In this series, we’ve explored bioelectronics and neuromodulation; how they’re changing the lives of patients, the major players in the industry, and how it’s regulated. We’ve also thrown in some thoughts from neuroscientists on how these new fields could benefit from looking at existing drug development processes for model translation and scaleup. This time, we’ll be looking at the future of precision medicine and the role that artificial intelligence and neural implant technology will play.
AI and ML in bioelectronic medicine
According to Health Tech World, the AI sector in the UK alone was worth over £15 billion in 2020 – and it’s set to grow the UK economy 10% by 2023. With its ability to learn and quickly crunch masses of data, artificial intelligence (AI) and machine learning (ML) technology have the potential to speed up or virtually eliminate many currently manual processes and drastically change healthcare for the better. Enabling healthcare specialists across the world (in a supportive capacity) to spend their time more efficiently and drastically changing healthcare for the better.
Just some of the currently used or proposed AI and ML applications in healthcare include:
Digital scribes
Patient/doctor conversations are recorded and the technology then deconstructs the text to autofill patient forms and records. The benefits here are threefold – freeing up time for the doctor that would otherwise be spent on admin, ensuring a level of form detail consistency, and human error mitigation that may occur when doctors are overburdened or exhausted.
Diagnosis predictions
A recent study in Germany (1) found promising results in preventing missed breast cancer with AI. By first being fed thousands of positive and negative screens, the technology learned how to differentiate between the two. Off the back of the success of this study, a triage system where the AI/ML would forward all positive or lower certainty screens to the radiologist for further assessment is now being proposed. This would greatly alleviate the current global shortage of radiologists caused by increasing demand for imaging studies/imaging needs due to the global pandemic and aging populations.
Pattern recognition and predictive analytics for decision making
Experts at the National Academy of Medicine (2) have indicated that AI/ML could be used to offer predictions on patient health or predispositions to certain diseases based on their genomic makeup, their patient records, and lifestyle information. This would empower clinician decision-making and help them offer preventative measures or suggestions for lifestyle changes before any illness occurs.
Finding drug molecules and predicting toxicity
One of the issues with drug discovery and development is that experimental drugs often inhibit the production of CP450; enzymes in the body that break down chemicals. Meaning the drugs build up in the body to toxic levels and end up causing more harm than the good they are designed for. To determine and discard these drug candidates as early as possible, toxicity tests are carried out in the lab.
Frustratingly, about a third of the time, the inhibition of CP450 isn’t discovered till the human clinical trial phase. So, it’s not surprising that pharmaceutical companies are now turning to AI to identify more promising drug candidates and screen out the drugs which are potentially toxic to humans before they progress far in the development pipeline. In using AI, millions of dollars and years of research are saved by cutting out the poor drug candidates early on.
From Pharmaceuticals to Neuroceuticals
However, the cost of drug discovery, development, and production in the last 20 years has risen exponentially and the fact that all the major molecules have already been found is making the search for new treatments much more difficult. This has led to scientists turning to new avenues for viable treatment alternatives. In particular, leveraging the nervous system.
The nervous system in the human body controls all our organs. When it malfunctions, it can cause major diseases such as heart disease, diabetes, arthritis, and respiratory diseases. For years the focus of research has been to affect these organs with pharmaceutical therapies. But studies using neuromodulation have proven it as a viable area for developing a new class of therapies. Classed as any therapy that takes insight from the nervous system in real-time, these ‘Neural Digital Therapies’ have already shown promise across a wide range of therapy areas with many successful studies being carried out globally.
In recognition of the potential of the nervous system, the US NIH Common Fund created the Stimulating Peripheral Activity to Relieve Conditions (SPARC) program. Working with leading researchers in both industry and academia, the SPARC program is accelerating the development of therapeutic devices that modulate electrical activity in nerves. – The primary objective being to improve organ function.
Neural Digital Therapies: The next frontier in precision medicine
Building on its work with the SPARC program, in 2020 BIOS Health launched its pioneering Autonomic Therapy Initiative (ATI). Which will see the company work with world-leading partners to deliver a new class of treatments for heart disease using AI-powered neural interfaces.
Having leveraged recent breakthroughs in AI and ML, BIOS is now working to translate the ‘language’ of the nervous system. By collecting and analysing neural data, BIOS is able to precisely link nerve activity to specific chronic conditions by isolating the neural biomarkers of disease and respond to them. – Effectively reading and writing the body’s neural signals in real time.
As valuable indicators of normal biological processes, biomarkers can be used for testing whether the body is responding to applied treatments. And as they are also used in drug development, neural biomarkers could be used by the pharmaceutical industry to reposition existing drugs. Repositioning (or ‘retargeting’) existing FDA approved drugs for different disorders would result in massive savings in research time and drug development costs.
“Many neurally mediated diseases are currently being treated fairly imprecisely with drugs. We’re doing this in two ways. Firstly, using AI software we can read the body’s neural signals and understand what is occurring in real-time to enable the software to then write correcting neural code to deliver software therapies. Secondly, our technology enables us to identify neural biomarkers, increasing our understanding of diseases further. We are improving traditional pharma therapies. If we can see what’s happening inside an individual in response to drugs, we can profile patients more effectively, target existing drugs more effectively, and ultimately deliver better therapies... neural digital therapies are levelling up precision medicine. -Delivering a whole new treatment platform via software. But we can also target existing drugs better with this information. - Even discover new drug options we hadn’t considered previously.”
Catherine Hanley ATI Programme Manager, BIOS Health
Future trends and applications in bioelectronic medicine
In the second blog of this series (THE NEUROMODULATION LANDSCAPE) we explored how neuromodulation implant devices are currently being used to treat a range of illnesses linked to the nervous system. We also looked at the innovation drivers in neuromodulation; and how closed-loop therapies and multi-indication systems are expected to increase as the market calls for more holistic treatments. Adding to this, it would be a fair bet to suggest that wearables such as the common heart monitor will play an important part too, for real-time data collection.
The rise in remote health wearables is already assisting clinicians in gathering quantifiable disease progression data. However, they are also a massive opportunity for the neural industry to obtain real-time organ performance insights. As we move towards a future of personalised medicine, these will form an important link with AI and patient records.
In this holistic view of future healthcare, ongoing therapy will be adaptable to long-term patient outcomes, resulting in less in-person consultations. To take this a step further, BIOS sees the combination of AI, ML, neural implants and neural digital therapies culminating in prescribed software algorithms from medical app stores – like a personalised healthtech Spotify!
“In the future we envision a seamless experience where Doctors aren’t forced to try a multitude of treatments and waiting to see which of them worked. But instead, make subtle and continuous interventions with new algorithms tailored to the patient’s needs and uploaded to the smart implant. We see neural information becoming a critical part of the drug development pipelines and also neural digital therapies as alternatives to drugs, stimulating nerves instead of drugs and giving second-by-second personalised treatments. This will mean fewer side effects and better, more targeted, more seamless treatments.” Emil Hewage CEO, BIOS Health
A future where neurally mediated chronic diseases are treated in a very personalised way with smart implants and software may sound fantastic. But the technology and science are already here. And the reality is that with advances in bioelectronics and neural digital therapies, 60 trillion dollars of chronic disease spending globally could suddenly become more affordable. And for the patient, the experience would become seamless, with fewer side effects, and more personalised than ever before…
Missed part 3? Catch up with the rest of the series here.
If you would like to know more about how BIOS is building the infrastructure for neural digital therapy development with the pharmaceutical industry to treat chronic diseases, or would like to become a partner, please contact our partnerships team.
References
1. Leibig C, Brehmer M, Bunk S, Byng D, Pinker K, Umutlu L. Combining the strengths of radiologists and AI for breast cancer screening: a retrospective analysis. Lancet Digit Health. 2022 Jul;4(7):e507-e519. doi: 10.1016/S2589-7500(22)00070-X. PMID: 35750400.
2. Matheny, M., Israni, S. T., Ahmed, M., & Whicher, D. (2019). Artificial intelligence in health care: The hope, the hype, the promise, the peril. Washington, DC: National Academy of Medicine.
3. Weigand LA, Undem BJ. Allergen-induced neuromodulation in the respiratory tract. Chem Immunol Allergy. 2012;98:142-162. doi:10.1159/000336508
4. Faulkner S, Jobling P, March B, Jiang CC, Hondermarck H. Tumor Neurobiology and the War of Nerves in Cancer. Cancer Discov. 2019;9(6):702-710. doi:10.1158/2159-8290.CD-18-1398
5. Chui RW, Buckley U, Rajendran PS, Vrabec T, Shivkumar K, Ardell JL. Bioelectronic block of paravertebral sympathetic nerves mitigates post-myocardial infarction ventricular arrhythmias. Heart Rhythm. 2017;14(11):1665-1672. doi:10.1016/j.hrthm.2017.06.025
6. Zanos TP, Silverman HA, Levy T, et al. Identification of cytokine-specific sensory neural signals by decoding murine vagus nerve activity. Proc Natl Acad Sci U S A. 2018;115(21):E4843-E4852. doi:10.1073/pnas.1719083115
7. Lambru S, Lanteri-Minet M. (2020). Neuromodulation in Headache and Facial Pain Management. ISBN : 978-3-030-14120-2
8. Guyot M, Simon T, Ceppo F, et al. Pancreatic nerve electrostimulation inhibits recent-onset autoimmune diabetes. Nat Biotechnol. 2019;37(12):1446-1451. doi:10.1038/s41587-019-0295-8