Low-Power Hardware-Based Deep-Learning Diagnostics Support Case Study

TL;DR

This paper presents a low-power, embedded hardware system using a compact deep learning model for rapid, accurate microscopy diagnostics of infectious diseases at the point of care.

Contribution

It introduces a hardware-embedded microscopy diagnostic system utilizing a Squeeze-Net model and quantization to achieve lab-level accuracy with significantly reduced power consumption.

Findings

6x more power-efficient than CPU-based systems

Inference time of approximately 3 ms per sample

Achieves laboratory expert-level diagnostic accuracy

Abstract

Deep learning research has generated widespread interest leading to emergence of a large variety of technological innovations and applications. As significant proportion of deep learning research focuses on vision based applications, there exists a potential for using some of these techniques to enable low-power portable health-care diagnostic support solutions. In this paper, we propose an embedded-hardware-based implementation of microscopy diagnostic support system for PoC case study on: (a) Malaria in thick blood smears, (b) Tuberculosis in sputum samples, and (c) Intestinal parasite infection in stool samples. We use a Squeeze-Net based model to reduce the network size and computation time. We also utilize the Trained Quantization technique to further reduce memory footprint of the learned models. This enables microscopy-based detection of pathogens that classifies with laboratory expert level accuracy as a standalone embedded hardware platform. The proposed implementation is 6x more power-efficient compared to conventional CPU-based implementation and has an inference time of $\sim$ 3 ms/sample.