Experimind | Turning failure into Success

Discovering the Hidden Value in Clinical Trials

ExperiMind Technologies aims to identify the best sub-population(s) within clinical trials for
whom the treatment effect is significantly greater than within the rest of the population.
ExperiMind Technologies has developed a set of unique and successfully tested data mining
and machine learning algorithms that go far beyond “conventional” biostatistics to discover
hidden relationships within clinical data. This new insight means ExperiMind Technologies
can identify the ideal sub-population(s) for a novel therapy early in its development program
and thereby improve the success of clinical trials.

The Need for ExperiMind Technologies

Many clinical trials end without achieving statistical significance for the pre-defined end-points, mainly due to one or more of the following reasons:

Sample Size too Small
Poor Efficiacy
Protocol Violation
Inclusive / Exclusive Criteria

“Doctors are men who prescribe medicines of which they know little, to cure diseases of which they know less, in human beings of whom they know nothing.”

Voltaire (1694 –1778) French philosopher

The Process

Pharmaceutics

Furthermore there is much heterogeneity among the individual outcomes and only a subset of patients responds to treatment. On average only about 50% of patients respond to prescribed drugs. In addition research and development costs are soaring and there are increasing risks in getting a new compound to market. Research and development expenses increased from US $2 billion in 1980 to US $26 billion in 2000.

Pharmaceutical companies need to be able to identify which patients will respond better to treatments and by tailoring therapeutics, the right drug can be delivered to the right patient with the right dose at the right time, drastically improving individual patient outcomes. Improvements in this domain would have important implications on drug development and could provide significant benefits for patients

To achieve the above, a drug development teams must endeavor to find the most relevant factors that define the largest sub-group(s) of patients that have a clinically meaningful benefit-risk profile for any given treatment regimen. ExperiMind’s technology provides a solution to finding the best “sub-population with enhanced treatment effect”.

“There are two realities in drug discovery,”

says Don Frail, chief scientific officer of Pfizer’s Indications Discovery Unit.

“The majority of candidates tested in development do not give the desired result, yet those drugs that do succeed typically have multiple uses.”

Pfizer, Washington University $ 22.5 M collaboration – May 2010.
(Pfizer Press Release – 5/17/2010)

Minimizing the Error

FDR analysis with confidence Interval

Overcome challenges in actual/agriculture fields

Crop yields are rising more slowly than they were several decades ago (J. A. Foley et al. Nature 478, 337–342; 2011). This is especially evident in Europe, and has serious implications for its responsibilities in global food production.Food production capacity is faced with an ever-growing number of challenges, including a world population expected to grow to nearly 9 billion by 2050 and a falling ratio of arable land to population. agricultural innovations focusing on seed, the most vital and critical agricultural input.

Using data mining in agriculture represents a new opportunity to improve process and plant performance. These methods which integrates machine learning, clustering and optimizations algorithms are helping to elucidate relations between traits and other markers.

ExperiMind’s pioneering technology has proved to assist in genetic marker based breeding ; as well as analyses results of crops to identify factors that affects yield and throughput.

“With 9 billion people expected to be alive in 2050, researchers are running out of time to develop new ways to feed them sustainably.”

Mueller, N. D. et al. Nature dx.doi.org/10.1038/nature11420 (2012).

ExperiMind’s Results: Single Variable

ExperiMind Technologies Benefits

ExperiMind’s solution enables pharmaceutical, medical device companies and other entities involved in clinical trials to:

Identify the patients that will derive a more pronounced benefit from a medical therapy

Learn about the efficacy, safety and potential benefits of new medicines earlier in the development process, thus enabling development of safer, more effective drugs

Lower costs by better planning and optimizing inclusion and exclusion criteria

Increase the probability of success by optimizing protocols and endpoints

Determine stopping rules to control overall costs per compound

Shorten the length of clinical trials through improved planning

Why ExperiMind Technologies?

Led by a highly experienced and multidisciplinary team, ExperiMind has developed a set of unique algorithms specifically tailored to meet healthcare industry needs.
ExperiMind’s powerful algorithms discover hidden insights far beyond current biostatistics due to their distinctive abilities to analyse clinical data in multiple ways.
ExperiMind’s outputs provide a win-win situation for our clients. ExperiMind can not only help identify new subsets of treatment responders but also help exclude non-responders. Through developing more targeted and efficient clinical trials, ExperiMind can increase the chance of clinical success and save companies valuable time and money as the trial progresses.
ExperiMind’s unsupervised approach to data mining and ability to work in parallel to the clinical analysis also means the technology is very easy to implement, results are delivered quickly and minimal support is required from your project team.
ExperiMind’s output is a set of recommendations for the design of specific clinical trials that supports a company’s senior management, clinical development, marketing and business development teams to make more informed clinical trial decisions and improve the cost-effectiveness of both in-house and
acquisition/investment programs.
ExperiMind’s technology has been successfully tested on various scaled databases.