The Latest AI Innovations in Clinical Research

Advancements in artificial intelligence continue to amaze, demonstrating vast capabilities which have attracted attention from the pharmaceutical industry and clinical organisations. The latest news demonstrates the potential impact of AI technologies in research and development and clinical trials. The field of neurodegenerative diseases is one of the latest examples, leveraging AI to support early diagnosis. 

Understanding AI 

AI is described as the simulation of human intelligence in machines that are programmed to process information like humans and to mimic their actions. Within AI is a subfield known as machine learning (ML). 

ML refers to a “subset of AI that enables computers to discover patterns in large data sets, make predictions and improve these predictions over time with repeated exposure to the data”. There are two main techniques used to apply ML: supervised and unsupervised learning. Unsupervised learning is a type of algorithm that learns patterns from data without tags. Supervised learning algorithms, on the other hand, are formed from labelled training data which consists of a set of training examples. 

In other words, supervised learning relies on human intervention to label data in order to train the model to search for a specific component – cancer image analysis for example. Unsupervised learning on the other hand, analyses vast amounts of data which has not been labelled in order to identify associations or trends.

AI technologies are becoming increasingly prevalent in clinical research, typically to support analysis of the ever-expanding data. Pattern recognition, evolutionary modelling, and data integration are a few of the real world applications in which AI can be applied to. 

Machine learning, for example, can discover patterns in data that don’t depend on the expertise of a human programmer. Not only does this reduce the time and cost of data interpretation, it can enable the identification of significant trends or patterns from the most complex data sets. 

AI has many applications for improving clinical research returns including patient identification; site selection; patient monitoring and cohort composition. The capabilities of AI to revolutionise the pharma industry and clinical fields are becoming increasingly evident, especially so with the latest advancements. 

Neurodegenerative diagnosis

Dementia represents one of the most prevalent diseases in the aging population, with 1 in 14 of those aged 65 years and over diagnosed with the disease. Currently, there is no cure for dementia and treatment available focuses on symptom management. 

Early diagnosis is incredibly useful for patients, as it can help them access the correct services and support they require to plan for the future, and potentially begin treatment to better control symptoms. In addition, it can eliminate the possibility of other, potentially treatable conditions with dementia-like symptoms. 

Unfortunately, the diagnosis of dementia is a lengthy process consisting of several MRI brain scans and a multitude of cognitive tests. This can take anything from four to 12 weeks depending on the waiting list.

Recently, however, researchers at Cambridge University in the UK are trialling an AI system that could potentially spot the signs of dementia after a single brain scan. The aim of the system is to enable treatment to begin earlier in order to slow the progression of dementia, including Azheimer’s disease. 

According to a recent article, the AI system has been trained using thousands of images from brain scans of patients with dementia, and uses an algorithm to identify key patterns that are undetectable to radiologists. The efficacy of the system will be tested using around 500 patients that will be enrolled into a clinical trial at Addenbrooke’s Hospital in Cambridge and other memory clinics across the UK.

One of the obvious advantages of this system is the speed with which the AI system could diagnose dementia. The diagnosis for neurodegenerative diseases like dementia is often dependent on the collaboration of radiologists, neurologists and healthcare scientists – the process of which can take months. A more rapid result enables patients to plan potential treatment regimens whilst they have the capacity to make decisions for themselves. 

In addition to supporting healthcare, this digital technology has huge potential to support research and development. Developing systems to identify patterns of early disease in specific neurodegenerative diseases could shape drug development, by better understanding pathogenesis. 

In terms of challenges, it is likely that some patients will be skeptical about the capabilities of an AI system and feel more comfortable with human analysis. Such uncertainties could arise during the upcoming clinical trial, which hopefully, will enable clinicians in the future to support patients through the process by explaining how the AI system works. 

Deep 6 AI platform

On June 29 2021, Deep 6 announced substantial expansion of its innovative AI-based clinical trial acceleration software. The software utilises AI to rapidly search through vast amounts of data including electronic medical reports and pathology reports to highlight important insights. In comparison to other software, this AI-based system can determine such insights in a matter of minutes, even from the most complex of data sets.

In terms of clinical research, this has the potential to revolutionise one of the more difficult tasks right now: the digitalisation of clinical trial recruitment. According to a recent report, “up to 90% of patient data is unstructured and much of it is siloed across different systems”. 

Extensive lab data, detailed pathology reports and medical records makes it extremely difficult to analyse and interpret this data as a whole and identify important trends or patterns. As a result, this comprises the efficiency of the patient recruitment process for clinical trials. 

In this system, Deep 6 has leveraged the ability of AI to analyse unstructured data (using natural language processing for example) and then fit the data into a model, which summarises the patient’s clinical journey. 

This is an example of the impact AI-based software could have on the speed and efficiency of clinical research. With an increasing number of clinical trials adopting a decentralised approach, an increasing amount of data is being digitised and stored across a number of different platforms. As technology continues to evolve, the integration of AI into clinical and pharmaceutical context will no doubt become more prevalent. 

Charlotte Di Salvo, Lead Medical Writer
PharmaFeatures