Difference between revisions of "2016 Summer Project Week/Needle Segmentation from MRI"
From NAMIC Wiki
| Line 23: | Line 23: | ||
<div style="width: 27%; float: left; padding-right: 3%;"> | <div style="width: 27%; float: left; padding-right: 3%;"> | ||
<h3>Approach, Plan</h3> | <h3>Approach, Plan</h3> | ||
| − | * | + | * First possible strategy: to develop a detection method based on Gaussian Mixture Models (GMM) supplied with different inputs (intensity, Frangi's filter, needle-tip cross-correlation and Hough transform). |
| + | * Second possible strategy: Convolutional Neural Network (CNN) over MRI patches. Two implementation can be investigated (pizel by pizel classification or entire patch classification). | ||
* Integrate algorithm to Slicer. | * Integrate algorithm to Slicer. | ||
</div> | </div> | ||
<div style="width: 27%; float: left; padding-right: 3%;"> | <div style="width: 27%; float: left; padding-right: 3%;"> | ||
<h3>Progress</h3> | <h3>Progress</h3> | ||
| − | * | + | * GMM strategy failed. |
| + | * CNN strategies has some accuracy problem but it can be the right way to get needles tips. In order to increase the accuracy some aspects can be investigated (patch size, dataset size, network architecture, data augmentation) | ||
</div> | </div> | ||
</div> | </div> | ||
Revision as of 07:40, 25 June 2016
Home < 2016 Summer Project Week < Needle Segmentation from MRI
Key Investigators
- Guillaume Pernelle, Imperial College
- Tina Kapur, BWH/HMS
- Paolo Zaffino, ImagEngLab
- Salvatore Scaramuzzino, ImagEngLab
- Maria Francesca Spadea, ImagEngLab
Project Description
NeedleFinder offers tools to segment needles from MRI/CT. It has mostly been tested on MRI from GYN brachytherapy cases. Currently the user must provide the needle tip for the segmentation to start. We aim to detect the tip automatically to make the needle segmentation fully automatic.
Objective
- Automatic detection of needle tips in MRI from GYN brachytherapy cases.
Approach, Plan
- First possible strategy: to develop a detection method based on Gaussian Mixture Models (GMM) supplied with different inputs (intensity, Frangi's filter, needle-tip cross-correlation and Hough transform).
- Second possible strategy: Convolutional Neural Network (CNN) over MRI patches. Two implementation can be investigated (pizel by pizel classification or entire patch classification).
- Integrate algorithm to Slicer.
Progress
- GMM strategy failed.
- CNN strategies has some accuracy problem but it can be the right way to get needles tips. In order to increase the accuracy some aspects can be investigated (patch size, dataset size, network architecture, data augmentation)
