Md Shamim Ahmed

Deep networks for optical coherence tomography (OCT) classification usually learn directly from image appearance, which leaves coherent speckle, depth-dependent attenuation, point-spread-function blur, and refractive distortion entangled with pathology inside a single feature representation. This work asks whether making that separation architecturally explicit is a useful inductive bias, rather than whether it raises benchmark accuracy. PhysioSpeck-Net routes a shared stem representation through four branches aligned one-to-one with distinct OCT image-formation effects, recombines them through cross-branch attention and a physics-guided gate, and constrains them with auxiliary objectives derived from a seven-layer retinal simulator. The framework is evaluated on a 1000-image balanced test set and, without fine-tuning, on 1400 images from a second OCT source. Under a common training protocol, the model is competitive with substantially larger convolutional and transformer baselines while using 8.80 million parameters, but the more informative results concern reliability: predictive entropy separates errors from correct predictions with error-detection AUCs of 0.9847 and 0.9688, expected calibration error remains near 4–5% on both sets, and Grad-CAM++ perturbation analysis shows that attribution faithfulness is class-dependent rather than uniformly high. Distributional analysis further shows a residual gap between simulated and clinical images. Leave-one-branch-out and loss ablations produced consistent performance reductions, paired significance testing confirmed gains over most conventional baselines, and duplicate-control analysis found no exact or confirmed near-duplicate images across the evaluated datasets. These findings support physics-guided feature separation as a compact and uncertainty-aware strategy for retinal OCT classification under dataset shift.

Rotational energy harvesters are often constrained by narrow operating bandwidths and sensitivity to specific rotational regimes, limiting their effectiveness under variable-speed conditions. This work presents an orientation-adaptive dual-mode piezoelectric rotational energy harvester capable of broadband energy extraction across diverse rotational and vibration environments. The proposed design combines gravity-induced magnetic excitation at low rotational speeds with centripetal-force-induced nonlinear dynamics at higher rotational speeds, enabling passive transition between operating modes without active tuning. A coupled nonlinear electromechanical model is developed to investigate the interactions among gravitational forcing, magnetic coupling, centripetal loading and piezoelectric transduction. Numerical simulations reveal the transition from gravity-dominated mono-stable behaviour to broadband nonlinear operation as rotational speed increases. Experimental validation is conducted using representative vibration profiles from aerospace, automotive, civil infrastructure and industrial environments. The results demonstrate clear orientation-dependent performance, with the downward cantilever configuration achieving a maximum average power output of 57.8 μW under aerospace elevation excitation, whilst the upward configuration exhibits improved robustness under broadband random vibrations. The proposed orientation-adaptive framework provides a compact, stator-independent solution for broadband rotational energy harvesting under realistic operating conditions.

A broadband piezoelectric rotational energy harvester with upward and downward cantilevers was developed. The system captures energy from both directions of rotation using PVDF materials. It was tested under two aerospace vibration profiles with distinct frequency content. The downward cantilever produced a peak average power of 57.8 µW. The upward cantilever reached 33.5 µW under different excitation. Results confirm broadband rotational energy harvesting capability for aerospace vibration environments.