Update README.md
Browse files
README.md
CHANGED
|
@@ -9,4 +9,36 @@ tags:
|
|
| 9 |
- echocardiography
|
| 10 |
- deeplearning
|
| 11 |
- fasterrcnnresnet50
|
| 12 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
- echocardiography
|
| 10 |
- deeplearning
|
| 11 |
- fasterrcnnresnet50
|
| 12 |
+
---
|
| 13 |
+
# 🫀 Aortic Valve Detection Model (AoVdetector)
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
::contentReference[oaicite:0]{index=0}
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
## Description
|
| 20 |
+
|
| 21 |
+
This model detects the **aortic valve in echocardiogram images** from the following views:
|
| 22 |
+
- PLAX (Parasternal Long Axis)
|
| 23 |
+
- PSAX (Parasternal Short Axis)
|
| 24 |
+
- 3CH (Three-Chamber View)
|
| 25 |
+
|
| 26 |
+
The detection model is based on **Faster R-CNN with a ResNet-50 backbone**, pretrained on the **COCO dataset (COCO_V1)** and fine-tuned for cardiac ultrasound images.
|
| 27 |
+
|
| 28 |
+
## Model Details
|
| 29 |
+
|
| 30 |
+
- **Repository:** https://github.com/CardiovascularImagingVallHebron/AoVdetector
|
| 31 |
+
- **Developed in:** Python
|
| 32 |
+
- **Model type:** Object Detection
|
| 33 |
+
- **Architecture:** Faster R-CNN (ResNet-50)
|
| 34 |
+
|
| 35 |
+
## Model Performance
|
| 36 |
+
|
| 37 |
+
- **Label Match Accuracy:** 92%
|
| 38 |
+
- **View Type Accuracy:** 98%
|
| 39 |
+
|
| 40 |
+
## Contact
|
| 41 |
+
|
| 42 |
+
For any questions or inquiries, feel free to reach out to:
|
| 43 |
+
**Pere Lopez-Gutierrez**
|
| 44 |
+
📧 pere.lopez@vhir.org
|