Butterfly raises $250M for portable full-body ultrasound scanner
venturebeat.com
venturebeat.com
Just months before hearing it, my pregnant wife and I were researching how hard ultrasound devices are to build, and were very surprised at how ultrasound can be used to detect/diagnosis an insane amount of common illnesses.
The hardware and physics behind ultrasound isn't nearly as technical or as advanced as you may think, and nowadays it is mostly a software issue. Combine the imaging with personal baselining and machine learning across similar body types fitted to 3D human model templates, you'd get Star Trek's medical scanner.
The technology here could be used to eliminate 99%+ waste and inefficiency in the medical industry, effectively bankrupting the current institutions and disrupting the industry with better, cheaper, healthier, less-annoying care. This is a future we should be fighting for.
Also, the article makes the common typo writing HIPPA instead of HIPAA. Otherwise great article.
Ouch. You'd think in 2018 spellcheck would have solved this for newspaper.
Tangential to this tangent: Google Docs spell check catches wrong use of "flower" vs "flour" based on context. Try this: open a new doc, type:
Flower
No error. Now add: Flower, eggs, sugar
Error on flower. But if you do: Flower, sugar, eggs
No error. Milk also works.ML? Or regex?
For reference, a bloom filter is an extremely space efficient, probabilistic data structure that acts a bit like a set and can answer the query 'does the bloom filter contain this entry'. The bloom filter will respond with either 'definitely not' or 'possibly/probably' depending on how it is tuned.
You could conceivably automatically populate this (still hardcoded) bloom filter by doing a brute force language corpus search for heavily correlated word pairs that have one or more of the two words having phonetically similar misspellings. E.g. 'sea' and 'breeze' would be heavily correlated. 'Sea' has a phonetically identical misspelling 'see'. You could then automatically add 'see + breeze' as a spurious pairing to the filter.
There was a period, maybe 1.5 years ago, during which text input prediction got substantially worse, then gradually improved. Along with the change came the ability for text input to change the estimated word after you entered the next word, using the combination of your entries to both words to estimate both simultaneously.
If they have language models which perform that task at a level worth pushing out to consumers they can do some smoothing of entries in a list.
At $2,000 this is a lot cheaper which will hopefully drive wider deployment.
Even in more developed countries, wider-scale use of imaging would be amazing: primary care being able to directly use ultrasound without referral to a specialist, and with smart image analysis, could cut diagnosis times significantly. In my case I waited 3 weeks for an US appointment on the NHS and that could have been avoided entirely.
"Computer vision algorithms ingest footage from the handset’s camera and detect the probe’s location in real time, directing users through augmented reality (AR) prompts precisely where to position it. (Butterfly calls it “Tele-Guidance”.)"
I know it's used for pregnancy, never saw other contexts for US so far, so would of course love to know if this will have more applications more easily available!
> The cloud storage service to which images are uploaded is AES 256-bit encrypted and SOC II certified
Um... Encrypting with aes 256 is easy. But will the keys be properly handled?
In a cardiac arrest patient, it's useful to know if the heart is actually moving or not (i.e. you may be seeing electrical signals on an EKG, but the heart isn't physically responding). It's also useful to look for a buildup of fluid around the heart, called cardiac tamponade (which squeezes the heart and prevents it from working effectively).
In trauma patients, a "focused assessment with sonography for trauma" or "FAST scan" is a quick way to check the common sources of major internal bleeding (which can provide useful guidance for treatment decisions in a very time sensitive setting).
It's also useful for "routine" stuff like the placement of IVs. Ultrasound can be used to positively identify and locate veins that are otherwise difficult to feel through the skin.