Transfer Learning improves MI BCI models classification accuracy in Parkinson's disease patients

Aleksandar Miladinović; Miloš Ajčević; Pierpaolo Busan; Joanna Jarmolowska; Giulia Silveri; Susanna Mezzarobba; Piero Paolo Battaglini; Agostino Accardo

arXiv · 2010.15899

Transfer Learning improves MI BCI models classification accuracy in Parkinson's disease patients

Aleksandar Miladinović·Miloš Ajčević·Pierpaolo Busan·Joanna Jarmolowska·Giulia Silveri·Susanna Mezzarobba·Piero Paolo Battaglini·Agostino Accardo

View original source ↗Read source document ↗Export metadata ↓

Abstract

Motor-Imagery based BCI (MI-BCI) neurorehabilitation can improve locomotor ability and reduce the deficit symptoms in Parkinson's Disease patients. Advanced Motor-Imagery BCI methods are needed to overcome the accuracy and time-related MI BCI calibration challenges in such patients. In this study, we proposed a Multi-session FBCSP (msFBCSP) based on inter-session transfer learning and we investigated its performance compared to the single-session based FBSCP. The main result of this study is the significantly improved accuracy obtained by proposed msFBCSP compared to single-session FBCSP in PD patients (median 81.3%, range 41.2-100.0% vs median 61.1%, range 25.0-100.0%, respectively; p<0.001). In conclusion, this study proposes a transfer learning-based multi-session based FBCSP approach which allowed to significantly improve calibration accuracy in MI BCI performed on PD patients.

Explore related subjects

eess.SP cs.LG

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Aleksandar Miladinović, Miloš Ajčević, Pierpaolo Busan, Joanna Jarmolowska, Giulia Silveri, Susanna Mezzarobba, Piero Paolo Battaglini, Agostino Accardo. 2020-10-29. Transfer Learning improves MI BCI models classification accuracy in Parkinson's disease patients. https://arxiv.org/abs/2010.15899

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Multi-Carrier Rydberg Atomic Quantum Receivers with Enhanced Bandwidth Feature for Communication and Sensing

Rydberg atomic quantum receivers (RAQRs) have attracted significant attention in recent years due to their ultra-high sensitivity. Although capable of precisely detecting the amplitude and phase of weak signals, conventional RAQRs face inherent limitations in accurately receiving wideband RF signals, due to the discrete nature of atomic energy levels and their intrinsic instantaneous bandwidth constraints. These limitations hinder their direct application to multi-carrier communication and sensing. To address this issue, this paper proposes a multi-carrier Rydberg atomic quantum receiver (MC-RAQR) structure with five energy levels. We derive the amplitude and phase of the MC-RAQR and extract the baseband electrical signal for signal processing. In terms of multi-carrier communication and sensing, we analyze the channel capacity and accuracy of angle of arrival (AoA) and distance parameters, respectively. Numerical results validate our proposed model, showing that the MC-RAQR can achieve up to a bandwidth of 11.7 MHz, which is 17-fold larger than the conventional RAQRs. As a result, the channel capacity and the resolution for multi-target sensing are improved significantly. Specifically, the channel capacity of MC-RAQR is 110-fold and 2.8-fold larger than the classical RF receivers and RAQRs, respectively. For sensing performance, the RMSE of AoA estimation for MC-RAQR exhibits 7.6-fold reduction, compared with the conventional RAQRs. Furthermore, the RMSE of distance estimation is $634$-fold smaller than that of the root-CRB of classical RF receivers, showing the superior performance of the MC-RAQR. This demonstrates its compatibility with waveforms such as orthogonal frequency-division multiplexing (OFDM) and its significant advantages for multi-carrier signal reception.

eess.SP↗

arXiv2026

Channel Estimation in MIMO Systems Aided by Microwave Linear Analog Computers (MiLACs)

Microwave linear analog computers (MiLACs) have recently emerged as a promising solution for future gigantic multiple-input multiple-output (MIMO) systems, enabling beamforming with greatly reduced hardware and computational cost. However, channel estimation for MiLAC-aided systems remains an open problem. Conventional least squares (LS) and minimum mean square error (MMSE) estimation rely on intensive digital computation, which undermines the computational advantage offered by MiLACs. In this letter, we propose efficient LS and MMSE channel estimation schemes for MiLAC-aided MIMO systems. By designing the training precoder and combiner implemented by lossless and reciprocal MiLACs, the proposed schemes perform LS and MMSE estimation in the analog domain, leaving only simple digital scaling. They achieve identical estimation performance to their digital counterparts while significantly reducing computational complexity. Numerical results verify the effectiveness of the proposed schemes.

eess.SP↗

arXiv2026

Joint Subcarrier Phase Recovery for Nonlinearity Mitigation

We propose a low-complexity phase recovery scheme that simultaneously mitigates laser phase noise and fiber nonlinearity across several subcarriers. In a long single-span link with Raman amplification, the scheme achieves 0.9 dB gain with 99 real multiplications per complex symbol.

eess.SP↗