profile
viewpoint
Ananya Jana ananyajana Rutgers University New Jersey

issue commentTai-Hsien/MeshSegNet

Flipping vtp file

Thank you @Tai-Hsien !

smoothumut

comment created time in 6 days

issue commentTai-Hsien/MeshSegNet

Flipping vtp file

Hi Tai Hsien, As easy_mesh_vtk is no longer maintained, could you please let us know how to flip a mesh using vedo?

smoothumut

comment created time in 7 days

issue commentTai-Hsien/MeshSegNet

Trouble in Data annotation

thanks @Tai-Hsien

hyzwj

comment created time in 10 days

issue commentTai-Hsien/MeshSegNet

Trouble in Data annotation

Hi @Tai-Hsien , Hope you are doing great. I have the entire mesh as well as the part meshes in form of stl file. could you guide on how to incorporate the labels in the entire/whole mesh file. The link https://github.com/Tai-Hsien/easy_mesh_vtk didn't work. How do you implement the mesh.set_cell_labels functions?

hyzwj

comment created time in 11 days

issue commentTai-Hsien/MeshSegNet

query related teeth_count

Thank you Tai Hsien for resolving my doubts

ananyajana

comment created time in a month

issue closedTai-Hsien/MeshSegNet

query related teeth_count

Hi Tai Hsien,

Thank you for your great work. I have a question regarding the tooth count in the upper jaw/lower jaw. I think fully grown human might have 16 teeth in each jaw. Does it not create problem for the segmentation, I see in the paper that only 14 of the teeth are numbered. Did you encounter this case where a jaw had 16 teeth and had to pick only 14 out of that as the segmentation is 15 class? Or did you filter out and take only the scans which has 14 teeth in each jaw?

Thanks, Ananya

closed time in a month

ananyajana

push eventananyajana/pre_hcc_classification

Ananya Jana

commit sha 58104853a7b9e16c9d4b4aa1b7c76b395a73bf47

code cleanup

view details

push time in a month

issue commentTai-Hsien/MeshSegNet

query related teeth_count

Thank you for your quick reply Tai Hsien. I have another question regarding the mesh decimation. I used the vedo to decimate the tooth meshes. I was expecting 10000 points, but I get more than 10k points Did you face this issue: this is my code snippet: s.decimate(N=10000, method='quadric') cnt_aftr = len(s.points())

for different meshes the cnt_aftr varies widely as can be seen from the log below: cnt_aftr: 10678 cnt_aftr: 10066 cnt_after: 10538 etc

Do you know how to deal with this situation? how to get exactly 10k points?

ananyajana

comment created time in a month

push eventananyajana/pre_hcc_classification

Ananya Jana

commit sha 0f8ad39453e36a3466e2b8642e1eb1523b4f7282

fixing minor error

view details

push time in a month

issue openedTai-Hsien/MeshSegNet

query related teeth_count

Hi Tai Hsien,

Thank you for your great work. I have a question regarding the tooth count in the upper jaw/lower jaw. I think fully grown human might have 16 teeth in each jaw. Does it not create problem for the segmentation, I see in the paper that only 14 of the teeth are numbered. Did you encounter this case where a jaw had 16 teeth and had to pick only 14 out of that as the segmentation is 15 class? Or did you filter out and take only the scans which has 14 teeth in each jaw?

Thanks, Ananya

created time in a month

startedTai-Hsien/MeshSegNet

started time in a month

push eventananyajana/pre_hcc_classification

Ananya Jana

commit sha 2413e4deba32cf02e033b356cfa3ed4043dbf8b9

code cleanup

view details

push time in a month

create barnchananyajana/pre_hcc_classification

branch : master

created branch time in a month

created repositoryananyajana/pre_hcc_classification

created time in a month

startedjanghyuncho/PiCIE

started time in a month

issue commentananyajana/HDenseUNet_pytorch

where is the 3d image processing?

@luckywyy - yes you are right

chen-yuu

comment created time in 2 months

issue openedLuchixiang/PCRL

query regarding downstream task data pre-processing

Hi Luchixiang,

Hope you are doing great. I have a question regarding the downstream task data pre-processing. Exactly which operation is performed on the 3D LUNA, 3D Brats and 3D LiTs volumes and the segmentation mask volumes? Are they resized or Are they resampled? which function do you use i.e. SimpleITK resample or skimage rescale or just array resizing? I assume that for the downstream tasks, all the volumes are resized to 64x64x32?

Thanks, Ananya

created time in 2 months

issue openedMrGiovanni/ModelsGenesis

query regarding the downstream task

Hi Zongwei,

This is a good work! I have a question regarding the downstream tasks. Do you resize or rescale or resample the LUNA and MRI 3D volumes and their segmentation masks for the downstream tasks?

Thanks, Ananya

created time in 2 months

startedHuaiChen-1994/LDLearning

started time in 2 months

issue commentLuchixiang/PCRL

regarding luna preprocessing

@Luchixiang thank you for your quick response, appreciate it!

ananyajana

comment created time in 2 months

startedLuchixiang/PCRL

started time in 2 months

issue openedLuchixiang/PCRL

regarding luna preprocessing

there is a check in the luna preprocessing if np.sum(d_img1) > config.lung_max * crop_deps1 * crop_rows1 * crop_cols1:

the size of d_img is input_rows, input_cols, input_depth which is 64x64x32 but it is being checked against a cube of size crop_deps1 * crop_rows1 * crop_cols1

these two could be different as there are options to choose i.e. col_size = [(64, 64, 32), (96, 96, 48), (96, 96, 96), (32, 32, 16), (112, 112, 64)]

so when the cube size is (32, 32, 16) there is a chance that the np.sum(d_img1) is always greater than config.lung_max * crop_deps1 * crop_rows1 * crop_cols1 and hence don't get selected

so the chances of selecting cubes of (32, 32,, 16) might actually be less?

created time in 2 months

more