Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
brandonb
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
42 ms
·
631.
▲
How Should We Evaluate Healthcare AI?
(outofpocket.health)
1 points
by
brandonb
1y ago
|
0 comments
632.
▲
New blood pressure guidelines recommend earlier treatment and skipping alcohol
(cnn.com)
4 points
by
brandonb
1y ago
|
1 comments
633.
▲
by
brandonb
1y ago
The FDA standard for blood oxygen sensing is within 6% absolute, 95% of the time. So variability in the sensing is pretty normal, and you want to look at long-term trends rather than individual measurements.
634.
▲
by
brandonb
1y ago
This Deep Learning textbook by Ian Goodfellow is a good intro to the fundamentals: https://www.deeplearningbook.org/ It does require some math, but has an appendix with a refresher. Then, read through this list from Ilya
635.
▲
by
brandonb
1y ago
Apple was in a patent dispute over this feature with Massimo. Their workaround is to calculate blood oxygen on the iPhone, using the sensors from Apple Watch. The Apple Watch hardware is otherwise the same. The back of the watch shines ligh
636.
▲
by
brandonb
1y ago
Apple was in a patent dispute over this feature with Massimo. Their workaround is to calculate blood oxygen on the iPhone, using the sensors from Apple Watch. The Apple Watch hardware is otherwise the same. The back of the watch shines ligh
637.
▲
Apple returns blood oxygen monitoring to the latest Apple Watches
(theverge.com)
32 points
by
brandonb
1y ago
|
8 comments
638.
▲
Smartwatches aren't confused about stress–but headlines and studies are
(wareable.substack.com)
3 points
by
brandonb
1y ago
|
0 comments
639.
▲
Samsung's 'Preventative Care' on the Galaxy Watch 8
(androidcentral.com)
2 points
by
brandonb
1y ago
|
0 comments
640.
▲
Am I Healthy? How to design a comprehensive health score
(empirical.health)
1 points
by
brandonb
1y ago
|
0 comments
641.
▲
Study Finds Smartwatches Aren't That Good at Measuring Stress
(gizmodo.com)
1 points
by
brandonb
1y ago
|
0 comments
642.
▲
Consequences and dangers of miscategorizing aging as a disease
(frontiersin.org)
3 points
by
brandonb
1y ago
|
0 comments
643.
▲
by
brandonb
1y ago
In a study of 200,000 people, apolipoprotein B (apoB) and lipoprotein(a) [Lp(a)] give a much earlier and more accurate read on heart disease risk than the usual cholesterol tests. People with the same cholesterol level could have totally di
644.
▲
New blood tests offer earlier and more accurate heart disease detection
(thebrighterside.news)
2 points
by
brandonb
1y ago
|
1 comments
645.
▲
When does the body start aging? The answer may surprise you
(usatoday.com)
4 points
by
brandonb
1y ago
|
0 comments
646.
▲
Poor sleep linked to 172 diseases, including dementia, Parkinson's, diabetes
(medicalnewstoday.com)
11 points
by
brandonb
1y ago
|
0 comments
647.
▲
by
brandonb
1y ago
If you’re in the top quartile of Lp(a), your overall risk of heart disease is about 114% higher. This is because each Lp(a) particle is about 6x more atherogenic than a "normal" LDL particle.
648.
▲
Lp(a) is another risk to your heart that's getting new attention
(washingtonpost.com)
2 points
by
brandonb
1y ago
|
1 comments
649.
▲
Interview with ChatGPT Agent Creators
(sequoiacap.com)
3 points
by
brandonb
1y ago
|
0 comments
650.
▲
by
brandonb
1y ago
Researchers adapted the PREVENT equations to calculate the age of your heart (published in JAMA Cardiology this week). The PREVENT equations are statistical models that predict your risk of a heart attack, stroke, or other cardiovascular ev
651.
▲
Most Americans' hearts are older than their chronological age, new tool says
(statnews.com)
3 points
by
brandonb
1y ago
|
1 comments
652.
▲
Deep Dive into Yann LeCun's JEPA
(rohitbandaru.github.io)
6 points
by
brandonb
1y ago
|
0 comments
653.
▲
Scientists created a new calculator for heart age
(washingtonpost.com)
4 points
by
brandonb
1y ago
|
0 comments
654.
▲
A Crisis Moment for Seed VC
(nextview.vc)
2 points
by
brandonb
1y ago
|
0 comments
655.
▲
Why VC-backed longevity startups are dying
(forbes.com)
2 points
by
brandonb
1y ago
|
0 comments
656.
▲
by
brandonb
1y ago
Co-Founder of the company (it's on my profile). Yes, you're right--of the biomarkers listed, only eGFR and HbA1c are standard in a routine health check.
657.
▲
by
brandonb
1y ago
That panel costs $190 (and includes 85 biomarkers). FWIW, similar bundles I've seen online are priced at $400-$500. All 85 biomarkers, if purchased separately, would cost a total of $1,490. ApoB, for example, usually costs $60 if done
658.
▲
by
brandonb
1y ago
I used to work with a cardiologist who joked that "we should just add statins to the water," so you have a point. The current guidelines for prescribing statins are based on your risk of a major cardiac event in the next 10 years
659.
▲
by
brandonb
1y ago
Eric Topol has a cool post on "machine eyes" -- using deep learning to identify subtle signals in ECGs and medical images that would be invisible to the naked eye: https://erictopol.substack.com/p/the-amazing-
660.
▲
by
brandonb
1y ago
There's a clinical trial for a new drug, lepodisiran, which lowered Lp(a) by 93.9%. Outside of that, if your Lp(a) is high, then the first strategy would be to choose a lower ApoB target than you would otherwise. (Every Lp(a) particle
More ›