Electrocardiogram (ECG) is one of the fundamental markers to detect different cardiovascular diseases (CVDs). Owing to the widespread availability of ECG sensors (single lead) as well as smartwatches with ECG recording capability, ECG classification using wearable devices to detect different CVDs has become a basic requirement for a smart healthcare ecosystem. We demonstrate that novel method of model compression with robust detection capability for CVDs from ECG signals can be aptly ported to the resource constrained micro-controller platform suitable for wearable devices while minimizing the performance loss. We employ knowledge distillation-based model compression approach where the baseline (teacher) deep neural network model is compressed to a TinyML (student) model using piece-wise linear approximation. Our proposed ECG TinyML has achieved ~156x compression factor to suit to the requirement of 100KB memory availability for model deployment on wearable devices including implantable loop recorder (ILR).
Schedule
Timezone: PST
Single Lead ECG Classification On Wearable and Implantable Devices
Arijit UKIL, Senior Scientist in TCS Research
Tata Consultancy Services, India
Electrocardiogram (ECG) is one of the fundamental markers to detect different cardiovascular diseases (CVDs). Owing to the widespread availability of ECG sensors (single lead) as well as smartwatches with ECG recording capability, ECG classification using wearable devices to detect different CVDs has become a basic requirement for a smart healthcare ecosystem. We demonstrate that novel method of model compression with robust detection capability for CVDs from ECG signals can be aptly ported to the resource constrained micro-controller platform suitable for wearable devices while minimizing the performance loss. We employ knowledge distillation-based model compression approach where the baseline (teacher) deep neural network model is compressed to a TinyML (student) model using piece-wise linear approximation. Our proposed ECG TinyML has achieved ~156x compression factor to suit to the requirement of 100KB memory availability for model deployment on wearable devices including implantable loop recorder (ILR).
Gitesh KULKARNI, Scientist, Embedded Devices, and Intelligent Systems
TCS Research, Bangalore, India
Electrocardiogram (ECG) is one of the fundamental markers to detect different cardiovascular diseases (CVDs). Owing to the widespread availability of ECG sensors (single lead) as well as smartwatches with ECG recording capability, ECG classification using wearable devices to detect different CVDs has become a basic requirement for a smart healthcare ecosystem. We demonstrate that novel method of model compression with robust detection capability for CVDs from ECG signals can be aptly ported to the resource constrained micro-controller platform suitable for wearable devices while minimizing the performance loss. We employ knowledge distillation-based model compression approach where the baseline (teacher) deep neural network model is compressed to a TinyML (student) model using piece-wise linear approximation. Our proposed ECG TinyML has achieved ~156x compression factor to suit to the requirement of 100KB memory availability for model deployment on wearable devices including implantable loop recorder (ILR).
Arijit UKIL, Senior Scientist in TCS Research
Tata Consultancy Services, India
Arijit Ukil is having more than 18 years of industrial research experience in different capacities. He is working as Senior Scientist in TCS Research, Tata Consultancy Services, India. He has published more than 50 research papers in distinguished conferences and journals. He has authored 4 book chapters. He has filed more than 40 patents with more than 30 grants in different geographies including Europe, China, USA, Japan, Australia, India. He holds Master’s in Engineering from Jadavpur University, Kolkata, India. He is a Senior Member, IEEE. He is steering committee member of HealthyIoT, 2016, 2017 and the General Chair in KDAH-CIKM-2018, 2019, 2020, 2021.
Gitesh KULKARNI, Scientist, Embedded Devices, and Intelligent Systems
TCS Research, Bangalore, India
Gitesh Kulkarni is working as a Scientist in TCS research. He is an accomplished designer of embedded systems and a Maker at heart. In one of his pioneering works, he was the lead designer of the world’s first industrial safety watch for the TATA group. He is a Master of Science in Electrical engineering from Colorado State University, Fort Collins, Colorado, USA. His research interests are at the cusp of edge computing, computer architectures, and sensing. He has several filed and granted patents in edge computing and wearable devices. Gitesh is a member of IEEE and ACM. Before joining TCS, Gitesh created embedded systems and systems products for leading technology companies with total experience of more than 19 years. He is a licensed Ham radio operator as well.
Schedule subject to change without notice.