Hugo Aerts

Big news: the thymus may be critical for adult health What if we missed something fundamental about the immune system? In two back-to-back papers in Nature, we show thymic health (AI on 30,000+ CT scans) links to longevity, disease risk & immunotherapy outcomes.

Thymus health is a predictor of lifelong well-being and immunotherapy effectiveness @Nature @Simon_Bernatz @aim_harvard @nbirkbak @AarhusUni @HugoAerts nature.com/articles/s4158… nature.com/articles/s4158… nature.com/articles/d4158…

Beyond all of this, in a 2nd @Nature paper by the same team thymus, health associated with improved response to cancer immunotherapy @HugoAerts @simonbernatz and colleagues nature.com/articles/s4158…

@EricTopol @Nature @simonbernatz Thank you for sharing this! These two Nature papers were a true team effort. Thank you for noting these striking findings was how strongly thymic health relates to immunotherapy outcomes and longevity. For more information, you can go here: x.com/HugoAerts/stat…

Big implications, folks, for today's new reports on the thymus gland and healthspan —Our thymus gland largely involutes after the teenage years, but AI used to determine its persistent level of health —Healthy adult thymus linked to an array of remarkably improved health outcomes in 2 cohorts —Cardiovascular incidence and mortality reduced (Figure); also less all-cause mortality —Pulmonary mortality, digestive disease, metabolic disease mortality reduced (Figure) —Reduced inflammation; reduced lung cancer; improved metabolic markers nature.com/articles/s4158…

👉 More to come - we’ll share deeper dives in the coming days

The second paper can be found here: nature.com/articles/s4158…

Big news: the thymus may be critical for adult health What if we missed something fundamental about the immune system? In two back-to-back papers in Nature, we show thymic health (AI on 30,000+ CT scans) links to longevity, disease risk & immunotherapy outcomes.

Using AI on CT data from 30,000+ people, we found that the health of the thymus varies strongly between people – and is linked to longevity, disease risk, and response to immunotherapy.

Did you know your face can reveal your biological age? Researchers at @MassGenBrigham developed FaceAge, an #AI algorithm that predicts biological age and survival outcomes for patients with cancer using a single photo. Patients with cancer appeared five years older than their actual age, and higher FaceAge predictions were linked to worse survival outcomes. Read more: massgeneralbrigham.org/en/about/newsr… #AI #CancerResearch #HealthTech

Did you know your face can reveal your biological age? @MGBResearchNews has developed FaceAge, an #AI algorithm that predicts biological age survival outcomes for patients with cancer using a single photo. Patients with cancer appeared five years older than their actual age, and higher FaceAge predictions were linked to worse survival outcomes. Learn more: spklr.io/6019CPed #CancerResearch #HealthTech

Cross-Care is our new #LLM benchmark for measuring safety/bias in the #healthcare domain. We find misalignments between real biomedical facts, pretraining data, and LLM knowledge representation. Read more (arxiv.org/abs/2405.05506) and explore our website (crosscare.net)

#TransferLearning + self-supervision to characterize glioma doi.org/10.1148/ryai.2… @HugoAerts @BrighamWomens @TYPoussaintMD #NeuroRad #ML #MachineLearning

Chest x-rays, the most frequently performed imaging test in medicine, have rich embedded information that humans can't see. Here #AI detects 10-year cardiovascular risk as well as the traditional (ASCVD) score! acpjournals.org/doi/10.7326/M2… @karljakobweiss @VineetRaghu2 @kparuchur

I'm excited to share a groundbreaking approach leveraging foundation models to develop high-performance radiomic biomarkers, even when confronted with limited training data. Check out our work in Nature Machine Intelligence: nature.com/articles/s4225…

1/ Check out this preprint that describes how we developed and validated a new #AI-based biomarker that we are calling “FaceAge”, which uses a simple face #photo to predict outcomes in #cancer patients! medrxiv.org/content/10.110… #oncology #AImedicine #FaceAge

Thanks for highlighting the studies - We are still in early days but so much potential. Looking forward to working towards responsible and equitable AI for healthcare as outlined in @atulbutte's wonderful editorial!

Is ChatGPT a reliable source for cancer treatment recommendations 🤖? Our study, out now in @JAMAOnc, found that the chatbot included guideline non-concordant recommendations for common cancer diagnoses 1/3 of the time: jamanetwork.com/journals/jamao… #ChatGPT #LLM #NLProc

Please check out our paper for more. ChatGPT output and scores are available at the project github github.com/AIM-Harvard/Ch…. Congrats to @shan23chen and thank you to wonderful coauthors! @MikeFooteMD @BenjaminKannMD @HugoAerts @Dr_RayMak @aim_harvard @Bos_CHIP

Also, check out the home page of this publication, where the deep learning model can be found: aim.hms.harvard.edu/cxr-lungrisk

Excited to share our new Nature Communications publication showing that deep learning algorithms can automatically predict the risk of lung disease mortality from a simple routine chest x-ray! nature.com/articles/s4146…