Please upgrade your browser

We built this website using the latest browser technologies to deliver the very best experience.

This makes the site run faster and easier to use. Unfortunately, your browser is out of date and will not support some of these technologies.

We recommend that you use a modern browser such as Google Chrome or Microsoft Edge to view this website.

Download ChromeDownload Microsoft Edge
Blog

How machine learning can help drive the next pharmaceutical research renaissance

November 11, 2021

The promise, that machine learning has the ability to derive critical insights from clinical data to rapidly advance medical research, has been recognized and embraced by early adopters. We are now in a position where the use of machine learning in the pharma industry is rising exponentially with a growing number of companies exploring and leveraging the new possibilities offered by machine learning. They see the incredible potential of Artificial Intelligence (AI) to accelerate the analysis of data, run large complex scientific queries in rapid succession, and then identify patterns, insights and proposed solutions that are impossible for expert human brains to do alone.

Such innovation is likely to have a significant impact on the life sciences landscape, playing a critical role throughout the entire drug discovery and development pipeline—from early-stage target identification to late-stage market access. Two very exciting areas where AI is already having an impact are patient stratification and synthetic control arms. Recent data and technology advances are starting to yield compelling results.

Patient stratification

Using AI approaches to analyze large real-world datasets can quickly provide scientists a clearer view of homogenous patient cohorts within diverse populations. Better cohort identification supports more refined patient stratification and allows scientists to look at patient characteristics over time to identify disease progression. This will help researchers better understand the underlying biology of diseases, develop more targeted therapeutics, and begin to deliver on the vision of personalized medicine.

This type of stratification is supported by machine learning-powered clustering. An excellent example of this can be seen in a recent project in the UK, where supervised and unsupervised deep clustering algorithms were applied to electronic medical record (EMR) data on large patient populations to identity relevant sub-grouping within a heart failure population. AI helped to more accurately determine which patients had experienced heart failure based on their vital signs and lab work. The researchers were then able to quickly identify more than 20 contributing factors, and group patients into five major sub-cohorts based on overlapping constellations of factors. 


Synthetic control arms

It’s well known that traditional clinical trial approaches are challenging and risky. They can take years to complete and are subject to high failure rates. Given the challenges associated with assembling participants in significant numbers, and ethical concerns around giving placebos to sick patients, the practice of creating virtual – aka ‘synthetic’ – patient populations to serve as control groups has a number of clear benefits. While the practice of creating virtual populations has been around for years, the level of sophistication required to use this technique effectively in clinical trials is a much more recent development, and while the FDA (US Food and Drug Administration) has recently been approving its use in certain circumstances, they are rightly applying very strict evaluation criteria to both algorithm and data set design.

In order to be effective and meet the very high standards currently being applied to synthetic controls by regulatory bodies, synthetic controls must be carefully constructed. Not only do the algorithms for modelling the population need to be extremely accurate and unbiased, but the data sets used to construct the virtual population must meet exacting requirements. Population size and diversity, as well as the depth of data (potentially including EMR, genomic and claims data), are critical criteria for constructing appropriate synthetic control groups.

As in the case of using clustering techniques for patient stratification, the key to significantly improving clinical trial design with synthetic control arms is combining deep, longitudinal patient, clinical research and clinical trial data with advanced AI and ML technologies. This approach offers the promise of faster and more efficient clinical trials, with much less time required to recruit suitable patients, and no need to administer placebos and diagnostics that can potentially be detrimental to the patient’s health. With the volume of available data continuing to increase, synthetic control arms could become the fastest and safest way for pharma companies to carry out research into diseases.

AI is on a clear path to play an ever more critical role in meeting many of the challenges faced in pharmaceutical and life sciences development. As the technical sophistication of AI continues to increase, its likely even very large pharma companies will keep partnering with external AI experts to ensure best practices and optimal results. Together, pharma and AI companies can work together in driving the next pharmaceutical research renaissance, using AI to focus research more effectively, and bringing products to market more quickly, safely, and effectively.

Derek Baird, President North America, Sensyne Health and Commercial Director, SENSE

This article originally appeared in PharmaLive 

Blog

How machine learning can help drive the next pharmaceutical research renaissance

November 11, 2021

The promise, that machine learning has the ability to derive critical insights from clinical data to rapidly advance medical research, has been recognized and embraced by early adopters. We are now in a position where the use of machine learning in the pharma industry is rising exponentially with a growing number of companies exploring and leveraging the new possibilities offered by machine learning. They see the incredible potential of Artificial Intelligence (AI) to accelerate the analysis of data, run large complex scientific queries in rapid succession, and then identify patterns, insights and proposed solutions that are impossible for expert human brains to do alone.

Such innovation is likely to have a significant impact on the life sciences landscape, playing a critical role throughout the entire drug discovery and development pipeline—from early-stage target identification to late-stage market access. Two very exciting areas where AI is already having an impact are patient stratification and synthetic control arms. Recent data and technology advances are starting to yield compelling results.

Patient stratification

Using AI approaches to analyze large real-world datasets can quickly provide scientists a clearer view of homogenous patient cohorts within diverse populations. Better cohort identification supports more refined patient stratification and allows scientists to look at patient characteristics over time to identify disease progression. This will help researchers better understand the underlying biology of diseases, develop more targeted therapeutics, and begin to deliver on the vision of personalized medicine.

This type of stratification is supported by machine learning-powered clustering. An excellent example of this can be seen in a recent project in the UK, where supervised and unsupervised deep clustering algorithms were applied to electronic medical record (EMR) data on large patient populations to identity relevant sub-grouping within a heart failure population. AI helped to more accurately determine which patients had experienced heart failure based on their vital signs and lab work. The researchers were then able to quickly identify more than 20 contributing factors, and group patients into five major sub-cohorts based on overlapping constellations of factors. 


Synthetic control arms

It’s well known that traditional clinical trial approaches are challenging and risky. They can take years to complete and are subject to high failure rates. Given the challenges associated with assembling participants in significant numbers, and ethical concerns around giving placebos to sick patients, the practice of creating virtual – aka ‘synthetic’ – patient populations to serve as control groups has a number of clear benefits. While the practice of creating virtual populations has been around for years, the level of sophistication required to use this technique effectively in clinical trials is a much more recent development, and while the FDA (US Food and Drug Administration) has recently been approving its use in certain circumstances, they are rightly applying very strict evaluation criteria to both algorithm and data set design.

In order to be effective and meet the very high standards currently being applied to synthetic controls by regulatory bodies, synthetic controls must be carefully constructed. Not only do the algorithms for modelling the population need to be extremely accurate and unbiased, but the data sets used to construct the virtual population must meet exacting requirements. Population size and diversity, as well as the depth of data (potentially including EMR, genomic and claims data), are critical criteria for constructing appropriate synthetic control groups.

As in the case of using clustering techniques for patient stratification, the key to significantly improving clinical trial design with synthetic control arms is combining deep, longitudinal patient, clinical research and clinical trial data with advanced AI and ML technologies. This approach offers the promise of faster and more efficient clinical trials, with much less time required to recruit suitable patients, and no need to administer placebos and diagnostics that can potentially be detrimental to the patient’s health. With the volume of available data continuing to increase, synthetic control arms could become the fastest and safest way for pharma companies to carry out research into diseases.

AI is on a clear path to play an ever more critical role in meeting many of the challenges faced in pharmaceutical and life sciences development. As the technical sophistication of AI continues to increase, its likely even very large pharma companies will keep partnering with external AI experts to ensure best practices and optimal results. Together, pharma and AI companies can work together in driving the next pharmaceutical research renaissance, using AI to focus research more effectively, and bringing products to market more quickly, safely, and effectively.

Derek Baird, President North America, Sensyne Health and Commercial Director, SENSE

This article originally appeared in PharmaLive 

Blog

How machine learning can help drive the next pharmaceutical research renaissance

How machine learning can help drive the next pharmaceutical research renaissance

November 11, 2021

The promise, that machine learning has the ability to derive critical insights from clinical data to rapidly advance medical research, has been recognized and embraced by early adopters. We are now in a position where the use of machine learning in the pharma industry is rising exponentially with a growing number of companies exploring and leveraging the new possibilities offered by machine learning. They see the incredible potential of Artificial Intelligence (AI) to accelerate the analysis of data, run large complex scientific queries in rapid succession, and then identify patterns, insights and proposed solutions that are impossible for expert human brains to do alone.

Such innovation is likely to have a significant impact on the life sciences landscape, playing a critical role throughout the entire drug discovery and development pipeline—from early-stage target identification to late-stage market access. Two very exciting areas where AI is already having an impact are patient stratification and synthetic control arms. Recent data and technology advances are starting to yield compelling results.

Patient stratification

Using AI approaches to analyze large real-world datasets can quickly provide scientists a clearer view of homogenous patient cohorts within diverse populations. Better cohort identification supports more refined patient stratification and allows scientists to look at patient characteristics over time to identify disease progression. This will help researchers better understand the underlying biology of diseases, develop more targeted therapeutics, and begin to deliver on the vision of personalized medicine.

This type of stratification is supported by machine learning-powered clustering. An excellent example of this can be seen in a recent project in the UK, where supervised and unsupervised deep clustering algorithms were applied to electronic medical record (EMR) data on large patient populations to identity relevant sub-grouping within a heart failure population. AI helped to more accurately determine which patients had experienced heart failure based on their vital signs and lab work. The researchers were then able to quickly identify more than 20 contributing factors, and group patients into five major sub-cohorts based on overlapping constellations of factors. 


Synthetic control arms

It’s well known that traditional clinical trial approaches are challenging and risky. They can take years to complete and are subject to high failure rates. Given the challenges associated with assembling participants in significant numbers, and ethical concerns around giving placebos to sick patients, the practice of creating virtual – aka ‘synthetic’ – patient populations to serve as control groups has a number of clear benefits. While the practice of creating virtual populations has been around for years, the level of sophistication required to use this technique effectively in clinical trials is a much more recent development, and while the FDA (US Food and Drug Administration) has recently been approving its use in certain circumstances, they are rightly applying very strict evaluation criteria to both algorithm and data set design.

In order to be effective and meet the very high standards currently being applied to synthetic controls by regulatory bodies, synthetic controls must be carefully constructed. Not only do the algorithms for modelling the population need to be extremely accurate and unbiased, but the data sets used to construct the virtual population must meet exacting requirements. Population size and diversity, as well as the depth of data (potentially including EMR, genomic and claims data), are critical criteria for constructing appropriate synthetic control groups.

As in the case of using clustering techniques for patient stratification, the key to significantly improving clinical trial design with synthetic control arms is combining deep, longitudinal patient, clinical research and clinical trial data with advanced AI and ML technologies. This approach offers the promise of faster and more efficient clinical trials, with much less time required to recruit suitable patients, and no need to administer placebos and diagnostics that can potentially be detrimental to the patient’s health. With the volume of available data continuing to increase, synthetic control arms could become the fastest and safest way for pharma companies to carry out research into diseases.

AI is on a clear path to play an ever more critical role in meeting many of the challenges faced in pharmaceutical and life sciences development. As the technical sophistication of AI continues to increase, its likely even very large pharma companies will keep partnering with external AI experts to ensure best practices and optimal results. Together, pharma and AI companies can work together in driving the next pharmaceutical research renaissance, using AI to focus research more effectively, and bringing products to market more quickly, safely, and effectively.

Derek Baird, President North America, Sensyne Health and Commercial Director, SENSE

This article originally appeared in PharmaLive 

Blog

How machine learning can help drive the next pharmaceutical research renaissance

How machine learning can help drive the next pharmaceutical research renaissance

The promise, that machine learning has the ability to derive critical insights from clinical data to rapidly advance medical research, has been recognized and embraced by early adopters. We are now in a position where the use of machine learning in the pharma industry is rising exponentially with a growing number of companies exploring and leveraging the new possibilities offered by machine learning. They see the incredible potential of Artificial Intelligence (AI) to accelerate the analysis of data, run large complex scientific queries in rapid succession, and then identify patterns, insights and proposed solutions that are impossible for expert human brains to do alone.

Such innovation is likely to have a significant impact on the life sciences landscape, playing a critical role throughout the entire drug discovery and development pipeline—from early-stage target identification to late-stage market access. Two very exciting areas where AI is already having an impact are patient stratification and synthetic control arms. Recent data and technology advances are starting to yield compelling results.

Patient stratification

Using AI approaches to analyze large real-world datasets can quickly provide scientists a clearer view of homogenous patient cohorts within diverse populations. Better cohort identification supports more refined patient stratification and allows scientists to look at patient characteristics over time to identify disease progression. This will help researchers better understand the underlying biology of diseases, develop more targeted therapeutics, and begin to deliver on the vision of personalized medicine.

This type of stratification is supported by machine learning-powered clustering. An excellent example of this can be seen in a recent project in the UK, where supervised and unsupervised deep clustering algorithms were applied to electronic medical record (EMR) data on large patient populations to identity relevant sub-grouping within a heart failure population. AI helped to more accurately determine which patients had experienced heart failure based on their vital signs and lab work. The researchers were then able to quickly identify more than 20 contributing factors, and group patients into five major sub-cohorts based on overlapping constellations of factors. 


Synthetic control arms

It’s well known that traditional clinical trial approaches are challenging and risky. They can take years to complete and are subject to high failure rates. Given the challenges associated with assembling participants in significant numbers, and ethical concerns around giving placebos to sick patients, the practice of creating virtual – aka ‘synthetic’ – patient populations to serve as control groups has a number of clear benefits. While the practice of creating virtual populations has been around for years, the level of sophistication required to use this technique effectively in clinical trials is a much more recent development, and while the FDA (US Food and Drug Administration) has recently been approving its use in certain circumstances, they are rightly applying very strict evaluation criteria to both algorithm and data set design.

In order to be effective and meet the very high standards currently being applied to synthetic controls by regulatory bodies, synthetic controls must be carefully constructed. Not only do the algorithms for modelling the population need to be extremely accurate and unbiased, but the data sets used to construct the virtual population must meet exacting requirements. Population size and diversity, as well as the depth of data (potentially including EMR, genomic and claims data), are critical criteria for constructing appropriate synthetic control groups.

As in the case of using clustering techniques for patient stratification, the key to significantly improving clinical trial design with synthetic control arms is combining deep, longitudinal patient, clinical research and clinical trial data with advanced AI and ML technologies. This approach offers the promise of faster and more efficient clinical trials, with much less time required to recruit suitable patients, and no need to administer placebos and diagnostics that can potentially be detrimental to the patient’s health. With the volume of available data continuing to increase, synthetic control arms could become the fastest and safest way for pharma companies to carry out research into diseases.

AI is on a clear path to play an ever more critical role in meeting many of the challenges faced in pharmaceutical and life sciences development. As the technical sophistication of AI continues to increase, its likely even very large pharma companies will keep partnering with external AI experts to ensure best practices and optimal results. Together, pharma and AI companies can work together in driving the next pharmaceutical research renaissance, using AI to focus research more effectively, and bringing products to market more quickly, safely, and effectively.

Derek Baird, President North America, Sensyne Health and Commercial Director, SENSE

This article originally appeared in PharmaLive 

Arrange to meet us
Blog

How machine learning can help drive the next pharmaceutical research renaissance

November 11, 2021

The promise, that machine learning has the ability to derive critical insights from clinical data to rapidly advance medical research, has been recognized and embraced by early adopters. We are now in a position where the use of machine learning in the pharma industry is rising exponentially with a growing number of companies exploring and leveraging the new possibilities offered by machine learning. They see the incredible potential of Artificial Intelligence (AI) to accelerate the analysis of data, run large complex scientific queries in rapid succession, and then identify patterns, insights and proposed solutions that are impossible for expert human brains to do alone.

Such innovation is likely to have a significant impact on the life sciences landscape, playing a critical role throughout the entire drug discovery and development pipeline—from early-stage target identification to late-stage market access. Two very exciting areas where AI is already having an impact are patient stratification and synthetic control arms. Recent data and technology advances are starting to yield compelling results.

Patient stratification

Using AI approaches to analyze large real-world datasets can quickly provide scientists a clearer view of homogenous patient cohorts within diverse populations. Better cohort identification supports more refined patient stratification and allows scientists to look at patient characteristics over time to identify disease progression. This will help researchers better understand the underlying biology of diseases, develop more targeted therapeutics, and begin to deliver on the vision of personalized medicine.

This type of stratification is supported by machine learning-powered clustering. An excellent example of this can be seen in a recent project in the UK, where supervised and unsupervised deep clustering algorithms were applied to electronic medical record (EMR) data on large patient populations to identity relevant sub-grouping within a heart failure population. AI helped to more accurately determine which patients had experienced heart failure based on their vital signs and lab work. The researchers were then able to quickly identify more than 20 contributing factors, and group patients into five major sub-cohorts based on overlapping constellations of factors. 


Synthetic control arms

It’s well known that traditional clinical trial approaches are challenging and risky. They can take years to complete and are subject to high failure rates. Given the challenges associated with assembling participants in significant numbers, and ethical concerns around giving placebos to sick patients, the practice of creating virtual – aka ‘synthetic’ – patient populations to serve as control groups has a number of clear benefits. While the practice of creating virtual populations has been around for years, the level of sophistication required to use this technique effectively in clinical trials is a much more recent development, and while the FDA (US Food and Drug Administration) has recently been approving its use in certain circumstances, they are rightly applying very strict evaluation criteria to both algorithm and data set design.

In order to be effective and meet the very high standards currently being applied to synthetic controls by regulatory bodies, synthetic controls must be carefully constructed. Not only do the algorithms for modelling the population need to be extremely accurate and unbiased, but the data sets used to construct the virtual population must meet exacting requirements. Population size and diversity, as well as the depth of data (potentially including EMR, genomic and claims data), are critical criteria for constructing appropriate synthetic control groups.

As in the case of using clustering techniques for patient stratification, the key to significantly improving clinical trial design with synthetic control arms is combining deep, longitudinal patient, clinical research and clinical trial data with advanced AI and ML technologies. This approach offers the promise of faster and more efficient clinical trials, with much less time required to recruit suitable patients, and no need to administer placebos and diagnostics that can potentially be detrimental to the patient’s health. With the volume of available data continuing to increase, synthetic control arms could become the fastest and safest way for pharma companies to carry out research into diseases.

AI is on a clear path to play an ever more critical role in meeting many of the challenges faced in pharmaceutical and life sciences development. As the technical sophistication of AI continues to increase, its likely even very large pharma companies will keep partnering with external AI experts to ensure best practices and optimal results. Together, pharma and AI companies can work together in driving the next pharmaceutical research renaissance, using AI to focus research more effectively, and bringing products to market more quickly, safely, and effectively.

Derek Baird, President North America, Sensyne Health and Commercial Director, SENSE

This article originally appeared in PharmaLive