# DeepHeart: Semi-Supervised Sequence Learning for Cardiovascular Risk Prediction

arXiv: https://arxiv.org/abs/1802.02511 · HTML: https://mlsanity.com/paper/1802.02511/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Feb 2018.

## Built on by (most cited first, 12 of 12)

- [On the Actionability of Outcome Prediction](https://mlsanity.com/paper/2309.04470/) (2023)
- [Semi-Supervised Learning and Data Augmentation in Wearable-based Momentary Stress Detection in the Wild](https://mlsanity.com/paper/2202.12935/) (2022)
- [The Benefit of the Doubt: Uncertainty Aware Sensing for Edge Computing Platforms](https://mlsanity.com/paper/2102.05956/) (2021)
- [Predicting Survival Outcomes in the Presence of Unlabeled Data](https://mlsanity.com/paper/2210.13891/) (2022)
- [HealthSLM-Bench: Benchmarking Small Language Models for Mobile and Wearable Healthcare Monitoring](https://mlsanity.com/paper/2509.07260/) (2025)
- [Anticipatory Detection of Compulsive Body-focused Repetitive Behaviors with Wearables](https://mlsanity.com/paper/2106.10970/) (2021)
- [Predicting Intraoperative Hypoxemia with Hybrid Inference Sequence Autoencoder Networks](https://mlsanity.com/paper/2104.14756/) (2021)
- [Associating Healthcare Teamwork with Patient Outcomes for Predictive Analysis](https://mlsanity.com/paper/2512.03296/) (2025)
- [Improving performance of heart rate time series classification by grouping subjects](https://mlsanity.com/paper/2311.13285/) (2023)
- [QS4D: Quantization-aware training for efficient hardware deployment of structured state-space sequential models](https://mlsanity.com/paper/2507.06079/) (2025)
- [Learning Unified Representations from Heterogeneous Data for Robust Heart Rate Modeling](https://mlsanity.com/paper/2508.21785/) (2025)
- [Zero-Shot Heart Rate Variability Forecasting from Consumer Wearables Using Time Series Foundation Models](https://mlsanity.com/paper/2607.20027/) (2026)

## Abstract (arXiv)

We train and validate a semi-supervised, multi-task LSTM on 57,675 person-weeks of data from off-the-shelf wearable heart rate sensors, showing high accuracy at detecting multiple medical conditions, including diabetes (0.8451), high cholesterol (0.7441), high blood pressure (0.8086), and sleep apnea (0.8298). We compare two semi-supervised train- ing methods, semi-supervised sequence learning and heuristic pretraining, and show they outperform hand-engineered biomarkers from the medical literature. We believe our work suggests a new approach to patient risk stratification based on cardiovascular risk scores derived from popular wearables such as Fitbit, Apple Watch, or Android Wear.