• Tackling the issue of underutilized “failed” clinical trial data.
• What lessons can be learned from “failed” clinical trial data and how can these data better inform future clinical studies?

• Automating highly manual processing tasks, translating and digitizing safety case processing and adverse drug reaction documents to make them more usable
• Discuss how optical character recognition (OCR), NLP and deep neural networks are being used to format this data
• Ultimately leading to faster assessment of subject, site and study risks

• Using AI to systematically evaluate the effect of different eligibility criteria on cancer trial populations and outcomes with real-world data
• Ultimately identifying a wider and more accurate pool of patients that could potentially benefit from treatments
• Facilitating the design of more inclusive trials while maintaining safeguards for patient safety

• This session provides the unique opportunity to listen to, and engage with, innovative start-up and middle market companies that are accelerating the integration of AI into clinical trials
• Six companies will take to the stage to deliver quick fire presentations about the work they are carrying out to enhance clinical trials

• Effectively applying deep learning methods to medical image segmentation and medical time series analysis
• Refining AI imaging studies via consistent selection of clinically meaningful endpoints such as survival, symptoms, and need for treatment models into the realm of statistical inference – particularly for prediction heterogenous treatment effects
• Understanding the promise and limitations of causal AI

• Highlighting the challenges for clinical trial design in areas where robust clinical trial data is lacking
• Leveraging real-world data (RWD) to inform QSP disease progression models
• Demonstrating how the QSP-RWD modeling framework was used to establish a target value for early go/no-go decision making

• Demonstrating the value of centralized-cloud management for the initial concept evaluation phase, development of the protocol, conducting and monitoring research, and intervening to make adjustments to the protocol
• Implementing cloud-based methods for protected clinical trial data sharing

• Generating clinical evidence using digitally simulated ‘predicted outcomes’
• Fully harnessing the power of simulation in each phase of the trial and utilizing AI tools to implement knowledge gained from real-world data

CXL as memory expansion is changing how data centers think about main memory.  Direct attached memory on DIMMs has flattened out, with DDR5 barely extending the footprint of older DDR4 systems at the same time that artificial intelligence and similar applications are demanding massive increases in total memory capacity.  CXL memory modules can dramatically grow the memory footprint, however CXL introduces new challenges that have users concerned.  Increased power requirements, longer access latency, limited per-module capacity, and increased danger of data loss on power failure have end users wondering… is it really worth it?  Non-volatile main memory can address these limitations of CXL memory expansion, encouraging data centers to lay the foundation for a future with whole new ways to think about memory pooling.