Efforts to integrate big data approaches and AI into healthcare will continue to positively impact the efficiency and effectiveness of clinical trials, as well as decision-making in diagnosis and treatment.  Early progress associated with ML, for instance, has been observed in supporting pathologists and their teams with interpreting medical images. There is now an opportunity to advance precision medicine by improving clinical trial quality and design through the adoption of AI/ML approaches for recruitment, development of modelling methods, and the integration of real world data.


Directors, heads, VPs and scientists from pharma/biotech, research and healthcare sectors in the following and related areas: clinical development,  computational health,  innovation, AI/ML, bioinformaticians, clinicians, radiologists, and software engineers.


AI, Machine Learning, Clinical Trials, Imaging, Clinical Study Design, Patient Recruitment, Real World Data, Modelling and Simulation.

TABLE 1: Optimising Clinical Trial Enrolment Strategy with Artificial Intelligence

With improving technologies and research around drug target identification and development phases, the number of drugs being entered into clinical trials is increasing. A crucial part of clinical trial design is patient recruitment. This task remains time-consuming and fallacious due to the manual and paper-based nature of the patient identification process. The use of AI approaches to automate these processes will identify more suitable patients, faster, within the clinical governance guidelines whilst maintaining patient confidentiality. Ultimately, the adoption of AI in clinical trial enrolment strategy will lead to safer trails, better decision-making and the faster approval of new drugs. 

TABLE 2:   Ensuring the Financial Sustainability of Clinical NGS Testing

The adoption of NGS into the clinic has been limited by challenges that have emerged with the growing complexity of genomic information and bioinformatics. The lack of standardisation of the quality and quantity of clinical samples, lack of appropriate tools, genomic counselling services and reporting of NGS test results by healthcare professionals, continues to impede the utilisation of NGS in the clinic. How can we ensure a standard of quality for individual tests to certify that an assay is useful and warrants funding from payers? Test results are often inconsistent and so it can be difficult to argue why any one test might be better than another. Without consistency, the value of the tests cannot be accurately judged and so payers are unwilling to promise reimbursement. 

TABLE 3: Implementation of Real World Data in Precision Medicine Efforts

When we consider that the majority of patients are treated for disease outside of a clinical trial setting, it is no surprise that there has been an increasing desire from health care providers and payers to access real world data. Clinical data, including longitudinal information, treatment history and outcomes, are continually being captured in electronic health records. Comparing such data to data collected in a clinical trial has shown that outcomes have sometimes misaligned, with the actual outcome deviating from the expectation. This highlights the need for use of real world data to improve the quality of routine care and economic evaluation of personalised medicine.  The use of real world data in this context remains limited by the lack of data standards, and incentives for data sharing.

TABLE 4: Improving Data Utility with Meta and Phenotypic Data

Healthcare systems are generating increasing amounts of genetic data. However, for the data to be useful for research and drug development, the data needs context and to be enriched with linked phenotypic information. It is especially important to get this right as genomics becomes a more established part of clinical care. Amassing genomic data in healthcare silos creates an access barrier to rich information that both researchers and drug developers could mine to understand and treat disease.