A Hierarchical Health Decision Support System Based on Wearable Medical Sensors and Machine Learning Ensembles

Web Published:

A Hierarchical Health Decision Support System based on Wearable Medical Sensors and Machine Learning Ensembles

Docket # 17-3310

Researchers at Princeton University in the Department of Electrical Engineering have developed a new hierarchical health decision support system (HDSS) that integrates health data from wearable medical sensors (WMSs) into clinical decision support systems (CDSSs) and can evaluate the health profile for anyone anywhere anytime.

HDSS is the first system that combines WMSs and CDSSs in a carefully structured information framework. The system has a multi-tier structure, starting with a WMS tier, backed by robust machine learning that enables diseases to be tracked individually by a disease diagnosis module. The system utilizes a novel procedure to generate disease diagnosis modules that can monitor various diseases in parallel. The feasibility of HDSS has been demonstrated through six accurate disease diagnosis modules aimed at four disease categories. HDSS can also be applied to other disease categories when datasets for those diseases become available. In addition, disease diagnosis modules for all 69,000 reported human diseases are estimated to need only 62 GB of storage space in the WMSs tier, which is practical even in terms of current technology. 

The invention can be used as a pervasive healthcare application on smartphone/personal computer/personal cloud for disease diagnosis and is compatible with commercially available WMSs such as Apple Watch, Microsoft band, Samsung watch, Samsung Band, etc.



•       Out-of-clinic patient data collector and interpreter

•       Personalized diagnosis/treatment/prescription system

•       Community/city/state/nation-level health trend analyzer

•       Health-related policy pricing



•       Bridge the clinical information gap

•       Accurate and practical to practice

•       Scalable disease-module based approach

•       Compatible with commercially available WMSs

•       Reduce preventable medical errors

•       Address the problem of unreliable patient recall of symptoms


Stage of development

We demonstrate the feasibility of the system through six disease modules aimed at four disease categories. We show that the system is scalable using five more disease categories. Just the WMS tier offers impressive diagnostic accuracies for five diseases: arrhythmia (86%), type-2 diabetes (78%), urinary bladder disorder (99%). Renal pelvis nephritis (94%), and hypothyroid (95%).



Hongxu Yin received his B.Eng degree in Electrical and Electronics Engineering from Nanyang Technological University, Singapore, in 2015. He is currently pursuing his Ph.D. in the Electrical Engineering Department at Princeton University. Hongxu Yin won the gold medal for best undergraduate thesis, and was on dean’s list for all academic years. His research interests include machine learning, artificial intelligence, wearable medical sensors, Internet-of- things, secure computing, and low-power analog and digital systems.


Niraj K. Jha, Professor of Electrical Engineering

Professor Niraj K. Jha completed his doctoral studies in Electrical Engineering at the University of Illinois at Urbana-Champaign in 1985.  He holds a M.S. in Electrical Engineering from the State University of New York at Stony Brook and a B.Tech. in Electronics and Electrical Communication Engineering from the Indian Institute of Technology.  He joined Princeton University in 1987, achieving the rank of Professor in 1998. Prof. Jha is a fellow of IEEE and the Association for Computing Machinery (ACM) and has served as the Editor-in-Chief of IEEE Transactions on VLSI Systems, and as an Associate Editor of several journals.  He has been the recipient of the AT&T Foundation Award, NEC Preceptorship Award for Research Excellence, the NCR Award for Teaching Excellence, the Princeton University Graduate Mentoring Award, and a Distinguished Alumnus Award from IIT, Kharagpur.  He has co-authored or co-edited five books, in addition to authoring or co-authoring 15 book chapters and more than 430 technical papers. He has won nine best paper awards and six best paper award nominations. In addition, his papers have been selected for “The Best of ICCAD: A collection of the best IEEE International Conference on Computer-Aided Design papers of the past 20 years,” by IEEE Micro Magazine as top picks from the 2005 and 2007 Computer Architecture conferences, and two were included among the most influential papers of the last 10 years at the IEEE Design Automation and Test in Europe Conference.  He holds 16 U.S. patents.

The research interests of the Jha lab include power-efficient chip multiprocessor (CMP) and multiprocessor system-on-chip (MPSoC) design, design algorithms and tools for FinFETs, three-dimensional integrated circuit (3D IC) design, embedded system analysis and design, field-programmable gate arrays (FPGAs), digital system testing, computer security, quantum circuit design, and energy-efficient buildings.


Intellectual Property Status

Patent protection is pending.

Industry collaborators are sought to further develop and commercialize this technology. A working prototype for the device is available along with a propriety lead small molecule.



Laurie Tzodikov

Princeton University Office of Technology Licensing • (609) 258-7256• tzodikov@princeton.edu

Sangeeta Bafna

Princeton University Office of Technology Licensing • (609) 258-5579• sbafna@princeton.edu


Patent Information:
For Information, Contact:
Laurie Tzodikov
Licensing Associates
Princeton University
Hongxu Yin
Niraj Jha