Innovative Data Solutions

Transforming multichannel time-series data into actionable insights through advanced machine learning techniques.

Data-Driven Insights

Transforming healthcare data into actionable insights through advanced machine learning techniques.

Dataset Construction Phase

Collect and preprocess multichannel time-series data for accurate healthcare analysis and anomaly detection.

A person wearing a smartwatch on their wrist. The watch screen displays a red heart and a waveform, suggesting a health or heart rate monitoring feature. The person's hand is adjusting the watch. In the background, a blue surgical face mask is partially visible.

IOMT-GAN Development

Build advanced architectures for generating synthetic healthcare data while ensuring high fidelity and reliability.

A person wearing a white shirt, tie, and surgical mask uses a stylus on a tablet device. They have a gray backpack and are in an interior space with shelves filled with books or similar items in the background.

“Application of GANs for ECG Synthesis & Anomaly Detection” (2022, lead author)

Introduced a time-series GAN for ECG data synthesis to augment arrhythmia detection training, boosting recall by 20%.

“Privacy-Preserving Medical Time-Series Data Synthesis” (2023, co-author, IEEE TDSC)

Developed a hybrid differential privacy and GAN approach, balancing privacy and utility. Attribute-inference evaluations showed a 40% reduction in disclosure risk.