Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Launch HN: Biodock (YC W21) – Better microscopy image analysis

news.ycombinator.com

Post reply on HN
Print view View original

11–19 of 19 posts

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#11

Post by mike210 »

svara wrote: post #9

Hi! We've been bootstrapping ariadne.ai [0] in the same space. Similar origin story, too - turning tools we built as grad students into products for the biomedical industry. Looks like we're thinking along somewhat convergent lines, but with some interesting differences as well. Feel free to reach out to my HN username at ariadne.ai if you'd like to chat. [0] https://ariadne.ai

Hey! Super cool. Will shoot you a message. Would love to chat and see what things we could talk about. Seems like you guys have built out some valuable assay analyses that we've heard people asking for, so congrats.

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#12

Post by andy99 »

mike210 wrote: post #10
andy99 wrote: post #8

Very cool. Do you have lots of unlabelled data, and if so, do you do any self-supervised pre-training? Have you ever considered releasing the backbone weights for the pre-trained models you have? No idea if this would be possible without giving up core IP, but I know I'm personally dying for an alternative to Imagenet (COCO for you?) trained on a big dataset. Are the images in your set diverse enough that you'd expec…

Hi Andy, We do have lots of unlabelled data, and we're also labeling a large portion of it. We do transfer learn for all of the models we're training, and the first backbone we use is partially self-supervised. Seems to help in overall performance, but it's not a huge effect in our experience. Maybe once we get a lot of models we can release the backbone weights for nuclear segmentation or at least a competition set…

Thanks for the reply! I'm actually working in a different domain but it seems to have a lot in common with yours - lots of unlabelled data, images that have nothing in common with Imagenet, in that they are all essentially of the same thing and we are looking for variations or features. We found that self-supervised pre-training (with various contrastive models) underperformed vs. starting with weights trained on Imagenet.

So a model that has been pretrained on something else, with enough variability to work as a feature extractor, but closer to the problem framing I mention would be of interest.

For state of the art computer vision stuff, most of the benchmarks use imagenet or similar datasets. But unfortunately I'm coming around to the realisation that those datasets are not representative of most real world problems (except general purpose scene / object recognition). So it becomes very challenging to pick out a potential technique to apply, and hope it transfers.

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#13

Post by mike210 »

andy99 wrote: post #12
mike210 wrote: post #10

Earlier quoted context omitted.

Hi Andy, We do have lots of unlabelled data, and we're also labeling a large portion of it. We do transfer learn for all of the models we're training, and the first backbone we use is partially self-supervised. Seems to help in overall performance, but it's not a huge effect in our experience. Maybe once we get a lot of models we can release the backbone weights for nuclear segmentation or at least a competition set…

Thanks for the reply! I'm actually working in a different domain but it seems to have a lot in common with yours - lots of unlabelled data, images that have nothing in common with Imagenet, in that they are all essentially of the same thing and we are looking for variations or features. We found that self-supervised pre-training (with various contrastive models) underperformed vs. starting with weights trained on Ima…

Interesting - it would be great to chat and find out more. Maybe there are things we can learn about each other. Can you shoot me an email at michael at biodock dot ai?

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#14

Post by mike210 »

itamarst wrote: post #3

To be fair, you can get really very far with non-machine-learning automated techniques (I did good-enough algorithms for an in-situ fluorescent gene sequencing image processing pipeline at one job). I suspect any form of good automated processing, regardless of whether it's AI, would be welcome to biologists. If you'd ever like to chat about the automation parts, I'd be interested in hearing how you're approaching it…

Also would love to chat if you ping me about the automation part - would love to get some feedback there. michael at biodock dot ai

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#16

Post by mike210 »

skwb wrote: post #15

Bioinformatics PhD student here: any plans to move to histology field? Or are you focusing more on the R&D space for now?

Hey! What kind of histology? If you mean doing AI analysis of histology images for research/etc. Then 100% - that's actually of the modules we're looking to do and if you work with those kind of images I would love for you to reach out at michael at biodock dot ai. We're totally free for academics, and a module like that would be free for academics as well.

If you mean clinical diagnostics histology - we're going to hold off a bit on that, although we're discussing some early partnerships there.

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#17

Post by gkk »

andy99 wrote: post #12
mike210 wrote: post #10

Earlier quoted context omitted.

Hi Andy, We do have lots of unlabelled data, and we're also labeling a large portion of it. We do transfer learn for all of the models we're training, and the first backbone we use is partially self-supervised. Seems to help in overall performance, but it's not a huge effect in our experience. Maybe once we get a lot of models we can release the backbone weights for nuclear segmentation or at least a competition set…

Thanks for the reply! I'm actually working in a different domain but it seems to have a lot in common with yours - lots of unlabelled data, images that have nothing in common with Imagenet, in that they are all essentially of the same thing and we are looking for variations or features. We found that self-supervised pre-training (with various contrastive models) underperformed vs. starting with weights trained on Ima…

Do I read it correctly that you're working with images with repeated pattern of instances of the same object on the image? I've been working with cell images, solving segmentation task - what biodock works on - and found interesting tricks to train models on vastly smaller number of labels than what you would think is possible with off-the-shelf models (e.g. Mask-RCNN or U-net + refinements).

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#18

Post by ArtWomb »

I think this is terrific you are trying to tackle this problem. The issue with using "50 year old" techniques is a detriment to progress. Interface looks beautiful as well.

My insights are probably things you've already considered. There are as many quantitative microscopy techniques as there are lab teams in the wild. Everyone builds their own correlations depending on their research dive of interest. So I think even if you just solved that problem: cloud storage of images combined with manual human analysis software. You could get some fraction of the million or so researchers across the globe. For many UI use cases, simplicity is desirable. But in this space, I think feature richness is key (even if 80% of users only need 20% of counting techniques).

The other obvious point is that image analysis can be similar across domains. From biology, to high energy physics. Even places you may not dream of: digital archeology. Keeping things general as possible may not be in your original mission or expertise, but I think it could help discover other pockets of interest.

Last point is around "AI enhanced microscopy". I've seen state of the art techniques like this termed "Medical Vision". Sort of a combination of "precision medicine" and "computer vision". As you try to break from the past, it may be something to ascribe to your services. Plus it sounds cool and futuristic ;)

Best of luck and really rooting for this to succeed. Definitely a problem that needs to be solved. And ripe with potential. Great job!

Re: Launch HN: Biodock (YC W21) – Better microscopy image analysis

#19

Post by mike210 »

ArtWomb wrote: post #18

I think this is terrific you are trying to tackle this problem. The issue with using "50 year old" techniques is a detriment to progress. Interface looks beautiful as well. My insights are probably things you've already considered. There are as many quantitative microscopy techniques as there are lab teams in the wild. Everyone builds their own correlations depending on their research dive of interest. So I think eve…

Hi! First, your last sentence made us really happy - thanks for the kind words.

As for your points: 1. Definitely. We are trying to get to feature parity ASAP with traditional methods so that our AI modules can really shine. We're building out a library right now so that scientists can put together their own pipelines.

2. We're keeping everything open at this point, and definitely see the possibility that other domains will need structured image analysis like the kind that we are building.

3. Interesting - might be a better way to phrase what we are working on.

Thanks! If you're interested in chatting, would love to get in touch. Reach out at michael at biodock dot ai.

Post reply on HN