Current Projects

BioSignals

Biomedical data are often inherently temporal, adding another layer of complexity to already challenging modeling problems. We recently developed a self-regularizing, parameter-efficient 1D quantum convolution method for time-series learning [1], with strong performance demonstrated in brain–computer interfacing. More broadly, we work with EEG [2-4], ECG [5,6], and activity data [7] to model a range of health and behavioral outcomes. 

1. Orka NA, Haque E, Awal MA, Moni MA. Fully quanvolutional networks for time series classification. ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 2025. [code][datasets]
2. Haque E, Orka NA, Jannat M, Moni MA. Generalized EEG representation learning captures deviations linked to major depressive disorder. In Preparation. [dataset1][dataset2][dataset3]
3. Orka NA, Haque E, Jannat M, Awal MA, Moni MA. QuanvNeXt: A lightweight quantum-classical hybrid neural network for EEG-based depression detection. Under Review. [dataset1][dataset2]
4. Orka NA, Haque E, Rahman MT. One representation, many minds. In Preparation. [dataset]
5. Orka NA, Bulbul AAM, Paul GK, Mosaddek ASM, Moni MA. A foundation model for scanned ECGs. In Preparation. 
6. Bulbul AAM, Awal MA, Sandhu M, Paul GK, Mosaddek ASM, Orka NA, Moni MA. iLEDSNet: Interpretable lightweight federated learning framework for MI detection on low-powered devices. Under Review. [dataset1][dataset2]
7. Rahman MT, Orka NA, Khan A, Moni MA. Temporal graph learning of wearable actigraphy and sleep traces for modeling adolescent crystallized intelligence. Under Review. [dataset]

BioImaging

Our recent work has largely focused on neuroimaging. We work with structural and functional MRI—including both resting-state and task-based imaging, and often their combination—to model human cognitive and behavioral phenotypes [1-3] and investigate brain health and disease [4-5] while prioritizing model interpretability.

1. Rahman MT, Orka NA, Khan A, Moni MA. A geometric deep learning framework for modeling human intelligence from structural brain organization. IEEE Transactions on Cognitive and Developmental Systems. 2026. [dataset]
2. Rahman MT, Orka NA, Khan A, Moni MA. Explainable deep learning of resting-state functional connectomes reveals network biomarkers of adolescent intelligence. Under Review. [dataset]
3. Rahman MT, Orka NA, Khan A, Moni MA. MAC-Net: A multi-task deep learning framework for modeling cognitive function from task-based fMRI. Under Review. [dataset]
4. Rahman MT, Shuva TF, Orka NA, Khan A, Moni MA. Brain-behavior modeling from multimodal MRI using opportunistic pretraining and common-cohort adaptation. Under Review. [dataset]
5. Ali MA, Hossain MA, Rahman MT, Shuva TF, Almoyad MA, Orka NA, Bulbul AAM, Khan RT, Kaiser MS, Moni MA. RGNN3D: A hybrid radiomic graph neural network for 3D MRI glioma grading. Knowledge-Based Systems. 2026. [dataset1][dataset2]
6. Tonni SI, Sheakh MA, Tahosin MS, Hasan MZ, Shuva TF, Bhuiyan T, Almoyad MA, Orka NA, Rahman MT, Khan RT, Kaiser MS. A hybrid transfer learning framework for brain tumor diagnosis. Advanced Intelligent Systems. 2025. [dataset]