Daniel Haehn PRO
Hi, I am a biomedical imaging and visualization researcher who investigates how computational methods can accelerate biological and medical research.
HPC Day 2022
Processing of
Biological Datasets at UMass Boston
MASSIVE
MACHINE
PSYCHOLOGY
DANIELHAEHN.COM
Biomedical Imaging
Chart Question Answering
We
Computers
GONZO
MONSTER
VAMPIRE
ZOMBIE
GARGOYLE
2x NVIDIA DGX A100
Thank you!
The Oregon-Massachusetts Mammography Database
OMAMA-DB
Daniel Haehn
Nurit Haspel
Marc Pomplun
Dan Simovici
Bill Lotter
Greg Sorensen
Ryan Zurrin
Pablo Bendiksen
Neha Goyal
Kendrick Khaev
Muskaan Manocha
Jeff Dusenberry
World's Largest Mammography Dataset
publicly available!
with labels!
World's Largest Mammography Dataset
we started with ~1 million images!
2000 x 2000 pixels
100 x 2000 x 2000 voxels
~8 MB
~500 MB
~8 TB
Intelligent Annotation Framework
Two Artificial Neural Networks..
..work together for Quality control
Discriminator finds error pattern of Classifier
Augmented Intelligence
Human + AI
import omama as O
pred = O.DeepSight.run(imgs)
+
=
Drained nodes...
...but it worked in the end!
Browser
Editor
Jupyter Notebook
SSHFS mounted Home Folder
TMUX on head node
SSH Tunnels (hop via head node to compute node)
Fly Retina Project
Jens Rister
Barry Stein
Daniel Haehn
Kunal Jain
Looking at the color-sensing photoreceptors
using an Electron Microscope
Kiran Balivada
Jeff Dusenberry
6144 x 4096 x 862 voxels
10 x 10 x 40 nm^3
How to transfer large datasets across institutions?
260+ GB
curl 'https://eastus1-mediap.svc.ms/transform/zip?cs=fFNQTw' \
-H 'authority: eastus1-mediap.svc.ms' \
-H 'accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9' \
-H 'accept-language: en-US,en;q=0.9,de;q=0.8' \
-H 'cache-control: max-age=0' \
-H 'content-type: application/x-www-form-urlencoded' \
-H 'origin: https://liveumb-my.sharepoint.com' \
-H 'sec-ch-ua: "Chromium";v="104", " Not A;Brand";v="99", "Google Chrome";v="104"' \
-H 'sec-ch-ua-mobile: ?0' \
-H 'sec-ch-ua-platform: "Linux"' \
-H 'sec-fetch-dest: iframe' \
-H 'sec-fetch-mode: navigate' \
-H 'sec-fetch-site: cross-site' \
-H 'upgrade-insecure-requests: 1' \
-H 'user-agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.0.0 Safari/537.36' \
--data-raw 'zipFileName=VSDataTileSet2.zip&guid=00c22792-c697-4def-b45e-ec0560018fe4&provider=spo&files=%7B%22items%22%3A%5B%7B%22name%22%3A%22VSDataTileSet2%22%2C%22size%22%3A0%2C%22docId%22%3A%22https%3A%2F%2Fliveumb-my.sharepoint.com%3A443%2F_api%2Fv2.0%2Fdrives%2Fb%21ZHO7Y3eH0EOqfeg_5VdUmSUuEmzqastElTPzN9v8GRKXPTMoP_jwQ44q0Km6lnis%2Fitems%2F01OSC7C5I2RBH7FQHPWFCY7B2FKLTC2YFF%3Fversion%3DPublished%26access_token%3DeyJ0eXAiOiJKV1QiLCJhbGciOiJub25lIn0.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.bTFzRnRSV3RaNE5MU2l2bUI3d3dOdW9tUm5yNmRJUTh3WUQvWUsybjhPaz0%22%2C%22isFolder%22%3Atrue%7D%5D%7D&oAuthToken=' \
--compressed
20 GB Transfer limit
Jeff Dusenberry
Director of Research Computing
at UMass Boston
SFTP
no shell access
MACHINE
PSYCHOLOGY
DANIELHAEHN.COM
Thank you!
By Daniel Haehn
Hi, I am a biomedical imaging and visualization researcher who investigates how computational methods can accelerate biological and medical research.