+
+</details>
+
+---
+
+## Build status
+
+| Stage | Scope | State |
+|---|---|:---:|
+| **0** | Clinical problem, scope, safety boundaries, evaluation plan, architecture, dataset cards | Done |
+| **1** | Docker Compose infrastructure, FastAPI, 14-table observation model, clinical workstation | Done |
+| **2** | Dataset ingestion, quality control, Cornerstone3D medical viewer, slice synchronisation | Done |
+| **3** | Measurement, lesion extraction, registration, cross-time matching, change detection, trajectory | Done |
+| **4** | 3D disease evolution, volume rendering, surface extraction tied to the time scrubber | Done |
+| **5** | Evidence engine, SafetyGuard, five clinical questions, audit trail | Done |
+
+**Deliberately not built:** MLOps automation (DVC, MLflow registry integration, Great
+Expectations), authentication and role-based access, deformable registration, progression
+classification, prognosis. The first is a project decision; the rest are scoped to later versions.
+
+---
+
+## Honesty as a design principle
+
+MEDTRACE is presented as a prototype, because that is what it is. Understanding why clinical
+validation matters, and saying so plainly, is a strength when talking to clinicians, not a
+weakness to hide behind confident language.
+
+Every number in this README is measured and reproducible from the repository. Where a result is
+unflattering, it is stated: the model's enhancing-tumour agreement on post-treatment data is
+poor, and that is why it is not in the measurement pipeline.
+
+---
+
+<div align="center">
+
+**Datasets:** BraTS 2023 GLI and LUMIERE, used under their respective research licences.
+LUMIERE is non-commercial.
+
+<br />
+
+*MEDTRACE reports measured change. It does not diagnose.*
+
+</div>
+
+---
+
+<div align="center">
+
+### Omar Rehan
+
+[![Portfolio](https://img.shields.io/badge/Portfolio-000000?style=for-the-badge&logo=vercel&logoColor=white)](https://omar-rehan.vercel.app/)
+[![GitHub](https://img.shields.io/badge/GitHub-181717?style=for-the-badge&logo=github&logoColor=white)](https://github.com/AIOmarRehan)
+[![LinkedIn](https://img.shields.io/badge/LinkedIn-0A66C2?style=for-the-badge&logo=linkedin&logoColor=white)](https://linkedin.com/in/omar-rehan-47b98636a)
+[![Hugging Face](https://img.shields.io/badge/Hugging%20Face-FFD21E?style=for-the-badge&logo=huggingface&logoColor=000000)](https://huggingface.co/AIOmarRehan)
+
+[![Kaggle](https://img.shields.io/badge/Kaggle-20BEFF?style=for-the-badge&logo=kaggle&logoColor=white)](https://kaggle.com/aiomarrehan)
+[![Medium](https://img.shields.io/badge/Medium-000000?style=for-the-badge&logo=medium&logoColor=white)](https://medium.com/@ai.omar.rehan)
+[![Tableau Public](https://img.shields.io/badge/Tableau%20Public-E97627?style=for-the-badge&logo=tableau&logoColor=white)](https://public.tableau.com/app/profile/omar.rehan)
+
+</div>

api/patients.json

@@ -0,0 +1 @@
+[{"age_at_baseline_years":65,"external_id":"RHUH-0001","id":"eb1d080b-101f-45ce-8455-cf9f4967dd55","is_synthetic":false,"lesion_count":2,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":65,"external_id":"RHUH-0002","id":"4709627e-6d9f-467a-9c36-76e45ed24071","is_synthetic":false,"lesion_count":5,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":75,"external_id":"RHUH-0003","id":"37ab76a9-f8d9-4ccb-8ae7-5b617bc333f8","is_synthetic":false,"lesion_count":4,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":64,"external_id":"RHUH-0004","id":"0e36f014-cba2-4b67-a60a-d5c720f13384","is_synthetic":false,"lesion_count":2,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":52,"external_id":"RHUH-0005","id":"a2cf6c6a-db6d-493a-9f48-ada357b32b83","is_synthetic":false,"lesion_count":2,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":72,"external_id":"RHUH-0006","id":"1644cc40-79e6-47bf-a4e4-51c8248a7ad9","is_synthetic":false,"lesion_count":3,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":51,"external_id":"RHUH-0007","id":"6bc843a9-e804-46d5-9d7d-d387a1f9c1ed","is_synthetic":false,"lesion_count":3,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":69,"external_id":"RHUH-0008","id":"e6fe336a-7b59-43a7-a419-3f01b51196bb","is_synthetic":false,"lesion_count":2,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":68,"external_id":"RHUH-0009","id":"8180e98b-ffa6-4f09-b73b-b35f9937a765","is_synthetic":false,"lesion_count":2,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":66,"external_id":"RHUH-0010","id":"75835d9b-32d4-4c9a-8cf9-8f57b066084b","is_synthetic":false,"lesion_count":2,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":71,"external_id":"RHUH-0011","id":"2043be8c-57f4-4845-88d0-cadadd106f19","is_synthetic":false,"lesion_count":3,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":64,"external_id":"RHUH-0012","id":"041c2146-8e58-4c2d-b5dd-6c888ff2447e","is_synthetic":false,"lesion_count":2,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":54,"external_id":"RHUH-0013","id":"31d3fa6c-d577-4972-9845-fd45c6d03801","is_synthetic":false,"lesion_count":4,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":78,"external_id":"RHUH-0014","id":"65c97660-6722-4a04-8a81-840ce91c1a8d","is_synthetic":false,"lesion_count":11,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":60,"external_id":"RHUH-0015","id":"92f0ea5a-eacc-42a8-a3c6-2345fc9e3722","is_synthetic":false,"lesion_count":4,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":68,"external_id":"RHUH-0016","id":"904c26dc-10ec-4454-9767-9758b9dc9789","is_synthetic":false,"lesion_count":3,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":47,"external_id":"RHUH-0017","id":"2be8a602-8102-4bf4-a6a0-87df5f84f3c1","is_synthetic":false,"lesion_count":4,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":57,"external_id":"RHUH-0018","id":"9bf9a4c8-66f6-470a-a03a-e425c4f5f438","is_synthetic":false,"lesion_count":8,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":61,"external_id":"RHUH-0019","id":"05e8f381-fc6b-4afa-a8ae-e1a8888b7be3","is_synthetic":false,"lesion_count":4,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":55,"external_id":"RHUH-0020","id":"add7899e-4dcb-4587-9733-ec0a4068cfa5","is_synthetic":false,"lesion_count":7,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":56,"external_id":"RHUH-0021","id":"8280c78d-5068-4b5a-ac27-d33c0839b167","is_synthetic":false,"lesion_count":1,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":78,"external_id":"RHUH-0022","id":"4ec108df-781d-4553-b5e0-56aa6c18c94a","is_synthetic":false,"lesion_count":1,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":59,"external_id":"RHUH-0023","id":"69ea08eb-e7e9-4728-b9f5-4eaeac564c4a","is_synthetic":false,"lesion_count":5,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":50,"external_id":"RHUH-0024","id":"5cbd85a6-01b0-40aa-b6ba-eb8a5dd942c3","is_synthetic":false,"lesion_count":3,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":49,"external_id":"RHUH-0025","id":"0a1ba0ca-23d8-46ed-9974-f86a08bf7027","is_synthetic":false,"lesion_count":1,"sex":"male","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":63,"external_id":"RHUH-0026","id":"d84b4c77-418f-4934-8b0b-257a47d73d31","is_synthetic":false,"lesion_count":5,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":77,"external_id":"RHUH-0027","id":"2166fc98-aac6-45a6-855a-bad61c98e806","is_synthetic":false,"lesion_count":4,"sex":"female","source_dataset":"RHUH_GBM","study_count":3},{"age_at_baseline_years":75,"external_id":"RHUH-0029","id":"cd9672c3-d41d-42ee-8245-d0579bcafe2b","is_synthetic":false,"lesion_count":8,"sex":"male","source_dataset":"RHUH_GBM","study_count":3}]

api/segmentations/009ae3f3-123e-4211-a6b9-019aef159d79.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"dc80cacc-f224-41a9-9922-3013494be267","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0003/1/RHUH-0003_1_segmentations.nii.gz","mesh_uri":"brain/dc80cacc-f224-41a9-9922-3013494be267.ply","source":"RHUH_GBM","study_id":"009ae3f3-123e-4211-a6b9-019aef159d79"}]

api/segmentations/02502ba4-5660-47b5-bffa-2670a34dfe7b.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"77293132-ff78-4a35-9870-de45465211b7","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0014/0/RHUH-0014_0_segmentations.nii.gz","mesh_uri":"brain/77293132-ff78-4a35-9870-de45465211b7.ply","source":"RHUH_GBM","study_id":"02502ba4-5660-47b5-bffa-2670a34dfe7b"}]

api/segmentations/026d5991-8bf0-4f22-9be3-a427cd017cdb.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"e4b5c98b-8b76-4f8d-b751-24cb9f9649ef","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0026/2/RHUH-0026_2_segmentations.nii.gz","mesh_uri":"brain/e4b5c98b-8b76-4f8d-b751-24cb9f9649ef.ply","source":"RHUH_GBM","study_id":"026d5991-8bf0-4f22-9be3-a427cd017cdb"}]

api/segmentations/059c5744-67cd-40d2-8d09-a44ccb563c16.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"4299b34b-d8ed-41fb-b20b-7b7b84377e47","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0021/0/RHUH-0021_0_segmentations.nii.gz","mesh_uri":"brain/4299b34b-d8ed-41fb-b20b-7b7b84377e47.ply","source":"RHUH_GBM","study_id":"059c5744-67cd-40d2-8d09-a44ccb563c16"}]

api/segmentations/0664cdeb-7fac-4da4-a80a-a55c6723ae2e.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"6c59d7cc-89de-4253-901e-c84dc994ba9b","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0011/0/RHUH-0011_0_segmentations.nii.gz","mesh_uri":"brain/6c59d7cc-89de-4253-901e-c84dc994ba9b.ply","source":"RHUH_GBM","study_id":"0664cdeb-7fac-4da4-a80a-a55c6723ae2e"}]

api/segmentations/06f843bf-75d5-45a7-8630-56ea34a8acc0.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"2a50800e-6e2c-40de-a8d9-9d709a165904","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0010/2/RHUH-0010_2_segmentations.nii.gz","mesh_uri":"brain/2a50800e-6e2c-40de-a8d9-9d709a165904.ply","source":"RHUH_GBM","study_id":"06f843bf-75d5-45a7-8630-56ea34a8acc0"}]

api/segmentations/071261ee-f4de-4fd3-a188-c9c1111010c3.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"f99740a9-da84-4aae-b754-43dd4e1e4c09","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0013/0/RHUH-0013_0_segmentations.nii.gz","mesh_uri":"brain/f99740a9-da84-4aae-b754-43dd4e1e4c09.ply","source":"RHUH_GBM","study_id":"071261ee-f4de-4fd3-a188-c9c1111010c3"}]

api/segmentations/0a15d8e4-145f-4cab-9167-8b13a9ff444f.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"36ac490e-b72b-46a1-ad06-ed6294dd8955","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0019/1/RHUH-0019_1_segmentations.nii.gz","mesh_uri":"brain/36ac490e-b72b-46a1-ad06-ed6294dd8955.ply","source":"RHUH_GBM","study_id":"0a15d8e4-145f-4cab-9167-8b13a9ff444f"}]

api/segmentations/0cd019eb-859f-4687-8c74-fa0b7a36e4e4.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"4a5ac6a0-ff86-47a7-bb3d-b4aa1254c00b","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0009/2/RHUH-0009_2_segmentations.nii.gz","mesh_uri":"brain/4a5ac6a0-ff86-47a7-bb3d-b4aa1254c00b.ply","source":"RHUH_GBM","study_id":"0cd019eb-859f-4687-8c74-fa0b7a36e4e4"}]

api/segmentations/0f6cdab3-55b8-48fa-82ee-f8b17515cab4.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"05001dda-3f4c-4e8f-95e3-cade43dba021","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0029/0/RHUH-0029_0_segmentations.nii.gz","mesh_uri":"brain/05001dda-3f4c-4e8f-95e3-cade43dba021.ply","source":"RHUH_GBM","study_id":"0f6cdab3-55b8-48fa-82ee-f8b17515cab4"}]

api/segmentations/11d9fd96-2cf5-4a80-8a39-8ad4ce075a86.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"c575a24f-ff31-4d44-8173-17ca676ab173","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0025/0/RHUH-0025_0_segmentations.nii.gz","mesh_uri":"brain/c575a24f-ff31-4d44-8173-17ca676ab173.ply","source":"RHUH_GBM","study_id":"11d9fd96-2cf5-4a80-8a39-8ad4ce075a86"}]

api/segmentations/1348e608-daf3-4473-a3f4-a8fb177d8534.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"f28d1a88-a293-4b37-81e8-5290d9b215c9","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0016/1/RHUH-0016_1_segmentations.nii.gz","mesh_uri":"brain/f28d1a88-a293-4b37-81e8-5290d9b215c9.ply","source":"RHUH_GBM","study_id":"1348e608-daf3-4473-a3f4-a8fb177d8534"}]

api/segmentations/164949bd-e96c-4a5f-9003-18a43fe53ec6.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"14eb0518-14d2-4f04-8335-ed449444b8bc","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0002/0/RHUH-0002_0_segmentations.nii.gz","mesh_uri":"brain/14eb0518-14d2-4f04-8335-ed449444b8bc.ply","source":"RHUH_GBM","study_id":"164949bd-e96c-4a5f-9003-18a43fe53ec6"}]

api/segmentations/169e9c49-2ef2-4c22-9245-3dc3ad32b545.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"87ee3851-7126-40ca-8e0c-447e099be7d7","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0001/0/RHUH-0001_0_segmentations.nii.gz","mesh_uri":"brain/87ee3851-7126-40ca-8e0c-447e099be7d7.ply","source":"RHUH_GBM","study_id":"169e9c49-2ef2-4c22-9245-3dc3ad32b545"}]

api/segmentations/1922d1cd-b1f5-441f-a0c2-ac2249440084.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"cfef0ad5-2a44-4d97-814d-b3dfe567339d","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0002/1/RHUH-0002_1_segmentations.nii.gz","mesh_uri":"brain/cfef0ad5-2a44-4d97-814d-b3dfe567339d.ply","source":"RHUH_GBM","study_id":"1922d1cd-b1f5-441f-a0c2-ac2249440084"}]

api/segmentations/1fec1237-f8c1-4712-a3d8-32d9a334f24e.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"a578b6e3-93b6-4b6f-9635-d99fcb942476","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0020/0/RHUH-0020_0_segmentations.nii.gz","mesh_uri":"brain/a578b6e3-93b6-4b6f-9635-d99fcb942476.ply","source":"RHUH_GBM","study_id":"1fec1237-f8c1-4712-a3d8-32d9a334f24e"}]

api/segmentations/2463dc35-d8b6-4998-b056-c2e12d7271dd.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"c1a1cba9-79d8-4c60-8c98-d2695545de18","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0023/0/RHUH-0023_0_segmentations.nii.gz","mesh_uri":"brain/c1a1cba9-79d8-4c60-8c98-d2695545de18.ply","source":"RHUH_GBM","study_id":"2463dc35-d8b6-4998-b056-c2e12d7271dd"}]

api/segmentations/25404115-694f-41e1-8308-c135a6327bb1.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"561474ec-d496-434c-9ed2-a59b32614966","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0005/1/RHUH-0005_1_segmentations.nii.gz","mesh_uri":"brain/561474ec-d496-434c-9ed2-a59b32614966.ply","source":"RHUH_GBM","study_id":"25404115-694f-41e1-8308-c135a6327bb1"}]

api/segmentations/25a4be4c-f837-4009-be28-7f070e8e5dd4.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"da250565-ef2d-4e4e-8743-688c26c7b343","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0018/0/RHUH-0018_0_segmentations.nii.gz","mesh_uri":"brain/da250565-ef2d-4e4e-8743-688c26c7b343.ply","source":"RHUH_GBM","study_id":"25a4be4c-f837-4009-be28-7f070e8e5dd4"}]

api/segmentations/26166fcc-7850-41ae-bd87-de6d814a5c30.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"2d176bc6-07ac-4857-8892-650d0db26a99","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0022/1/RHUH-0022_1_segmentations.nii.gz","mesh_uri":"brain/2d176bc6-07ac-4857-8892-650d0db26a99.ply","source":"RHUH_GBM","study_id":"26166fcc-7850-41ae-bd87-de6d814a5c30"}]

api/segmentations/26272b5b-80dd-4201-8001-3de172691935.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"6732aaa8-1337-4ae5-9109-1c649c92f9e0","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0025/2/RHUH-0025_2_segmentations.nii.gz","mesh_uri":"brain/6732aaa8-1337-4ae5-9109-1c649c92f9e0.ply","source":"RHUH_GBM","study_id":"26272b5b-80dd-4201-8001-3de172691935"}]

api/segmentations/27e46bbd-9461-4c3d-8fde-414291f5bc72.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"e8899c64-4be7-40cd-80cc-45b2fa5a8763","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0019/0/RHUH-0019_0_segmentations.nii.gz","mesh_uri":"brain/e8899c64-4be7-40cd-80cc-45b2fa5a8763.ply","source":"RHUH_GBM","study_id":"27e46bbd-9461-4c3d-8fde-414291f5bc72"}]

api/segmentations/2a721993-2746-49fd-ab08-363c33187bc1.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"cd37e9ea-50e5-44df-b8ad-ea15e9a3d796","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0025/1/RHUH-0025_1_segmentations.nii.gz","mesh_uri":"brain/cd37e9ea-50e5-44df-b8ad-ea15e9a3d796.ply","source":"RHUH_GBM","study_id":"2a721993-2746-49fd-ab08-363c33187bc1"}]

api/segmentations/2db7f389-59e8-4f51-892a-45aa64724a4d.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"a71b3592-bfe9-4e0b-9068-bb631384002b","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0024/1/RHUH-0024_1_segmentations.nii.gz","mesh_uri":"brain/a71b3592-bfe9-4e0b-9068-bb631384002b.ply","source":"RHUH_GBM","study_id":"2db7f389-59e8-4f51-892a-45aa64724a4d"}]

api/segmentations/2f4c8574-9ea0-4334-b96f-29f82fd556d5.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"2821cef8-9b27-4b9a-84fa-7f915fa4a612","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0011/1/RHUH-0011_1_segmentations.nii.gz","mesh_uri":"brain/2821cef8-9b27-4b9a-84fa-7f915fa4a612.ply","source":"RHUH_GBM","study_id":"2f4c8574-9ea0-4334-b96f-29f82fd556d5"}]

api/segmentations/311dcc60-732d-49ef-98e8-146429b35fa8.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"4f07bf81-09ac-4ac2-b4c8-37304675e8fc","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0004/1/RHUH-0004_1_segmentations.nii.gz","mesh_uri":"brain/4f07bf81-09ac-4ac2-b4c8-37304675e8fc.ply","source":"RHUH_GBM","study_id":"311dcc60-732d-49ef-98e8-146429b35fa8"}]

api/segmentations/33bf232d-ced9-495b-9152-eb6d149877d3.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"d702746d-798e-406c-8902-464971c4c159","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0014/1/RHUH-0014_1_segmentations.nii.gz","mesh_uri":"brain/d702746d-798e-406c-8902-464971c4c159.ply","source":"RHUH_GBM","study_id":"33bf232d-ced9-495b-9152-eb6d149877d3"}]

api/segmentations/3c71eff6-5c5c-4e23-a140-eeb7fc48c2aa.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"2287d11f-6502-4012-a204-40ef3210bb0f","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0005/0/RHUH-0005_0_segmentations.nii.gz","mesh_uri":"brain/2287d11f-6502-4012-a204-40ef3210bb0f.ply","source":"RHUH_GBM","study_id":"3c71eff6-5c5c-4e23-a140-eeb7fc48c2aa"}]

api/segmentations/3d314d09-a3c4-4ca8-be24-8d35bfcab419.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"ecab29f4-6f5c-496c-94ec-c8401bcf4447","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0029/2/RHUH-0029_2_segmentations.nii.gz","mesh_uri":"brain/ecab29f4-6f5c-496c-94ec-c8401bcf4447.ply","source":"RHUH_GBM","study_id":"3d314d09-a3c4-4ca8-be24-8d35bfcab419"}]

api/segmentations/4815a12f-0b7c-467f-ba55-afcd5064394a.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"54ba54f9-c979-42ef-9fc1-761ad8a30ceb","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0008/0/RHUH-0008_0_segmentations.nii.gz","mesh_uri":"brain/54ba54f9-c979-42ef-9fc1-761ad8a30ceb.ply","source":"RHUH_GBM","study_id":"4815a12f-0b7c-467f-ba55-afcd5064394a"}]

api/segmentations/495eb8a6-8859-4a12-9c11-b0b88f665c76.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"33609cc1-994c-471f-9eeb-dab5e914b109","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0006/1/RHUH-0006_1_segmentations.nii.gz","mesh_uri":"brain/33609cc1-994c-471f-9eeb-dab5e914b109.ply","source":"RHUH_GBM","study_id":"495eb8a6-8859-4a12-9c11-b0b88f665c76"}]

api/segmentations/49fff636-3d0c-4bbc-866d-fd34845e8803.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"719f4e7e-d5c3-44f6-a3de-aac16d9958e1","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0002/2/RHUH-0002_2_segmentations.nii.gz","mesh_uri":"brain/719f4e7e-d5c3-44f6-a3de-aac16d9958e1.ply","source":"RHUH_GBM","study_id":"49fff636-3d0c-4bbc-866d-fd34845e8803"}]

api/segmentations/4f456804-060e-47f1-a54b-c714bfcf58dc.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"767168e0-94f5-45c9-a271-3990b020975a","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0016/2/RHUH-0016_2_segmentations.nii.gz","mesh_uri":"brain/767168e0-94f5-45c9-a271-3990b020975a.ply","source":"RHUH_GBM","study_id":"4f456804-060e-47f1-a54b-c714bfcf58dc"}]

api/segmentations/50c7e8b0-067e-4c68-b476-5f78dfcd7a62.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"ae882a7f-ab0b-486c-b363-3a626b4a0e2b","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0003/0/RHUH-0003_0_segmentations.nii.gz","mesh_uri":"brain/ae882a7f-ab0b-486c-b363-3a626b4a0e2b.ply","source":"RHUH_GBM","study_id":"50c7e8b0-067e-4c68-b476-5f78dfcd7a62"}]

api/segmentations/53b1674c-bbf3-4bfd-aa8c-61517d151eb1.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"9979af6c-73a5-4951-b6da-280581cc4d42","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0013/1/RHUH-0013_1_segmentations.nii.gz","mesh_uri":"brain/9979af6c-73a5-4951-b6da-280581cc4d42.ply","source":"RHUH_GBM","study_id":"53b1674c-bbf3-4bfd-aa8c-61517d151eb1"}]

api/segmentations/5b364a21-a771-4f24-b34f-648c56159bb3.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"0f5451b0-55c8-430c-9ef8-132956220c72","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0008/2/RHUH-0008_2_segmentations.nii.gz","mesh_uri":"brain/0f5451b0-55c8-430c-9ef8-132956220c72.ply","source":"RHUH_GBM","study_id":"5b364a21-a771-4f24-b34f-648c56159bb3"}]

api/segmentations/62ec3a94-03fb-4477-860a-e4068c3f2bad.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"8f9ce7cd-dcf7-4d44-bd78-39cbfb7420f4","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0013/2/RHUH-0013_2_segmentations.nii.gz","mesh_uri":"brain/8f9ce7cd-dcf7-4d44-bd78-39cbfb7420f4.ply","source":"RHUH_GBM","study_id":"62ec3a94-03fb-4477-860a-e4068c3f2bad"}]

api/segmentations/6327e26a-4d95-4c6c-855d-f24ae5f91ce5.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"7c9f5b8d-3525-4057-8c8b-951ed932b830","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0007/0/RHUH-0007_0_segmentations.nii.gz","mesh_uri":"brain/7c9f5b8d-3525-4057-8c8b-951ed932b830.ply","source":"RHUH_GBM","study_id":"6327e26a-4d95-4c6c-855d-f24ae5f91ce5"}]

api/segmentations/63ad442f-86f8-4a05-abb1-2aa4b1f9f5d5.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"283eba5f-5195-4775-89e7-b5d69a161841","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0029/1/RHUH-0029_1_segmentations.nii.gz","mesh_uri":"brain/283eba5f-5195-4775-89e7-b5d69a161841.ply","source":"RHUH_GBM","study_id":"63ad442f-86f8-4a05-abb1-2aa4b1f9f5d5"}]

api/segmentations/6b329524-c37b-4581-9859-84a55b0f7d84.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"4fcde278-7d47-47bd-869a-1227360a9af0","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0006/2/RHUH-0006_2_segmentations.nii.gz","mesh_uri":"brain/4fcde278-7d47-47bd-869a-1227360a9af0.ply","source":"RHUH_GBM","study_id":"6b329524-c37b-4581-9859-84a55b0f7d84"}]

api/segmentations/71f83359-3065-4843-8c63-6d27998c4099.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"7df6f57a-064b-4f01-bda5-b832ec861a81","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0017/0/RHUH-0017_0_segmentations.nii.gz","mesh_uri":"brain/7df6f57a-064b-4f01-bda5-b832ec861a81.ply","source":"RHUH_GBM","study_id":"71f83359-3065-4843-8c63-6d27998c4099"}]

api/segmentations/7388fa67-3ca2-4c5a-a0bb-2b7dd8374ed5.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"9e572f36-bb89-4fab-b3b1-e59d03c32f1f","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0027/1/RHUH-0027_1_segmentations.nii.gz","mesh_uri":"brain/9e572f36-bb89-4fab-b3b1-e59d03c32f1f.ply","source":"RHUH_GBM","study_id":"7388fa67-3ca2-4c5a-a0bb-2b7dd8374ed5"}]

api/segmentations/76fe4d95-b610-42dc-8276-4caa3700a292.json

@@ -0,0 +1 @@
+[{"confidence":null,"has_label_map":true,"has_mesh":true,"has_volume":true,"id":"220f0d84-1fdb-47f3-990b-ec35187caec9","label_definitions":{"ENHANCING":3,"NECROTIC_CORE":1,"OEDEMA":2},"label_map_uri":"RHUH-GBM/RHUH-0005/2/RHUH-0005_2_segmentations.nii.gz","mesh_uri":"brain/220f0d84-1fdb-47f3-990b-ec35187caec9.ply","source":"RHUH_GBM","study_id":"76fe4d95-b610-42dc-8276-4caa3700a292"}]