coolwulf

Will change it to https as you suggest. Just need to get the ssl certificate.

For research purpose only. And please use anonymized / de-identified data. The server is not keeping any data.

I just sent you an email from my personal gmail.

I'm hosting on my own server.

I will write a blog post of details of the implementation. Will post later on HN. Thanks for the interests.

I built a native rust based hacker news client (open source on github):

http://fastHNreader.com

And I saw this post from it :)

Please run it in terminal and change it as executable file.

Experience lightning-fast performance with our native Hacker News reader built with Rust. Stay up-to-date with tech news, discussions, and insights without any lag or slowdowns.

That could be another tool :p

Pretty much. That is why this was a tiny personal project I did. The key for me is to be a le to track article and comments on HN for myself. I found it useful and that's why I decided to open source and make it available for everyone else.

Please use the thread title link

Thank you. I will work to make it available to more people and make this system more powerful and extremely fast.

Well I'm implementing a LLM based auto summary for each article and other features which focus on usability / readability / tracking for HN users. And this is more specialized and cannot be achieved by other tools.

Because I can customize a lot of functions related to Hacker News which usual RSS reader cannot

I built http://chess-notation.com, a free web app that turns handwritten chess scoresheets into PGN files you can instantly import into Lichess or Chess.com.

I'm a professor at UTSW Medical Center working on AI agents for digitizing handwritten medical records using Vision Transformers. I realized the same tech could solve another problem: messy, error-prone chess notation sheets from my son’s tournaments.

So I adapted the same model architecture — with custom tuning and an auto-fix layer powered by the PyChess PGN library — to build a tool that is more accurate and robust than any existing OCR solution for chess.

Key features:

Upload a photo of a handwritten chess scoresheet.

The AI extracts moves, validates legality, and corrects errors.

Play back the game on an interactive board.

Export PGN and import with one click to Lichess or Chess.com.

This came from a real need — we had a pile of paper notations, some half-legible from my son, and manual entry was painful. Now it’s seconds.

Would love feedback on the UX, accuracy, and how to improve it further. Open to collaborations, too!

My son has been playing chess with more than 5 years and accumulated a lot of notation sheets. I just happened to find out them scattered everywhere in his room. Since I'm leading some AI efforts to digitize some handwritten medical notes for healthcare applications, I decided to apply my expertise in converting his notation sheets into easy to manage pgn in his lichess account. So far this app works great on doing this with vision transformers. I decided to make this FREE for everyone so no need to hand type them yourself, especially for kids. Enjoy!

NeuralRad is the leading platform for Stereotactic Radiosurgery workflow, using deep-learning based auto-segmentation, auto-labeling to assist neurosurgeon and oncologists to make clinical decisions and build better treatment plan for each patient.

Great to meet another NERS student here on HN!

You might should have considered Nuclear Engineering department at University of Michigan - Ann Arbor, which is ranked No. 1 in the US (in contrary, MIT's nuclear engineering department is ranked No. 2). It's a better program and you might learn more things than what you could do at MIT.

I am talking in terms of my experience Treating multiple brain Mets patients. They have lesions as small as 0.01cc which we treat using either GammaKnife or CyberKnife with zero margin for CTV. This accuracy won't be achieved easily with tape measurements AFAIK.

Thank you!

Thanks for your comment. However I don't think finding lesions is trivia especially if you are dealing with a patient with multiple brain Metastasis of more than 10 or even 20, 30 in counts.

For SRS, accurate localization and contour, label are crucial for the better outcome.

Thank you for pointing out the typo. I am on mobile with fat fingers :p

At Stanford, the platform is used to help Physicians and Medical Physicists to contour Brain Mets. In addition, the platform never approves any dicom RT Structures. Physicians can use the generated RT structure to assist their SRS workflow.

I'm actually based in US.

To clarify, we have several research collaborations on-going in US clinics.

NeuralRad is being used during the clinic practice workflow at the moment. The platform currently is not FDA 510k cleared so we had to establish an IRB with the clinics which are building the platform into their clinical workflow.

Thank you for your comment and I would like to clarify a few things:

1) NeuralRad could be running locally inside your own clinic. Currently, Stanford hospital, UTSW, Jefferson and several other clinics in the US are using it in house. In this configuration, the data never transferred outside the clinic.

2) The whole platform is HIPAA-complaint. All Dicom data are anonymized before submitting to NeuralRad server for inference if using the NeuralRad cloud server.

3) NeuralRad server is within US, not in "other" country.

Thanks.

Also I don't think you could DM on HN (maybe I'm missing something). But if your side wants to collaborate or implement this into your clinical workflow (We do provide free local installation of the whole platform inside your clinic network.), you could email me at coolwulf@gmail.com or hao.jiang@neuralrad.com

Best,

Thanks for the reply.

1. We have already developed the module for the longitudinal comparisons of the follow-up MR images. The module is not yet published but you can see the follow-up tab on the top of the dashboard (now grayed)

2. There's an on-going research between us and Stanford hospital on the prediction of radiation necrosis vs. brain mets. Our publication could be found in last year's AAPM conference.

3. Sure that's a function could be added soon.

Thanks.

Hao

All patient data are anonymized locally inside user's browser before sending to NeuralRad server for deep learning inference.