Related Experiment Video
Updated: May 21, 2025
15:48
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
Published on: December 15, 2014
22.3K
AI-assisted diffuse correlation tomography for identifying breast cancer
Ruizhi Zhang1, Jianju Lu2, Wenqi Di1
1North University of China, State Key Laboratory of Dynamic Measurement Technology, Taiyuan, China.
Journal of Biomedical Optics
|May 19, 2025
Summary
Artificial intelligence (AI)-assisted diffuse correlation tomography (DCT) accurately distinguishes between benign and malignant breast lesions at 97% accuracy. This technique shows promise for early breast cancer diagnosis and treatment monitoring by assessing microvascular blood flow.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Oncology
Background:
- Diffuse correlation tomography (DCT) is an emerging noninvasive technique for measuring breast microvascular blood flow.
- Validation of DCT for differentiating benign and malignant breast lesions is limited due to instrumentation and algorithmic challenges.
Purpose of the Study:
- To develop and validate an AI-assisted DCT system for classifying breast lesions.
- To assess the performance of AI models in analyzing DCT-derived blood flow data for lesion characterization.
Main Methods:
- An AI-assisted DCT system was developed using a novel source-detector array and image reconstruction algorithm.
- DCT images were acquired from 61 female patients.
- AI models were trained using blood flow images as feature parameters or global inputs for lesion classification.
Main Results:
- DCT measurements demonstrated stability in healthy subjects through longitudinal monitoring.
- AI-assisted DCT achieved 97% accuracy in distinguishing between benign and malignant breast lesions.
- The system effectively identified functional abnormalities linked to metabolic demands in breast diseases.
Conclusions:
- AI-assisted DCT shows significant potential for the early diagnosis of breast cancer.
- This technique can aid in timely therapeutic assessment by detecting functional changes before significant tumor or vascular network development.
- The study highlights the capability of AI-assisted DCT in characterizing breast lesions noninvasively.
