brandonb

28 ms·

601.
by brandonb 1y ago
This is really cool. Have you (or your colleagues) written anything about what you learned about ML for drug discovery?
602.
by brandonb 1y ago
This study followed 5,294 people for 20 years. People in the highest third of inflammation (hs-CRP > 5) had 32% higher chance of heart disease and 25% higher mortality from any cause. Inflammation was a predictor even when you take chole
603.
Inflammation predicts 25% higher mortality(thelancet.com)
24 points by brandonb 1y ago|4 comments
604.
by brandonb 1y ago
Sometimes people assume heart disease is a modern problem that didn't plague ancient societies. This is one study that shows it's not.
605.
34% of ancient mummies showed atherosclerosis in a CT scan(pubmed.ncbi.nlm.nih.gov)
4 points by brandonb 1y ago|2 comments
606.
by brandonb 1y ago
At least in the US, health insurers can’t raise rates or deny coverage based on pre-existing conditions. That was a major part of the Affordable Care Act.
607.
by brandonb 1y ago
What’s your source on Apple not using the neural network for VO2Max estimation? They’ve been using on-device neural networks for various biomarkers for several years now (even for seemingly simple metrics like heart rate). FWIW, the articl
608.
by brandonb 1y ago
FWIW, we're working on something similar (you wouldn't necessarily need to write R or Python). Feel free to email me at bmb@empirical.health and I can add you to a beta once we have it ready!
609.
by brandonb 1y ago
By 2018, the concept was definitely in the air since you had GPT-1 (2018) and BERT (2018). You could argue even Word2Vec (2013) had the core concept of pre-training on an unsupervised or self-supervised objective leading to performance on a
610.
by brandonb 1y ago
In the paper, they say they can't release the weights due to terms of consent with study participants (this is from the Apple Heart and Movement study).
611.
by brandonb 1y ago
Not yet -- this one is just a research study. Some of their previous research has made it into product features. For example, Apple Watch VO2Max (cardio fitness) is based on a deep neural network published in 2023: https://www.
612.
by brandonb 1y ago
I worked on one of the first wearable foundation models in 2018. The innovation of this 2025 paper from Apple is moving up to a higher level of abstraction: instead of training on raw sensor data (PPG, accelerometer), it trains on a timeser
613.
Beyond sensor data: Foundation models of behavioral data from wearables(arxiv.org)
230 points by brandonb 1y ago|54 comments
614.
At-home ECGs will detect early heart issues and save thousands, say doctors(theguardian.com)
4 points by brandonb 1y ago|0 comments
615.
by brandonb 1y ago
In medicine, there's often tradeoff between health and cost. GLP-1s are an obvious example, costing $10,000 per year but saving lives. This study shows an exception to that rule -- testing Lp(a), the strongest hereditary biomarker of h
616.
Study: Lp(a) testing would save both lives and dollars(sciencedirect.com)
1 points by brandonb 1y ago|1 comments
617.
Large neck size predicts the most common abnormal heart rhythm(msn.com)
2 points by brandonb 1y ago|0 comments
618.
Adaptive deep SVM for detecting early heart disease among cardiac patients(nature.com)
2 points by brandonb 1y ago|0 comments
619.
by brandonb 1y ago
This happens a lot in medicine due to believing observational data rather than randomized control trials. An entire book was written on Medical Reversal by the person who's now head of CBER at FDA: https://www.amazon.com&#x2
620.
Even mild Covid may leave blood vessels five years older(sciencedaily.com)
3 points by brandonb 1y ago|0 comments
621.
Simple blood cholesterol test may better predict coronary artery disease risk(medicalnewstoday.com)
1 points by brandonb 1y ago|0 comments
622.
UnitedHealthcare rolling out new benefit aimed at early cancer detection(fiercehealthcare.com)
3 points by brandonb 1y ago|0 comments
623.
Deep learning outperforms cardiologists in identifying structural heart disease(physicsworld.com)
3 points by brandonb 1y ago|0 comments
624.
OpenEvidence Scores 100 on US Medical Licensing Exam(openevidence.com)
6 points by brandonb 1y ago|0 comments
625.
A personal health large language model for sleep and fitness coaching(nature.com)
3 points by brandonb 1y ago|0 comments
626.
DeepHeart: Semi-Supervised Sequence Learning for Cardiovascular Risk Prediction [pdf](cdn.aaai.org)
3 points by brandonb 1y ago|0 comments
627.
Whoop pushes back on FDA over blood pressure feature(mobihealthnews.com)
3 points by brandonb 1y ago|0 comments
628.
by brandonb 1y ago
This paper is relevant to Long Covid, where many patients have reduced parasympathetic function and suffer post-exertional malaise (PEM). The fact that vagus nerve stimulation improves VO2Max in healthy volunteers is a nice proof of concept
629.
Distillation Scaling Laws(arxiv.org)
5 points by brandonb 1y ago|0 comments
630.
by brandonb 1y ago
These new guidelines make use of a new heart disease risk model, PREVENT. PREVENT predicts the risk of a heart attack, stroke, or other major cardiovascular event over the next 10 years or 30 years. Blood pressure is the strongest individua

More ›