Changing a patient's destiny in the age of AI

This blog is dedicated to you: the patients, the caregivers and the people who wake up every morning with hope and optimism. This blog is dedicated to all those who accept living each day with uncertainty.


The life of a patient is never easy, but living with a rare condition is particularly challenging. Rare diseases are, by definition, rare. But what does rare really mean? In Europe, about 30 million people are living with one of the 6,000 to 8,000 rare diseases we currently know about [1] [2]. It doesn't seem so rare anymore, does it? Most of the rare diseases have a strong genetic component, and even then, getting a correct diagnosis is very difficult. This year at FutureHealth Basel, we met a patient who had waited 15 years to get diagnosed with Myhre Syndrome.


Why does it take so long? How can this be somebody's destiny? What are the current healthcare systems doing wrong and what can we do to solve this problem?


Changing a patient's destiny

Artificial Intelligence (AI) has shown to be an incredibly powerful tool that can revolutionise diagnostics, and it has been confirmed that deep learning models can have an equivalent diagnostic performance to that of healthcare professionals [3]. AI is much closer to humans than we think and although this technology can change our lives for the better, AI is still not widely adopted in life sciences.


In order to experience precision medicine as a reality, we need to step into a time machine.

Let's go to 2030!


It's March 3rd 2030 and Sophie has just been born. The hospital where Sophie is born is one of many that is now powered by AI. As part of the routine checks performed on newborns, Sophie's doctor takes a sample of her DNA for genetic screening. The test reveals that Sophie has a mutation of a very important gene, and that this mutation can change her life completely. Understandably worried, the doctor uses an AI system to check against a genomic database and is instantly alerted to the fact that this mutation is associated with a specific rare disease.


A few months pass when Sophie’s doctor decides to use an AI system that has been trained to diagnose rare diseases based on facial features [4]. Unfortunately, this test further confirms the genetic results, and Sophie is diagnosed with a rare disease. In 2030, with the help of data and algorithms, accurate diagnosis can be achieved at speed, increasing the awareness for Sophie, her caregivers, and the doctors.


Achieving an early, accurate diagnosis for rare disease patients is simply life-changing. Nobody can express it better than a Myhre Syndrome patient themselves: "Nothing is worse than not knowing what you have. Even if there is no cure or therapy." For the patient and the caregiver, awareness means everything. It means that a proactive approach towards anticipating future disease complications can be taken.


In 2030, with the use of AI, Sophie's family will know about her condition just a few months after her birth. In 2020, however, rare disease patients are waiting years to find out what disease they are suffering from.


Having to wait for almost 15 years to get an accurate diagnosis can no longer be somebody's destiny, not in the age of AI.

Building bridges

AI solutions need to open minds and keep the focus on patients. They must connect individuals, and provide insights that increase knowledge and empower action to be taken. In 2020 the problem is not the lack of technology, the problem is the mindset, the tunnel vision.


In healthcare, key stakeholders across the ecosystem are stuck in their own incentive bubble, and between stakeholders, information is both lost and misunderstood, and the patient's needs are sometimes forgotten.

So at OKRA, how are we solving this dilemma? How are we helping to move into a world where Sophie won't have to wait 15 years for a diagnosis?


SiteGuide

Let's paint the following picture: a drug that can potentially treat a rare disease has been identified, so we decide to run a clinical trial. How can we find the right investigators and the best sites to provide rare disease patients with the best possible clinical trial?


OKRA's SiteGuide is an AI product that integrates all historical clinical trial data with scientific literature, and helps determine the best place to conduct a clinical trial, bringing together rare disease patients worldwide.
SiteGuide helps decipher the true reasons and recommendations behind factors affecting site selection, empowering feasibility teams to target high-performing investigator sites. SiteGuide is designed to change the future of clinical trial operations, by empowering feasibility teams and site monitors to make quick and informed decisions based on all the available data.


With SiteGuide, patients can be confident that their clinical trial will be performed in the best possible site.


MedCompass

The clinical trial for the rare disease drug has started and results are positive. This drug could really make a difference. How do we make sure that rare disease specialists know about this at the appropriate time?


MedCompass facilitates communication between rare disease specialists and Medical Affairs teams. MedCompass ensures that MSLs engage with the right stakeholders, predicting scientific need and proactively supporting the delivery of information to satisfy expert needs. At any given moment, we identify the best opportunity to engage, using behind-the-scenes AI to combine medical strategy with real-time scientific information (multiple sources) and market conditions.


With MedCompass, patients and caregivers can become more confident that rare disease specialists and life sciences companies are working together to bring the right drug to the market.


FieldFocus

The treatment for the rare disease has been validated and approved, so now we must bring it to the patients who may benefit from it. How? We must empower sales representatives to make sure that every doctor working with rare disease patients is made aware of the treatment at the right time and with tailored content.


OKRA's FieldFocus is an AI system that learns from historic data (sales, CRM and prevalence) ensuring that sales teams make the right decisions consistently. Using AI, we help identify the best opportunities to engage with healthcare practitioners at any given moment. FieldFocus uses multiple datasets to suggest the conversations that will further elevate engagement, the urgency of the conversation whilst also providing the reasons why.


Thanks to FieldFocus, sales representatives are empowered to reach doctors effectively through communication channels that are preferred, enabling the delivery of the latest treatments to those who actually need it.


Embracing uncertainty to leap ahead

At OKRA, we believe that to effectively build bridges among healthcare stakeholders, AI has to be humanised - it has to be explainable, actionable and predictive. Only when AI outputs are explained, can users take action and proactively anticipate predicted outcomes. Explainability in AI is key to ensure user trust: data should be used in an ethical way, providing transparency and user empowerment.


AI is available today, we don't have to wait for 2030 to make precision medicine a reality. The choice is ours: what kind of world do we want to live in? Are we okay with patients having to wait for 15 years to get the right diagnosis?

Now is the time: we need to leap ahead in the intelligence race.

If we dare to embrace uncertainty and adopt trustworthy AI in healthcare, we will change the destiny of many patients for the better.



References

[1] European Commission (2020). Rare diseases [online] Available at: https://ec.europa.eu/info/research-and-innovation/research-area/health-research-and-innovation/rare-diseases_en [Accessed 2 March 2020].

[2] Nguengang Wakap, S., Lambert, D.M., Olry, A. et al. Estimating cumulative point prevalence of rare diseases: analysis of the Orphanet database. Eur J Hum Genet 28, 165–173 (2020). DOI: https://doi.org/10.1038/s41431-019-0508-0

[3] Liu, X. et al. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digital Health 2019; 1: e271–97. DOI: https://doi.org/10.1016/S2589-7500(19)30123-2

[4] Gurovich Y. et al., Identifying facial phenotypes of genetic disorders using deep learning, Nat Med. 2019 Jan;25(1):60-64. DOI: https://doi.org/10.1038/s41591-018-0279-0

Request a Demo

Do you want to learn more?
Leave your email here and we will contact you.

For any other requests including press and speaking opportunities, please email hello@okra.ai