Day Two
Thursday 18th April 2024
8:30 am Networking & Light Breakfast
8:50 am Chair’s Opening Remarks
Aligning Digital Pathology & Omics to Overcome Slow Adoption for Real- World Applications
9:00 am Panel Discussion: Emerging Frontiers: The Intersection of Digital Pathology & Omics to Pioneer Unprecedented Diagnostic Capabilities
Synopsis
- Exploring the synergy between digital pathology and omics in understanding diseases at a molecular level
- Discussing how integrating genomic data with digital pathology images enhances treatment decisions
- Addressing challenges and ethical considerations in managing large-scale omic and pathology dataset
10:00 am Overcoming Real-World Obstacles to Scaling AI & Digital Pathology
Synopsis
- Slow adoption despite promise
- Navigating regulatory requirements (CLIA and FDA)
- Promoting health equity and mitigating biases
- Generalizability gaps of AI
- Human-centered design considerations
10:30 am Morning Networking Break
Unveiling the Power of Digital Pathology to Expedite Preclinical Research for Improved Safety & Efficacy of Novel Drugs
11:30 am Understanding the Adoption, Utility & Need for Digitalization in Toxicologic Pathology & Nonclinical Safety Assessment
Synopsis
- Function and role of toxicologic pathologists in the drug development process, specifically in safety assessment and GLP environments
- Current uses of digital pathology in toxicology pathology, barriers to increased adoption, and strategies to increase adoption
- Near-future digital pathology tools that will improve workflows and aid decision-making
12:00 pm High-Throughput Analysis in Preclinical Digital Pathology: Accelerating Research & Discovery
Synopsis
- Explore how high-throughput digital pathology platforms facilitate rapid and comprehensive analysis of tissue samples in preclinical studies
- Discuss the impact of high-throughput techniques on data acquisition, processing, and interpretation in preclinical research
- Harnessing digital pathology for quantification via tissue sample image analysis
12:30 pm Overcoming Challenges in Preclinical Toxicologic Pathology
Synopsis
- Understanding and overcoming challenges in investigative toxicology
- Adapting analytical tools for effective safety outcome assessment in diverse animal models
1:00 pm Lunch Break & Networking
Leveraging AI-Enabled Digital Pathology to Accurately Stratify & Select Patients for Improved Drug Response Rate
2:00 pm Harnessing Deep Learning Models in Digital Pathology for Patient Stratification in Immunotherapy
Synopsis
- How can computational analysis of pathologic data translate to accurately stratified patients?
- Integrating multi-omics to determine precision therapies
- Exploring the challenges regarding implementing AI for stratification
2:30 pm Reproducible and Quality Data for Harnessing AI-enhanced Digital Pathology to Develop Companion Diagnostics to Enhance Drug Response Rate
Synopsis
- How is AI harnessing pathological data for the identification of novel biomarkers to accurately predict drug response rate?
- How can different algorithms permit accurate patient selection based on pathological data to determine responders versus non-responders?
- How can multiple data sources be accurately validated to ensure accurate Companion Diagnostics generation?
3:00 pm Afternoon Break & Networking
Unveiling the Path to Better Pathological Data Handling to Improve Image Quality Whilst Minimizing Cost & Variabilities
3:30 pm Delving into the Essential Validation Steps to Uphold Exceptional Image Quality for Improved Pathological Reliability
Synopsis
- What is needed to ensure scanner technology is validated for high-quality and consistent imaging?
- Ensuring consistent validation of tissue samples and slides staining quality to maintain high accuracy rates
- Considerations for training and evaluating the performance of quantitative image analysis algorithms
4:00 pm Unsupervised Learning Approach for Detailed Cell Type Mapping in H&EStained Images
Synopsis
- Generating datasets that accurately and exhaustively delineate cells in H&E images is challenging and time-consuming, particularly for achieving comprehensive coverage of the entire cellulome
- We propose a machine learning framework that utilizes state-of-the-art cell segmentation models in conjunction with a modified self-supervised learning model to capture intricate embeddings of individual cells in H&E whole slide images
- Our framework is designed to cluster similar cell types, enabling their differentiation in an automated manner. This paves the way for comprehensive, automated classification of cell types in H&E-stained images