hey
#boston
ppl! I’m co-organizing the
#c_house
hackathon in Cambridge, MA on july 12-14th!
-
^ part of a 48hr multi-house hackathon with hackers in sf, berlin, bangalore & berkeley... plus, we have prizes🏆
-
interested? shoot me a DM ;)
ML progress (day 5)
-Surprise, surprise- I read more reaction-diffusion brain tumor papers…
-> this has been a current deep dive of mine, really trying to understand the implications of this in possible neuro-onc clinical cases - w/a deep learning lens
Also, have been cooking
ML Progress (day 2)
- Mainly focused on overviewing the status quo within diffusion leveraging diffusion regression models for image-driven simulation - particularly in brain tumours via reaction-diffusion equations.
^looking forward to boosting my theoretical understanding of
Elsabeth Black's 🇨🇦 advice to her younger teammates before Worlds: “You have to enjoy the moment. You need to enjoy the process and not focus too much on the end result – just focus on your
#gymnastics
, what’s in your control and leave the rest out."
#WGC2022
#ARTWorlds2022
ML progress (day 11+12)
-
Had some fun testing out and putting to work what I’ve been working on learning within ML thus far.
Worked on training an brain tumor classification model from MRI images in PyTorch :)
So happy to transfer my knowledge consumption into actual
#SNO2023
So honoured to have met
@michelle_monje
earlier this month
@NeuroOnc
's 28th Annual Meeting!
I've been in awe of your research in pediatric neuro-onc. It was incredible getting to chat with w/you excited to see where the future of
@CancerNeurosci
heads.
i'm bringing back my daily ML progress updates as an effort to
#buildinpublic
as I build my technical depth within AI applications in medicine.
-
ML progress (day 1)
- talked thru my latest ai x neuro-onc project blockers w/
@nick_linck
+ got perspective on abt building back
hey!! 'm going to be starting daily progress for my technical skillsets in machine learning (ml), esp. in medicine :)
ML Progress (day 1)
- looked into some classification + diffusion regression model papers. I wasn't as productive as I would've liked, will keep building :)
@PopCulture2000s
@SamHaney12
I feel like it was pre-hipster (like before the tumblr era w/ chevron and moustaches, but after like the juicy tracksuit + Paris Hilton inspired looks era).
what excites me about applying to
@_TheResidency
? (esp. for the summer 24 house in Cambridge?)
-
I hope to get the chance to build in neuro-onc x ai and return to where my biotech dreams all started...
(yes, that picture is the aftermath of 13y/o me visiting Cambridge, MA)
(lack of) ML progress (day 5)
-
this is what happens when developers touch grass… in all seriousness I attempted to scooter 🛴 on route to build out my model but ended up wiping out lol.
-
learned my lesson the hard way lol
ML progress (day 9/10)
Explored fine tuning + LLMs, pretty cool stuff in the world of data development/processing etc.
Still have a bias towards computer vision stuff in medtech imo but excited to keep exploring and opening my mind to new topics :)
Let's get to building 👊🫡
ML Progress (day 3)
- Today I worked on a Brain Tumor MRI classification model + continiued my hyperfixation on reaction diffusion applications too.
Looking into possibly doing a deep dive into different model architectures. Open to resource suggestions / advice / tips etc :)
ML Progress (day 6/7)
Yesterday, I slacked off on my update - I spent about 6hrs looking more into some of the current medtech applications - specifically within neuroimaging + tumor cell density etc.
Aside from yesterday’s lack of post - today marks 1 week of my ML
#bip
#buildinpublic
(2/21) w/another
#tbt
to
#sno2023
@RadRudie
's (AWESOME) presentation on "Deep learning segmentation of infiltrative and enhancing cellular tumour in post-treatment glioblastoma patients with multi-shell diffusion MRI...
had some fun this week and gave my 1st undergraduate research presentation!
I was the only freshman out of the selected presentations ;) who knows maybe one day I’ll be defending my own PhD thesis 🤓😅
ML progress (day 3 + 4)
-
yesterday I wrote less code and “touched” grass per say LOL by attending one of
@manoliskellis
“salon” nights
@MITMuseum
which hosted
@JamieMetzl
+ focused on the future of what deploying cutting edge biotech might look like from an systems + social
ML progress (day 8)
Struggled with my mindset - I know not technically code-base related but definitely something I need to work on. In actual terms, I spent another day diving into deep learning applications for HGGs.
on friday, I volunteered + attended (
@MIT_CSAIL
) IIA+Symposium on AI Frontiers & Implications
@MIT
@MIT_SCC
-
from meeting world-class AI researchers like
@manoliskellis
to hearing
@iamwill
share how emerging tech will shape the music industry.
-
🧵here's my takeaways...
hey
#boston
interested in biotech, ai wearables, edutech or neurotech (+ more).
-
Join us this Friday (8/16) for the
@_TheResidency
project pitch + demo day! DM me for details :)
I am excited to document my skillset progression + project development in ai/ml in medicine.
at the end of this project stretch, I aim to apply these concepts + hopefully collaborate w/other builders in the MedTech space🤝
cc:
@_buildspace
@_nightsweekends
ML Progress (day 4)
- Continued fixing some bugs that I found in my MRI model, ended up detouring and taking a trip back to revisit some my foundations in python.
Also started the course [
@jeremyphoward
]
Starting to really enjoy these daily ‘check-ins’
as a proud kwk alum from last summer
@karliekloss
no joke, kode with klossy was the best experience + I never have felt so seen + comp sci since KWK - if you need any alums (like me) to help share kwk
@kodewithklossy
please lmkkkkk (it would make my entire year)
ML progress (day 2)
-
Just getting started - focused on revisiting some ML basics — specifically looked into reviewing what characteristics are req to ensure my model is the most robust and generalizable example that it can be :)
I found this cool computational pathology blog
Tumor talk time 💬🧠“Enhanced sensitivity of ECoG during awake craniotomies w/circular electrical grid”
@LenoxNeurosurg
-
Awesome installment of
#tumortalk
today by the Journal of neuro-oncology (
#buildinpublic
day 6/21)
reply ANY pic of the outside around you! 🌎🔎
i'll use a geospatial vision LLM to predict your location within twenty miles. 🕵️♂️
no landmarks needed. random trees/scenery works!
putting
@georainbolt
out of work
@geoguessr
#MedTwitter
, interested in neuro-onc x AI?
... My latest yt vid explores all the exciting advancements in AI/ML in non-invasive pLGG mutational status prediction!
--
stay tuned as I share wkly vids & build my technical depth 👇+ open to feedback/thoughts
Currently in the midst of attending my first (and hopefully not last) society of neuro-oncology meeting!
#SNO2023
-
all made possible by cure glioblastoma x Elsa’s touch 🦋💜🧠🎗️
@NeuroOnc
!!
last summer I had the privilege of working on neurovivid w
@EdGE_at_TERC
to inspire young neurodivergent students about
#neurotech
+ non invasive
#BCI
applications!
-
so rewarding to see this in action!!
#buildinpublic
(4/21) todays post is less technical and more so an reminder to focus on the little things in life! It’s so easy to get consumed w/work and day to day commitments
- if you can, go outside and recenter when things become too overwhelming.
If there’s one thing, I’ve learnt this week. It’s never to underestimate the good that exists in humanity!! As a young builder in
#STEM
, I’m so grateful for people like Tom. He honoured my enthusiasm for building in tech… (More in thread⬇️)
@manoliskellis
the beauty and insights found in the paper hold true for those who choose to take the time to explore!! science does not happen in isolation, continuous collaboration is vital and the notes + depth of this paper illustrates the importance of this :)
My latest neuro-onc x ai project looks at meningioma tumour segmentation using the
@3DSlicerApp
@aim_harvard
@harvardmed
We can better understand meningiomas & develop detailed 3D visualizations of imaging data using pre-existing MRI data.
video & article are coming soon
#buildinpublic
(7/21) on days when none of my code seems to run or I just feel unmotivated, I like to look back at moments like this ⬇️
*yes my
@ChooseMuse
headset was upside down in this video haha) - throwback to last December w/my ASD diagnostic project!
02/23 | my 1st video newsletter!! go check it out !! I talk abt my
@LenoxNeurosurg
virtual OR article + neurophysiological ASD biomarkers + how I'm learning the importance of "self-love" 🫶