Analog In-Memory Computing with Uncertainty Quantification for Efficient Edge-based Medical Imaging Segmentation
Preprint
Abstract
This work investigates the role of the emerging Analog In-memory computing (AIMC) paradigm in enabling Medical AI analysis and improving the certainty of these models at the edge. It contrasts AIMC's efficiency with traditional digital computing's limitations in power, speed, and scalability. Our comprehensive evaluation focuses on brain tumor analysis, spleen segmentation, and nuclei detection. The study highlights the superior robustness of isotropic architectures, which exhibit a minimal accuracy drop (0.04) in analog-aware training, compared to significant drops (up to 0.15) in pyramidal structures. Additionally, the paper emphasizes IMC's effective data pipelining, reducing latency and increasing throughput as well as the exploitation of inherent noise within AIMC, strategically harnessed to augment model certainty.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Bringing uncertainty quantification to the extreme-edge with memristor-based Bayesian neural networksNature Communications, 2023
- A 64-core mixed-signal in-memory compute chip based on phase-change memory for deep neural network inferenceNature Electronics, 2023
- A compute-in-memory chip based on resistive random-access memoryNature, 2022
- Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images2022
- A 40-nm, 2M-Cell, 8b-Precision, Hybrid SLC-MLC PCM Computing-in-Memory Macro with 20.5 - 65.0TOPS/W for Tiny-Al Edge Devices2022
- nnU-Net: a self-configuring method for deep learning-based biomedical image segmentationNature Methods, 2020
- Monolithically Integrated RRAM- and CMOS-Based In-Memory Computing Optimizations for Efficient Deep LearningIEEE Micro, 2019
- Nucleus segmentation across imaging experiments: the 2018 Data Science BowlNature Methods, 2019
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015