Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: I built Sinkove, a Midjourney for medical images (sinkove.com)
1 point by pedro_sanchez 3 months ago | hide | past | favorite
Hi HN,

Today, we are launching Sinkove with a foundation GenAI model to generate biomedical images from text!

After years of working in AI for medical imaging, I’ve consistently struggled with acquiring large, unbiased, and diverse datasets. The need for better, more accessible data has driven me to build Sinkove to create AI-generated medical images.

Our technology is based on latent diffusion models (LDM), similar to Stable Diffusion. Today, we are launching by deploying two state-of-the-art open-source models released under the Apache 2 license: Chest X-ray and Brain MRI.

We have some partnerships lined up and we will be releasing other amazing models soon. There are many applications of synthetic medical images, what is your favourite? What do you wanna use it for? Which features would you like to see? I am looking forward to hearing your feedback.

PS: Sinkove means ‘Synthetic life’ inspired by a native Brazilian language.

Pedro + Sinkove team!




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: