Research by Prof. Paul P. Sotiriadis's Group received the Best Student Paper Award @FAIEMA 2025

We are pleased to announce that the paper entitled "An Analog Low-Power Decision Tree Classifier Architecture for Parkinson’s disease prediction" received the Best Paper Award (PhD Symposium) in the 3rd International Conference on Frontiers of Artificial Intelligence, Ethics, and Multidisciplinary Applications (FAIEMA), which took place in Stavanger, Norway on 18 - 19 September, 2025.

The Award-Winning paper was co-authored by Vasileios Moustakas (PhD Student, ECE-NTUA and Researcher Archimedes/Athena RC), Konstantinos Cheliotis (Diploma Student, ECE-NTUA and Researcher Archimedes/Athena RC), Anna Mylona (Diploma Student, ECE- NTUA and Researcher Archimedes/Athena RC) , Dr. Vassilis Alimisis (Collaborating Researcher, ECE-NTUA and Postdoctoral Researcher Archimedes/Athena RC),and Paul P. Sotiriadis (Professor, ECE-NTUA; Lead Researcher, Archimedes/ Athena RC; IEEE, AIIA & AAIA Fellow ).

ABSTRACT: This work presents a low-power analog integrated decision tree classifier for real-time Parkinson’s disease prediction. To achieve exceptionally low power consumption of 833 nW and a classification speed of 640 K inferences per second, the proposed design utilizes sub-threshold analog circuitry, including a ReLU circuit, sigmoid function circuit, tunable current mirrors, and a current comparator. It is implemented using the TSMC 90nm CMOS process. When tested on a Parkinson’s disease dataset, the classifier consistently maintained high performance despite variations in process, voltage, and temperature, achieving an average accuracy of 93.65%. Monte Carlo analysis and Process-Voltage-Temperature corner testing confirm its robustness and high accuracy, making it well-suited for low-power biomedical engineering applications as a front-end processor.