Automated OCT-Based Tissue Screening System
This invention is a fully automated Optical Coherence Tomography (OCT)-based tissue screening system designed for high-throughput analysis of ex vivo tissue cultures. It provides rapid, label-free, and quantitative morphological readouts, enabling more efficient and reliable drug discovery and evaluation than traditional methods.
Description
The core technology is an integrated system that automates the process of tissue screening using OCT imaging and sophisticated deep learning algorithms. The system comprises a sample arm and a motorized platform that enables precise 3-D manipulation of the tissue sample. A processing apparatus, which may be implemented on a computer or other device, contains three key components: object detection, imaging depth optimization, and segmentation. The object detection component uses a neural network, such as one based on the Single Shot MultiBox Detector (SSD) deep learning algorithm, to determine the exact location of the tissue sample using an image from a camera. The imaging depth optimization component determines the optimal imaging depth for the OCT scan by analyzing the average intensity across B-scans at multiple depths. Finally, the segmentation component employs a hybrid deep learning architecture, combining a plurality of transformer neural network blocks for global feature extraction and a plurality of residual neural network blocks for local feature extraction, to delineate tissue regions in the 3-D B-scan volume. This integration of automated 3-D positioning with sophisticated deep learning ensures accurate and reproducible quantitative morphological readouts, such as thickness, area, and volume, overcoming limitations of manual or traditional histology methods.Applications
- Drug Discovery and Development: High-throughput screening of therapeutic agents using ex vivo tissue cultures.- Ophthalmological Diagnostics: Potential application in quantitative characterization of tissue responses to variables.
- Biotechnology Research: Applications requiring quantitative, reliable parameter readout in tissue analysis.
- Pre-Clinical Testing: Evaluation of drug effects and disease mechanisms using ex vivo imaging of tissue models.
- Medical Device Development: Integration into OCT systems for automated analysis in research or clinical settings.
Advantages
- High-Throughput and Automation: Provides a fully automated solution for rapid ex vivo tissue screening.- Quantitative and Reliable Readouts: Generates quantitative morphological readouts (thickness, area, volume) which are highly reliable and reproducible.
- Non-Invasive and Label-Free: Uses OCT, a non-invasive, high-resolution, and 3-D imaging technique, which does not require tissue processing steps like traditional histology.
- Enhanced Efficacy Screening: Enables superior screening of therapeutic agents, potential therapies, and optimal protocols.
- Advanced Deep Learning: Utilizes a custom, hybrid deep learning framework (Transformer and Residual Network blocks) for accurate tissue detection and segmentation in 3-D.