Intelligent diagnostic system for Cryptococcus: Switch-controllable nanocatcher and CNN-based artificial intelligence

Feng Gu

doi:10.1016/j.cej.2023.142674

Intelligent diagnostic system for Cryptococcus: Switch-controllable nanocatcher and CNN-based artificial intelligence

2023-06-01

0

PRE

AI

DOI：10.1016/j.cej.2023.142674

Original

Original request for help

Save

Abstract

En 中文

Each year, one million Cryptococcal infections occur among people living with HIV/AIDS, resulting in nearly 625,000 fatalities due to low efficiency of clinical diagnostic procedures and misdiagnosis caused by faulty subjective judgment of manually enumerated imaging-based infection assays. In order to improve the diagnostic efficiency of non-immunodeficient patients infected with Cryptococcus, the present study developed an intelligent diagnostic system combined with a switch-controllable nanocatcher (MNP@PNIPAMAA-CAS) with high capture performance for Cryptococcus and convolutional neural network (CNN)-based artificial intelligence (AI), which was utilized to analyze the collected pathological images for subsequent diagnosis. This system benefitted from the high adsorption efficiency of the MNP@PNIPAMAA-CAS and the objective data processing ability of AI. The system's detection limit was 1-4 cells /mL, the specificity was 97%, and the detection time was only 6 min. The intelligent diagnostic system will provide significant benefit to the diagnosis of clinical fungi due to its high sensitivity, high specificity, and superfast detection speed.

Keywords:

Cryptococcus

Artificial intelligence

MNP@PNIPAMAA-CAS

Diagnostics

YOLO

Intelligent diagnostic system for Cryptococcus: Switch-controllable nanocatcher and CNN-based artificial intelligence

Abstract

Journal

Organization