Human Breast Cancer-Related Patterns Revealed by Artificial Intelligence

Shi, Derek

Permalink

https://hdl.handle.net/2142/114285 Copy

Description

Title

Human Breast Cancer-Related Patterns Revealed by Artificial Intelligence

Author(s)

Shi, Derek

Issue Date

2022

Keyword(s)

Electrical and Computer Engineering

Abstract

With the latest advancements in optical bioimaging, rich structural and functional information is captured from biological samples, which calls for powerful computational tools to identify patterns and uncover relationships between optical characteristics and various biological conditions. Artificial Intelligence (AI) excels at extracting patterns and insights from large volumes of data that can hardly be perceived by the human eye and brain. Here we developed a weakly-supervised deep learning framework for human breast cancer-related optical biomarker discovery based on simultaneous label-free autofluorescence-multiharmonic (SLAM) microscopy. The latent feature space of cancer-related SLAM patch images found by the model is visualized in 2-D using t-distributed stochastic neighbor embedding (t-SNE). The AI model revealed both known cancer-related patterns (shown in the left column, e.g., tumor cells, tumor-associated fibrillar collagen) and non-obvious patterns (shown in the right column, e.g., NADH-rich extracellular vesicles), which facilitate new insights into tumor micro-environment and field cancerization and inspire new hypotheses for cancer biomarker discovery.

Type of Resource

• Text
• Image

Language

eng

Permalink

http://hdl.handle.net/2142/114285

Copyright and License Information

Copyright 2022 Derek Shi

Owning Collections