FunKANLite cuts model size and boosts speed for medical imaging
Hardware-Aware Functional Kolmogorov-Arnold Networks for Efficient Medical Image Enhancement and Segmentation
Computer Vision and Pattern RecognitionHardware ArchitectureMachine Learning
Summary
Medical image tools often need large and slow models that can't run well on small devices near patients. The authors showed how to shrink a powerful model called FunKAN into simpler, faster versions called FunKANLite-TR and FunKANLite-ST without losing much accuracy. These smaller models run efficiently on devices like Raspberry Pi and Jetson, using less energy and processing images faster. This could help bring advanced image cleanup and segmentation to clinics and remote settings.
What this means in practice
- •For medical device engineers: Make diagnostic imaging apps that run accurate artifact removal and segmentation efficiently on portable devices used in clinics.$Commercial implications: Enables design of compact medical imaging software for handheld or bedside devices with limited computing power.
- •For embedded system developers: Build accelerated medical image processing systems for edge hardware using the compressed FunKANLite models to reduce energy use and latency.
Authors
Mohammad Sadegh Sirjani
Abstract
Functional Kolmogorov-Arnold Networks (FunKAN) achieve state-of-the-art accuracy on MRI Gibbs artifact removal and anatomical segmentation, but their 11.6 M parameters and 8.7 GFLOPs are too large for edge medical devices. We present FunKANLite, a two-stage, hardware-aware compression of FunKAN for point-of-care use. FunKANLite-TR reduces the spatial prior and replaces the ResBlock offset predictor with a depthwise-separable block. It has 1.9x fewer parameters than FunKAN and no loss in accuracy. We then distill FunKANLite-TR into FunKANLite-ST, which lowers the Hermite basis rank, factorizes the spatial prior into a low-rank form, and halves the filter widths. FunKANLite-ST has 5.6x fewer parameters and 3.7x fewer GFLOPs than FunKAN. It stays within 1.4 percentage points IoU of FunKAN on BUSI, GlaS, and CVC-ClinicDB, and reaches 33.95 dB PSNR on IXI. On an NVIDIA Jetson Orin Nano and a Raspberry Pi 5, FunKANLite-ST reduces energy per inference by up to 68% and raises throughput by 2.9x.