Sensing Auditory Evoked Potentials

Protecting Against Tinnitus With Big Data

Thomas Rouse - Senior Consultant, Medical & Healthcare

By: Thomas Rouse
Senior Consultant, Medical & Healthcare

30th August 2017

Home » Thomas Rouse

We are being continuously monitored in our daily lives; from search engines tracking browsing habits, shops analysing purchases via loyalty cards or online accounts, and social media targeting adverts based upon our friends, conversations, and activities. While we may accept this as the cost of entry to the modern world, few would deny that it is evocative of a dystopian, Big Brotheresque hierarchy, where the monitoring is unlikely to be for our benefit. Health monitoring may be a nobler goal, however even a seemingly altruistic project, DeepMind’s collaboration with the Royal Free London NHS Foundation Trust to reduce preventable deaths from acute kidney incidents, has fallen foul of the public perception and the Information Commissioner’s Office.

There is a lot of excitement about data-driven health innovation, especially where the data can be collected automatically, and potentially uploaded or aggregated. This could allow for improved outcomes, more accurate diagnoses, early warning of conditions, advanced recovery monitoring, fewer hospital visits, and ultimately revolutionise the understanding and treatment of many diseases and conditions.

We have developed a wonderful technology which is able to automatically provide detailed characterisation of a user’s auditory system by detecting electrical signals from the cochlea and auditory brainstem. No more user interaction is required other than putting on a set of headphones, and no clinical supervision is necessary. It can detect permanent or temporary changes, and, with regular use at home or work, provide early warning of the onset of hearing loss and tinnitus.

The applications are driven by who wants to use the data, and why. The technology was originally developed to allow employers more cost effectively and conveniently meet their health surveillance duty, under the ‘The Control of Noise at Work Regulations. Workers with high levels of noise exposure need to have regular audiometric tests. It may be disruptive or impractical to send staff to a testing centre. There are also requirements after the tests have taken place. The purpose of the regulations is to protect workers, and if an issue is detected, action should be taken to prevent further damage.

The employer must also keep health records of the outcome of the surveillance; however, these cannot contain any confidential medical records. With our technology, these tasks can be automated without any need to leave the workplace. Beyond compliance, there is also a potential upside for the employer if testing can be made before and after each shift. In a case of litigation relating to hearing loss, it is likely that it could be shown whether the damage occurred inside or outside of work hours.

Individuals may be concerned about their own or their loved one’s hearing. Building the technology into consumer headphones was also one of the original motivations. For example, a smartphone app would be able to take regular snapshots of a user’s hearing and alert them or a parent if there is any change, long before symptoms become apparent.

Medical trials of drugs which may have a side effect of tinnitus would be able to use the technology to objectively monitor and record the state of the auditory system instead of having to rely on the subject’s subjective assessment.

Perhaps most interestingly of all, the data could allow the technology itself to improve. This double-edged information sword needs to be handled carefully. It is essential that no-one feels their data has been abused, so this needs to be balanced with the potentially significant benefits. The signal we record was previously only obtainable by an expert practitioner in a clinic, so comparative studies over time are limited. Long term data from a large number of subjects is likely to improve system performance and the wider understanding. It may shed light upon insidious conditions, such as hidden hearing loss and tinnitus, and provide a vital additional tool for audiologists as part of an integrated healthcare system.

Save

Save

Save

Save

Save

Save

Save

We are being continuously monitored in our daily lives; from search engines tracking browsing habits, shops analysing purchases via loyalty cards or online accounts, and social media targeting adverts based upon our friends, conversations, and activities. While we may accept this as the cost of entry to the modern world, few would deny that it is evocative of a dystopian, Big Brotheresque hierarchy, where the monitoring is unlikely to be for our benefit. Health monitoring may be a nobler goal, however even a seemingly altruistic project, DeepMind’s collaboration with the Royal Free London NHS Foundation Trust to reduce preventable deaths from acute kidney incidents, has fallen foul of the public perception and the Information Commissioner’s Office.

There is a lot of excitement about data-driven health innovation, especially where the data can be collected automatically, and potentially uploaded or aggregated. This could allow for improved outcomes, more accurate diagnoses, early warning of conditions, advanced recovery monitoring, fewer hospital visits, and ultimately revolutionise the understanding and treatment of many diseases and conditions.

We have developed a wonderful technology which is able to automatically provide detailed characterisation of a user’s auditory system by detecting electrical signals from the cochlea and auditory brainstem. No more user interaction is required other than putting on a set of headphones, and no clinical supervision is necessary. It can detect permanent or temporary changes, and, with regular use at home or work, provide early warning of the onset of hearing loss and tinnitus.

The applications are driven by who wants to use the data, and why. The technology was originally developed to allow employers more cost effectively and conveniently meet their health surveillance duty, under the ‘The Control of Noise at Work Regulations. Workers with high levels of noise exposure need to have regular audiometric tests. It may be disruptive or impractical to send staff to a testing centre. There are also requirements after the tests have taken place. The purpose of the regulations is to protect workers, and if an issue is detected, action should be taken to prevent further damage.

The employer must also keep health records of the outcome of the surveillance; however, these cannot contain any confidential medical records. With our technology, these tasks can be automated without any need to leave the workplace. Beyond compliance, there is also a potential upside for the employer if testing can be made before and after each shift. In a case of litigation relating to hearing loss, it is likely that it could be shown whether the damage occurred inside or outside of work hours.

Individuals may be concerned about their own or their loved one’s hearing. Building the technology into consumer headphones was also one of the original motivations. For example, a smartphone app would be able to take regular snapshots of a user’s hearing and alert them or a parent if there is any change, long before symptoms become apparent.

Medical trials of drugs which may have a side effect of tinnitus would be able to use the technology to objectively monitor and record the state of the auditory system instead of having to rely on the subject’s subjective assessment.

Perhaps most interestingly of all, the data could allow the technology itself to improve. This double-edged information sword needs to be handled carefully. It is essential that no-one feels their data has been abused, so this needs to be balanced with the potentially significant benefits. The signal we record was previously only obtainable by an expert practitioner in a clinic, so comparative studies over time are limited. Long term data from a large number of subjects is likely to improve system performance and the wider understanding. It may shed light upon insidious conditions, such as hidden hearing loss and tinnitus, and provide a vital additional tool for audiologists as part of an integrated healthcare system.

Save

Save

Save

Save

Save

Save

Save

What does your body have to say about you?

What does your body have to say about you?

Thomas Rouse - Senior Consultant, Medical & Healthcare

By: Thomas Rouse
Senior Consultant, Medical & Healthcare

12th April 2017

Home » Thomas Rouse

Many moons ago, while writing up my thesis, I had been burning the candle at both ends. Unsurprisingly, I was starting to feel tired and run down. A wise friend encouraged me to heed some sage advice; “When your body speaks, listen!”.

Somehow, that little mantra lodged in my mind, and I could no longer disregard all the hints and messages I was receiving from myself but had previously managed to ignore.

The image of organs communicating their status to our conscious mind, like a conversation or shouted warning has stuck with me. Anthropomorphizing my own subsystems helped me to see them in a different way. It made the internal advice friendly, convincing me to work with my body not against it. As time has passed, that image has developed into a full blown intrigue.

There are many methods of expression within us, with seemingly no end to the fascinating tales biologist friends have provided of the incredible processes that take place. These interactions make us who we are, and reveal how we function. In just the area I specialise in, electrical signals, there is a vast amount of information passing around. It is tempting to say mind-boggling, but that would be incorrect, as we (hopefully) cope without being in a permanently boggled state! Our brain is a very effective filter, alerting us only when there is something of interest.

This is because it’s not just at the limits of our endurance that information appears. It would be nice if there was a simple alarm which only triggered when we have exceeded our specifications, are outside normal parameters, or have a failure. The signals need interpretation, and this is not a trivial matter.

What does your body have to say about you?Take for example the electrocardiogram (ECG), we can examine it on a number of levels. Our heart rate can be extracted by looking at the time between the largest spikes in the signal. This can be evaluated as an absolute value, checking it is inside acceptable limits. The safe range can vary depending on factors such as age, or what we are doing at the time. From this, we can then study how the heart rate changes with time. Is the spacing even, and how quickly does it recover after exercise? This is useful information on its own, but is just the tip of the iceberg. The ECG is composed of several features which correspond to different processes in the cardiac cycle. The height and spacing of the different bumps and troughs can provide a vital insight into how the heart is functioning. A skilled interpreter can analyse them to help diagnose a range of conditions. There are enough differences between us, however, that even with a healthy heart, this signal can be used as a biometric identifier.

We already have wearable devices which can show signals such as the ECG outside a hospital environment. The ECG is relatively loud, yet is easily masked by noises generated when moving, or electrical interference such as our 50/60Hz mains supply. Other signals of interest can be a million times quieter, making their reliable extraction something of a black art. Even with a perfect waveform, in most circumstances, diagnosis or monitoring of a condition is difficult or impossible to automate without some expert human intervention. If a specialist is involved, our brain has probably already determined that something is wrong.

What does your body have to say about you?The ability to listen and understand the subtle inflections in our internal communications, by benevolently eavesdropping at the point of care or in our daily lives could provide huge benefits from early detection of conditions, or allowing people to leave hospital sooner.

At each stage in a physiological process a signal may be generated. For example, after the initial stimulation, impulses could be detected travelling though the nervous system, when they reach the brainstem, or as they pass towards the higher levels of brain function. Like a game of Chinese whispers, each new wave may be an altered interpretation of the original. We can use this to create more data, by understanding and giving meaning to the interaction between the points in the chain, or simplify it by spying on the neural pathway which extracts what we want.

All these challenges; determining what is being said, getting the raw data, and working out what it means are at a fascinating stage where they are beginning to become possible outside specialist clinical environments. Our body has started the conversation, how are we going to respond?

Save

Many moons ago, while writing up my thesis, I had been burning the candle at both ends. Unsurprisingly, I was starting to feel tired and run down. A wise friend encouraged me to heed some sage advice; “When your body speaks, listen!”.

Somehow, that little mantra lodged in my mind, and I could no longer disregard all the hints and messages I was receiving from myself but had previously managed to ignore.

The image of organs communicating their status to our conscious mind, like a conversation or shouted warning has stuck with me. Anthropomorphizing my own subsystems helped me to see them in a different way. It made the internal advice friendly, convincing me to work with my body not against it. As time has passed, that image has developed into a full blown intrigue.

There are many methods of expression within us, with seemingly no end to the fascinating tales biologist friends have provided of the incredible processes that take place. These interactions make us who we are, and reveal how we function. In just the area I specialise in, electrical signals, there is a vast amount of information passing around. It is tempting to say mind-boggling, but that would be incorrect, as we (hopefully) cope without being in a permanently boggled state! Our brain is a very effective filter, alerting us only when there is something of interest.

This is because it’s not just at the limits of our endurance that information appears. It would be nice if there was a simple alarm which only triggered when we have exceeded our specifications, are outside normal parameters, or have a failure. The signals need interpretation, and this is not a trivial matter.

What does your body have to say about you?Take for example the electrocardiogram (ECG), we can examine it on a number of levels. Our heart rate can be extracted by looking at the time between the largest spikes in the signal. This can be evaluated as an absolute value, checking it is inside acceptable limits. The safe range can vary depending on factors such as age, or what we are doing at the time. From this, we can then study how the heart rate changes with time. Is the spacing even, and how quickly does it recover after exercise? This is useful information on its own, but is just the tip of the iceberg. The ECG is composed of several features which correspond to different processes in the cardiac cycle. The height and spacing of the different bumps and troughs can provide a vital insight into how the heart is functioning. A skilled interpreter can analyse them to help diagnose a range of conditions. There are enough differences between us, however, that even with a healthy heart, this signal can be used as a biometric identifier.

We already have wearable devices which can show signals such as the ECG outside a hospital environment. The ECG is relatively loud, yet is easily masked by noises generated when moving, or electrical interference such as our 50/60Hz mains supply. Other signals of interest can be a million times quieter, making their reliable extraction something of a black art. Even with a perfect waveform, in most circumstances, diagnosis or monitoring of a condition is difficult or impossible to automate without some expert human intervention. If a specialist is involved, our brain has probably already determined that something is wrong.

What does your body have to say about you?The ability to listen and understand the subtle inflections in our internal communications, by benevolently eavesdropping at the point of care or in our daily lives could provide huge benefits from early detection of conditions, or allowing people to leave hospital sooner.

At each stage in a physiological process a signal may be generated. For example, after the initial stimulation, impulses could be detected travelling though the nervous system, when they reach the brainstem, or as they pass towards the higher levels of brain function. Like a game of Chinese whispers, each new wave may be an altered interpretation of the original. We can use this to create more data, by understanding and giving meaning to the interaction between the points in the chain, or simplify it by spying on the neural pathway which extracts what we want.

All these challenges; determining what is being said, getting the raw data, and working out what it means are at a fascinating stage where they are beginning to become possible outside specialist clinical environments. Our body has started the conversation, how are we going to respond?

Save

Save

Save