MACE Lab

The Multi-omics and Artificial intelligence for Cardiac Electrophysiology (MACE) lab @shaan_khurshid @broad_institute @MGHHeartHealth @broad_ml4h

[Shaan Khurshid @shaan_khurshid]

Just because you exercise a lot doesn't mean you can ignore your cardiovascular risk score...

New work in @JACCJournals led by all-🌟 team @BradleyPetek @AjufoEzim, showing remarkably consistent performance of CV risk scores across measured PA levels

Don't ignore your score

@uk_biobank @broadinstitute @MGBResearchNews @MGHHeartHealth @MACE_Lab_AI @JSawallaGusehMD @patrick_ellinor

[Mitsuaki Sawano, MD, PhD, FACC @MitsuakiSawano]

🎙️ #JACCばらん Ep.63 | 電気的除細動後のAF再発、AIは予測できるのか？新規診断AFに対する電気的除細動は一般的な初期治療ですが、多くの患者でAFが再発します。今回の論文では、12誘導心電図1枚から再発リスクを予測するECG-AIの可能性に迫ります。

🎙️ ゲスト
臼田 圭佑 先生（Massachusetts General Hospital / Broad Institute of MIT and Harvard）

約60万人規模コホートから、新規AFに対して初回DCCVを受けた1293例を解析。中央値4.4年追跡で57%が再発しました。ECG-AIは、従来の臨床スコアよりもAF再発予測において優れた識別能を���しました。

👉 5年AF再発率：63%
👉 75歳以上：80.8% vs 65歳未満：48.9%
👉 AUROC：ECG-AI 0.771 vs clinical score 0.686 vs age/sex 0.655

🧠 Take-home
・AF再発リスクは年���で大きく異なる
・若年例では“一律追加治療”ではなく個別化戦略の可能性
・ECG-AIが除細動後管理を変えるかもしれない

Youtube: https://t.co/ThObNUJIFX
Video: https://t.co/BtYTCltC9y
Podcast: https://t.co/SJnOtdxmK6

論文：Electrocardiogram-Based Artificial Intelligence to Predict Recurrence After Cardioversion for Newly Diagnosed Atrial Fibrillation. J Am Coll Cardiol EP. https://t.co/w3lnSQ8tN9

@MitsuakiSawano @Nobu0129 @KenEjiri @SS_cardiol @SK2798 @ACCinTouch @JACCJournals @shaan_khurshid
#JACCばらん #AF #AtrialFibrillation #AI #ECGAI #Electrophysiology #JACC #CardioTwitter

[Shaan Khurshid @shaan_khurshid]

Can you predict future stroke using the ECG?

New @JACC_journals we introduce ECG2Stroke, a publicly available AI neural network model that estimates 10-year stroke risk with comparable performance to dedicated clinical risk models by encoding markers of atrial cardiomyopathy.

@broad_institute @broad_ml4h @MGHHeartHealth @MGBResearchNews @MACE_Lab_AI

[MACE Lab @MACE_Lab_AI]

New work @JACC_journals on SCA prediction using ECG-AI and EHR data!

[Shaan Khurshid @shaan_khurshid]

Sudden cardiac arrest (SCA) prevention is a paradox ➡️ we focus on a few people at highest risk (e.g., low EF), but most SCA occurs in people without known disease.

Here, in @JACC_journals #Advances we present an AI approach to population-based SCA prediction using ECG+EHR.

A fantastic cross-country collaboration @broadinstitute @broad_ml4h @UW @MGHHeartHealth @MGBResearchNews @MACE_Lab_AI

[Shaan Khurshid @shaan_khurshid]

Sudden cardiac arrest (SCA) prevention is a paradox ➡️ we focus on a few people at highest risk (e.g., low EF), but most SCA occurs in people without known disease.

Here, in @JACC_journals #Advances we present an AI approach to population-based SCA prediction using ECG+EHR.

A fantastic cross-country collaboration @broadinstitute @broad_ml4h @UW @MGHHeartHealth @MGBResearchNews @MACE_Lab_AI

[Shaan Khurshid @shaan_khurshid]

Amazing presentation by ⭐️ @SarnoffCardio and @MACE_Lab_AI fellow @axshahid on his impactful project investigating multimodal prediction of heart failure including ECG-based #AI, clinical factors, and genetics.

@broadinstitute @broad_ml4h @MGHHeartHealth @MGBResearchNews

[Shaan Khurshid @shaan_khurshid]

The ability to stratify risk of future #afib in NICM patients is increasingly important. "Keeping score" with a PRS may be an important tool in the kit.

Cool work @JACCJournals #EP by @saif_dababneh and colleagues. Check it out, along with our editorial with @marstonMD, links ⬇️⬇️

@MGHHeartHealth @MGBResearchNews @broadinstitute

[Shaan Khurshid @shaan_khurshid]

VITAL-AF was a negative RCT...1L ECG screening in primary care did not ⬆️ #afib diagnosis.

But IF we used ECG-AI to guide screening, could it have gone differently?

Our new analysis @JACCJournals led by ⭐ fellow @nat_ved sheds light into the potential future of #afib screening

[Shaan Khurshid @shaan_khurshid]

We know too much sitting is harmful. But just how bad is it??

New @JACCJournals #Advances work led by ⭐ @MGBResearchNews #EPeeps fellow @AjufoEzim quantifies effects of measured sedentary behavior on >700 future diseases

@uk_biobank @broadinstitute @MGHHeartHealth @MACE_Lab_AI

[Shaan Khurshid @shaan_khurshid]

Struggling a bit (like me) with the #ACC26 app? We'll make it easy for you - check out our @MACE_Lab_AI ACC content below!

@DrCardioCon @alalusim_md @ShinoKany @MGHHeartHealth @broadinstitute @broad_ml4h @patrick_ellinor

[Shaan Khurshid @shaan_khurshid]

Your patient develops transient #afib in the setting of physiologic stress ("stressor-associated AF").

Do they "really" have AF? Will it come back? How to treat?

NEW analysis @JAHA_AHA led by 🌟 fellow @MGHHeartHealth @JHaimovichMD shows ECG-based #AI is here to help! Thread ⬇️

[Shaan Khurshid @shaan_khurshid]

Re: exercise, does it matter for CV risk if you're a "morning person"? Or if you run in the evening after work? Or you're a midday gym rat?

New study @ESC_Journals #EJPC led by ⭐ #EPeeps fellow @AjufoEzim says NO - just go out there and do it!

@MGHHeartHealth @broadinstitute

[Shaan Khurshid @shaan_khurshid]

Great work in @JACCJournals #JACCEP led by 🌟 @MACE_Lab_AI post-doc Keisuke Usuda on ECG-AI to stratify risk of #afib recurrence after DCCV of a first-time episode

ECG-AI may prioritize post-DCCV pts for monitoring or upfront rhythm control @broad_institute @MGHHeartHealth

[Shaan Khurshid @shaan_khurshid]

Can ECG-based AI predict future heart failure?

NEW in @CircHF, we introduce ECG2HF: an ECG-AI model to predict 10-year HF, validated in 3 datasets

Led w @lucidtronix, Shino Kany, @JenHoCardiology, @patrick_ellinor, @broad_ml4h, @broad_institute, @MGHHeartHealth, @MACE_Lab_AI

[Pradeep Natarajan @pnatarajanmd]

Individuals w clonal hematopoiesis of indeterminate potential have nl labs, incl CBC, & are thus hard to identify beyond age. Innovative work from @shaan_khurshid @lucidtronix shows how routine ECGs can help prioritize those high yield for CHIP testing https://t.co/PmeIktt1Ym @JACCJournals #JACCAdvances

[Shaan Khurshid @shaan_khurshid]

NEW in @JACCJournals #JACCBTS, ECG-based AI age associates with CHIP beyond chronological age - ECG-AI may possess value in prioritizing individuals for CHIP screening/detection.

Co-led w/ @lucidtronix @MACE_Lab_AI @broadinstitute @MGHHeartHealth @patrick_ellinor @mchonig

[Shaan Khurshid @shaan_khurshid]

Honored to present our work using ECG-AI to predict adjudicated cardiac arrest as featured science #AHA25

@MGHHeartHealth @MGBResearchNews @broadinstitute @broad_ml4h

[JAMA Cardiology @JAMACardio]

Moderate to vigorous physical activity in guideline-recommended ranges was associated with lower risk of adverse cardiovascular outcomes and did not increase arrhythmia risk or pathologic cardiac remodeling in G+P− carriers.

#AHA25 @AHAScience

https://t.co/9AHYnpe90u

[Ezim Ajufo @AjufoEzim]

Pleased to share our work on the clinical (CV / CMP / VA) and subclinical associations of #wearable-measured physical activity in the growing gene-positive, phenotype-negative CMP variant carrier population presented at #AHA25 @MGBResearchNews @broadinstitute @JAMACardio

[Circulation @CircAHA]

#AHA25: VITAL-AF trial: Stressor-associated atrial fibrillation (AF) carries similar risks of stroke, bleeding, and death as primary AF ; yet patients are less likely to receive oral anticoa... https://t.co/KHusqqdQNr @JHaimovichMD @shaan_khurshid @MGHHeartHealth @MGBResearchNews