MirrorHR Epilepsy Research Kit

• Epilepsy is a chronic non-communicable disease of the brain that affects people of all ages.
• Around 50 million people worldwide have epilepsy, making it the fourth most common neurological disease globally. 
• Nearly 80% of people with epilepsy live in low and middle-income countries. 
• Estimated that up to 70% of people living with epilepsy could live seizure-free if adequately diagnosed and treated. 
• The risk of premature death in people with epilepsy is up to three times higher than in the general population. 
• Three-quarters of people with epilepsy living in low-income countries do not get the treatment they need.
• In Italy, 1 in 100 people is affected by epilepsy, at least 125,000 people suffer from a form of epilepsy that does not respond to drugs, 25 thousand new cases are diagnosed a year, 80% of epilepsies arise in childhood and adolescence
• Epilepsy impact an individual's health and quality of life: having seizures and epilepsy can affect safety, relationships, work, driving, and more.
• Epilepsy has an impact on the life, health, and well-being of caregivers. Family life is adapted to meet the patient's needs, and this can involve challenging life decisions such as modifying or losing employment, moving to a new place, significant financial implications, depression, sleep deprivation, and, in some cases, loss of primary relationships, and divorce.
• Current epilepsy treatments have many side effects and 1 in 4 patients are not responding to the drugs.
• Seizure's triggers (both contextual and biometric) are many and usually challenging to exploit, considering the lack of a proper recording of the pre-seizure events.
• Total epilepsy costs worldwide are up to €210 billion/year, being partly supported by patients themselves and partly by the health care system; they are mainly due to outpatient visits, hospitalizations, and ER access

By daily tracking biometric and contextual data through a wearable device (Apple Watch worn by the patient) paired to a phone app (used by parent/caregiver), MirrorHR provides caregivers insights into potential seizure triggers (diet, medication time, stress, etc.), alert them at the onset of the seizure so that it can be handled on time at home, by administering medication and providing support and helps them change habits to mitigate triggers, reducing the number of seizures, from birth to adult life. 
A 1-min video diary where the patient or caregiver explains what happened each day is processed by an Artificial Intelligence (AI) extracting the relevant data about potential triggers and recording it in a timeline of events, together with the daily biometric data contributing to personalized care. 
A secure cloud infrastructure collects all data to feed AI models and help improve seizure patterns identification for better disease management.

Epilepsy is the 4th neurological disease in the world, affecting 50M people of all ages, 6M in the EU. It is mainly characterized by a wide range of unpredictable seizures making drug efficiency vary from person to person. At times the patient dies during or after an attack, which is usually preventable if taken on time but still increases three times the risk of premature death in the general population. Seizure triggers are many (missed medications, sleep deprivation, stress, routines, etc.) and usually challenging to diagnose, as the only data that doctors get from patients is some biometrics in the hospital and a piece of paper diary with insufficient contextual data, as usually many days, are forgotten to be filled given the high level of stress and social stigma those families live with. A patient's life becomes constant stress as seizures cannot be predicted, which is also transmitted to the patient's environment, especially when the patient is a child, highly impacting on their life.

Current epilepsy drugs are not very efficient as they have many side effects and 1 in 4 patients are not free of seizures. Surgical procedures are an alternative, but they are invasive and highly risky, sometimes causing memory loss or paralysis. The Ketogenic diet may help reduce the number of seizures, but with its large number of food restrictions, following it is extremely difficult. 

Lastly, total epilepsy costs worldwide are up to €210 billion/year, being partly supported by patients themselves.

With MirrorHR, the expectations are to have an efficient, affordable, side effects free, user-friendly and non-invasive device that monitors epileptic patients, records biometric data and contextual information which is processed and shared with physicians, so that epilepsy seizure triggers are better understood. Once understood, seizure triggers will be mitigated, leading to fewer and lower intensity seizures. 

The expectations are to warn the patients suffering from epilepsy in advance of any potential seizures so that they can take measures to avoid them and live a better life without anxiety and fear. MirrorHR consists of consumer-grade hardware (smartwatch and smartphone) and proprietary apps. Its innovative and time-efficient 1 min VideoDiary concept makes collecting biometric and contextual patient data a reality. 

At the same time, hospitals, health systems, health insurers, and patients expect lower epilepsy healthcare costs thanks to MirrorHR.

Francesca Fedeli’s life (CEO & co-founder) changed with the birth of her son, Mario, who was diagnosed with Cerebral Palsy after birth. Together with her husband (Roberto D’Angelo presented below) founded the social enterprise FightTheStroke Foundation in 2014 to advocate for young stroke survivors. She is an Ashoka, Eisenhower, Global Good Fund Fellow and a TED speaker. She has worked in marketing for over 15 years for many large brands, such as ING, American Express, Nestle, and Barilla. She Is a Board Director in various Patients and Disability associations.

Roberto D’Angelo (CTO & co-founder) has over 25 years of experience as an IT expert, leading a multi-award-winning team at the international level for Microsoft. Has also gained over six years of experience in social entrepreneurship with the FightTheStroke foundation.

Gabriel Variane (Chief Medical Officer) is the medical director of the Neonatal Neuro-Critical Care program in Sao Paulo (Brazil) and has over ten years of neonatologist experience. he put his expertise to good use in starting PBSF, whose goal is to provide education and implementation of neuroprotection strategies for babies at risk of brain injury. In addition, with his medical background, Gabriel will lead MirrorHR’s clinical trials in several countries.

Jacopo Mangiavacchi (Chief Data Scientist) and Pritika Mehra (Chief Family Experience Officer) have tens of years of experience in the AI/ML fields, while Dave Hill (Chief Architect), having spent over 25 years in the blue-chip industry, will share his knowledge about the development of software apps.

Ricardo Wagner (Chief Marketing Officer) completes the picture having over 15 years in marketing, design, and business management at Microsoft. In addition, he is an expert in business solutions, accessibility, and productivity tools.

 MirrorHR has an excellent team with deep technical, medical, and commercial expertise able to design, prototype, pilot, and spread digital solutions born with patients, caregivers, and clinicians for patients, caregivers, and clinicians.

Through several initiatives (surveys, congresses, webinars, workshops), these communities are constantly engaged; they provide excellent feedback and contribute to improving the service with updated versions.

I want to impact the most significant number of people and families living with epilepsy, in order to let them be in charge again of their most useful health data and to manage their own life.

MirrorHR is the first remote seizure management service helping to reduce the number and severity of seizures for epilepsy patients, mainly focusing on the youngest ones. MirrorHR works through ML, NLP, cloud computing, a mobile app, and a consumer-grade wearable device, making it disruptive and unique. The system gives doctors and caregivers the power to gain insights into potential triggers and fully personalize alerts for seizure detection from biometric and contextual data. Contextual data (kids' environment events) is obtained via a 1 min VideoDiary, where the caregiver openly talks about the patient's day without pre-made formularies, which is then analyzed by 2 AI layers that automatically write those events on a timeline shared with doctors. Mixing biometric data (like a heartbeat, oxygen saturation, and respiratory rest rate.) with contextual data (like emotions, stress, events, and symptoms), it's possible to identify and reduce triggers to improve the patient QoL.

There are no high-tech, efficient, easy-to-use end-2-end solutions based on consumer-grade devices on the market. Apps/bracelets are trying to alert of upcoming seizures but only work in 1 out of 10 cases. They are black boxes getting biometric data you cannot control, resulting in impersonalized diagnostics, leading to false alarms and seizures not detected. Other solutions try to track potential triggers by manual diary entries. However, data sharing capabilities with doctors are cumbersome; they miss many vital triggers and take time to fill. 

On the other hand, MirrorHR's innovative solution combines biometric data obtained every few seconds from the patient's wearable with contextual insights extracted from the daily 1 min VideoDiary via a creative AI. Doctors and caregivers are given the tools to read and understand the data in real-time to personalize the alert threshold to prevent false alarms while recognizing actual seizures.

MirrorHR does not bring the cure but a way to co-live with the disease, reducing its numbers and severity with only a 1 min video per day. Initial beta testers with night seizures report they didn't need to call an ambulance in 85% of the cases and saw a reduction in several seizures between 25% and 50%.

MirrorHR will be exploited through a subscription fee. The user version of the software will be free for families (MirrorHR is the only digital device free for users in this market), and thanks to its numerous advantages, they will rapidly incorporate it into their lives and push so that the institutions incorporate it into the health system. The enterprise version, to be used by doctors to obtain all the data from the patient and to be able to give optimum recommendations/treatments, will be covered by the public health system (in those countries such as Italy, Spain, and the UK where it is possible), health insurers and other private health companies in the US and South America.

On top of developing the technology as mentioned above and knowledge, FightTheSroke plans to natively store data in the mobile device in the FHIR format. This approach is highly innovative as it puts the user in complete control of their data and enables anonymization, encryption, and management at the user level.

Once the late majority adopts MirrorHR, the innovative solution would have become the new gold standard for improving the QoL for patients with epilepsy. At this point, the technology would also represent an attractive product for the laggards.

FightTheStroke aims to make sure that all patients suffering from epilepsy can benefit from innovation in the medical industry while resisting the pressure of increasing costs. MirrorHR is an affordable solution that can drastically improve the QoL for patients suffering from epilepsy, helping them to keep their seizures under control with a bright, innovative, risk-free solution.

FightTheStroke has a comprehensive clinical trial and regulatory strategy to ensure compliance with all regulations/standards. It will finish preparing the documents necessary for the CE Mark & FDA approval, including technical information (specifications and validation data), the study folder, device master record documents, work order specification documents, and risk analysis.

MirrorHR will be commercialized in 2024 in Italy, Spain, and the UK. By the end of 2027, it plans to cover most of the EU market. 

Later on, in 2028, India, Brazil, the US, and Canada will be sought, considering the high number of patients suffering from epilepsy. 

In order to cross the chasm and target the early majority, FightTheStroke has defined a comprehensive dissemination and commercialization strategy. Firstly, efficient high-level communication will be implemented to amplify the impact of the project’s results and increase the visibility of the project at the EU and international levels. The planned communication actions will include online marketing through website and social media and participation in industry-related events.

In addition, FightTheStroke will study foreign markets & define a go-to-market strategy for each region, setting priorities and time frames. At the same time, the company will identify the resources (new offices, hiring, communication material) necessary to launch its product in other markets.

FightTheStroke solution has a solid scale-up potential, its potential impact being, therefore, game-changing.

• Decrease 50% risk of a seizure within 12 months
• Decrease Epilepsy health expenditures five times within 12 months
• Users satisfaction: positive feedback evaluated through ad hoc surveys for patients/caregivers and clinicians
• Improvement of the QoL of patients and caregivers evaluated through an ad hoc methodology that uses an adequate QoL scale

Insight and Overarching Goal: Patients with epilepsy and their caregivers are under constant stress, anxiety, and confusion, and this stressful situation has an impact on their life, health, and well-being of all. Family life is adapted to meet the needs of the patient, this can involve challenging life decisions such as modifying or losing employment, moving to a new place, significant financial implications, depression, sleep deprivation, and, in some cases, loss of primary relationships, and divorce. MirrorHR helps reduce family stress factors, therefore impacting seizure triggers while collecting relevant data for therapy compliance and research.

Activities: 

1) Patients/caregivers input their daily data through the video diary on the mobile phone. 

2) Patients wear regular watches that can constantly measure vital parameters and send seizure alerts. 

3) Patients/caregivers download their data and transmit them to the doctors before the monthly checkup.

Outputs:

1) Patients/caregivers collect, store and measure their health data.

2) Patients and Caregivers reduce their stress levels in their everyday routine. 

3) Doctors can rely on trusted data for adjusting the therapy, avoiding triggers, and researching potential new drugs.

Outcomes:

1) Patients are in charge of their well-being through an everyday data collection with the video diary.

2) We ignite a behavioral change in patients' habits.

3) We enhance patients-doctors communication through a reliable exchange of information.

----> reduction of seizures and risk of SUDEP, improvement in developing new therapies for epilepsy.

MirrorHR augments the biometric data collected directly from the patient with crucial contextual information for understanding the triggers that increase the probability of seizures, thanks to its intelligent diary feature implementation. A 1-minute video diary is recorded daily by the patient/caregiver, where details of what happened during the day are offered. Its speech is then transformed by an ML model and is processed by another ML model to extract the relevant data about potential epilepsy triggers. All the data (video diary + biometric data) is chronologically organized into a calendar of events (i) to provide doctors with a complete data overview of what is going on inside and around the kid in every single moment, to give insights on potential seizure triggers, identify and prevent them from reducing seizure number and (ii) to train the ML model to improve seizure detection and trigger discovery. State-of-the-art ML techniques in Natural Language Understanding capture entities, labels, and associated “quantitative and qualitative data” to build a calendar of events, which are analyzed to get potential triggers and recommend behavioral changes to reduce their impact. Zero-Shot Learning helps to accelerate the experimentation phase. A TensorFlow BERT (Bidirectional Encoder Representations from Transformers) QnA model trained on the SQuAD (Stanford Question Answering Dataset) dataset is used for the front-end data extraction. In contrast, a dedicated classifier and NER (Named Entity Recognition and Classification), scored in parallel, are used for the back-end classifier.

From the pilot to the hackathon till now, we've always hired a cross-functional team with high attention to the level of diversity, starting from the gender of our female CEO: now, in our working teams we have persons with disabilities and different conditions of neurodiversity, geographically dispersed, and we will continue prioritizing the hiring of persons with disabilities even in the future.

MirrorHR improves the quality of life for patients suffering from epilepsy by offering an innovative, high-tech solution that not only detects seizures but also reduces their number and intensity. It uses commercial-grade hardware, and thanks to its 1 min video diary function and proprietary AI algorithms, it can help shed light on epilepsy triggers while offering peace of mind to epileptic patients.

MirrorHR will be exploited through a subscription fee. While the user version of the software will be free for families with members suffering from epilepsy, the enterprise version, which will be used by doctors to obtain all the data from the patients and to be able to give optimum recommendations/treatments, will have an annual price per patient of €480, representing 13% of the savings it generates to the healthcare system by reducing number and severity of seizures.

MirrorHR will work with partners such as PBSF, FightTheStroke Foundation, and International Alliance for Pediatric Stroke, who will act as a communication channel to reach customers such as hospitals and clinics, public health systems, and health insurers. We will have a committed support team to help customers with training. 

MirrorHR’s technology was developed during the Microsoft 2019 Global Hackathon and Microsoft’s 2020 Garage. Some milestones achieved are 1) several versions of MirrorHR’s patient app have been successfully built; 2) the technology has received positive feedback from beta end-users; 3) FTS found partners for clinical trials in Italy, Spain, the UK, and Brazil.

MirrorHR expects rapid growth, considering that the innovativeness of its solution is in line with the current market trends and that it has received excellent feedback from beta end-users from around the world. 

An effective communication strategy will contribute to the growth of sales. Once proven in Europe, the solution will potentially be used on a global scale, with a second expansion to India, Brazil, the US, and Canada. 

The business proposition is inherently scalable since it is a software product installed on commercial-grade and widely available hardware. Furthermore, FTS intends to invest in R&D to maintain its technological edge over its direct competitors.

MirrorHR won the Microsoft 2019 Global Hackathon, the largest company hackathon in the world, among +9000 projects, leading to a six-month incubation period in Microsoft’s 2020 Garage Residency program (1Ml USD ca as estimated value). Microsoft’s CEO, Satya Nadella, particularly encouraged MirrorHR to move forward. The technology has been proven successful: hundreds of families from over 31 countries using the app (available in the App Store) have provided excellent feedback. Currently available in 13 languages.