Upload folder using huggingface_hub

README.md

@@ -20,33 +20,72 @@ library_name: promptm-unet
 [![GitHub Repository](https://img.shields.io/badge/GitHub-Repository-black?logo=github)](https://github.com/kiuyha/PromptM-UNet)
 [![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
 
-**PromptM-UNet** is an ultra-lightweight, multi-modal 3D medical image segmentation framework that integrates **Residual Vision Mamba (SSM)** with **multi-stage language conditioning**. It enables accurate, prompt-guided volumetric CT segmentation while maintaining linear $O(N)$ computational complexity and a minimal memory footprint feasible for standard consumer GPUs.
+**PromptM-UNet** is an ultra-lightweight, multi-modal 3D medical image segmentation framework that integrates **Residual Vision Mamba (SSM)** with **multi-stage language conditioning**. It enables accurate, prompt-guided volumetric CT segmentation while maintaining linear \(O(N)\) computational complexity and a minimal memory footprint feasible for standard consumer GPUs.
 
 * **GitHub Repository**: [https://github.com/kiuyha/PromptM-UNet](https://github.com/kiuyha/PromptM-UNet)
-* **Target Organ (v1.0)**: Spleen (TotalSegmentator CT Cohort)
+* **Target Organ**: Spleen (TotalSegmentator CT Cohort)
 
 ---
 
 ## 🌟 Key Features
 
-* **Linear $O(N)$ Volumetric Modeling**: Built on 3D Residual Vision Mamba (RVM) blocks from LightM-UNet, providing global receptive field modeling without quadratic $O(N^2)$ Transformer memory explosion.
-* **Ultra-Lightweight Footprint**: Active visual backbone is only **~1.87 Million parameters** ($16\times$ smaller than standard 3D nnU-Net, $100\times$ smaller than SegVol).
+* **Linear \(O(N)\) Volumetric Modeling**: Built on 3D Residual Vision Mamba (RVM) blocks from LightM-UNet, providing global receptive field modeling without quadratic \(O(N^2)\) Transformer memory explosion.
+* **Ultra-Lightweight Footprint**: Active visual backbone is only **~1.87 Million parameters** (\(16\times\) smaller than standard 3D nnU-Net, \(100\times\) smaller than SegVol).
 * **Zero-Overhead Language Conditioning**: Decoupled text embedding caching ensures frozen language models (CLIP, BioBERT, Sentence-BERT) consume 0 extra GPU memory during training/inference.
 * **Dual-Resolution Spatial Zoom**: SegVol-inspired dual-scale pipeline (3mm global context + 1.5mm high-resolution target crop).
-* **Clinical Performance**: Achieves **$0.862+$ Dice Similarity Coefficient (DSC)** and **$0.787+$ Normalized Surface Distance (NSD)** on TotalSegmentator CT scans.
+* **Clinical Performance**: Achieves **\(0.862+\) Dice Similarity Coefficient (DSC)** and **\(0.792+\) Normalized Surface Distance (NSD)** on TotalSegmentator CT scans with under **4.3 GB** validation VRAM.
 
 ---
 
-## 📊 Benchmark Results
+## 📊 Benchmark & Ablation Results
 
-Evaluated on the official TotalSegmentator spleen validation/test cohort across multi-tier clinical text prompts ($N$: Name, $NS$: Name+Synonym, $NL$: Name+Location, $NSL$: Name+Synonym+Location):
+The following results are evaluated on the official TotalSegmentator spleen validation cohort across 30 training epochs (from [`notebooks/ablation_promptm_unet.ipynb`](notebooks/ablation_promptm_unet.ipynb)):
 
-| Configuration | Parameters | Peak DSC | Peak NSD | Inference VRAM |
-| :--- | :---: | :---: | :---: | :---: |
-| **PromptM-UNet (Bottleneck Fusion, 50:50 BCE)** | **1.87M** | **`0.8622`** | **`0.7928`** | **< 4.3 GB** |
-| **PromptM-UNet (All-Stage Fusion, 50:50 BCE)** | **1.91M** | **`0.8619`** | **`0.7871`** | **< 4.3 GB** |
-| **LightM-UNet Baseline (Vision-Only)** | 1.87M | `0.7820` | `0.6950` | < 4.3 GB |
-| **SegVol (Foundation Model)** | 181M | `0.8590` | `0.7710` | > 24 GB |
+### 1. Fusion Strategy Ablation
+*Evaluated using Multiplication operation, CLIP text encoder, and 50:50 BCE supervision:*
+
+| Fusion Strategy | Peak Epoch | Peak DSC | Peak NSD | Final DSC (Epoch 30) | Final NSD (Epoch 30) | Peak Val VRAM |
+| :--- | :---: | :---: | :---: | :---: | :---: | :---: |
+| **Early Fusion** | **19** | **`0.8622`** | **`0.7928`** | `0.8196` | `0.7514` | **< 4.3 GB** |
+| **All-Stage Fusion** | 24 | `0.8619` | `0.7871` | `0.7737` | `0.6848` | < 4.3 GB |
+| **Late Fusion** | 12 | `0.8576` | `0.7675` | `0.7347` | `0.6095` | < 4.3 GB |
+
+### 2. Loss Function Ablation
+*Evaluated across compound loss formulations:*
+
+| Loss Configuration | Peak Epoch | Peak DSC | Peak NSD | Final DSC (Epoch 30) | Final NSD (Epoch 30) |
+| :--- | :---: | :---: | :---: | :---: | :---: |
+| **Focal Loss (50:50)** | **20** | **`0.8627`** | `0.7827` | `0.8045` | `0.7195` |
+| **Standard BCE (50:50)** | 24 | `0.8619` | **`0.7871`** | `0.7737` | `0.6848` |
+| **Dynamic Loss (80:20 \(\to\) 65:35)** | 27 | `0.8537` | `0.7668` | `0.8280` | `0.7299` |
+| **Focal Loss (80:20)** | 20 | `0.7827` | `0.6821` | `0.7358` | `0.6393` |
+
+### 3. Fusion Operation Ablation
+*Evaluated using 50:50 BCE supervision:*
+
+| Operation | Peak Epoch | Peak DSC | Peak NSD | Final DSC (Epoch 30) | Final NSD (Epoch 30) |
+| :--- | :---: | :---: | :---: | :---: | :---: |
+| **Multiplication** | **24** | **`0.8619`** | **`0.7871`** | `0.7737` | `0.6848` |
+| **Concatenation** | 26 | `0.8381` | `0.7647` | `0.8120` | `0.7387` |
+| **Hybrid** | 28 | `0.8152` | `0.7301` | `0.7962` | `0.7041` |
+
+### 4. Text Encoder Ablation
+*Evaluated with Early Fusion:*
+
+| Text Encoder | Peak Epoch | Peak DSC | Peak NSD | Final DSC (Epoch 30) | Final NSD (Epoch 30) |
+| :--- | :---: | :---: | :---: | :---: | :---: |
+| **CLIP** | **19** | **`0.8622`** | **`0.7928`** | `0.8196` | `0.7514` |
+| **BioBERT** | 19 | `0.8466` | `0.7584` | `0.7968` | `0.7148` |
+
+### 5. Multi-Tier Clinical Prompt Robustness
+*Evaluated on Best Early Fusion + CLIP checkpoint across prompt complexity tiers:*
+
+| Prompt Tier | Clinical Description | Peak DSC |
+| :--- | :--- | :---: |
+| **Tier \(N\)** | Organ Name (*e.g., "spleen"*) | `0.8620` |
+| **Tier \(NS\)** | Name + Synonym (*e.g., "spleen, lien"*) | `0.8621` |
+| **Tier \(NL\)** | Name + Location (*e.g., "spleen in left upper quadrant"*) | **`0.8624`** |
+| **Tier \(NSL\)** | Name + Synonym + Location (*e.g., "spleen, lien in upper left abdomen"*) | `0.8622` |
 
 ---