Skip to main content

iTRUST: In-time TReatment of acUte ischaemic STroke

Periodic Reporting for period 1 - iTRUST (iTRUST: In-time TReatment of acUte ischaemic STroke)

Reporting period: 2017-02-01 to 2017-07-31

In Acute Ischaemic Stroke (AIS), a blot clot suddenly blocks a major vessel supplying the brain with oxygen, killing 1.9 million nerve cells (neurons) every minute. Thus, “time is brain” in AIS and it is therefore critical that the patient instantly receives the best treatment. At the same time, stroke development and response to treatment are medically complex and depend on a vast range of patient-specific characteristics, which puts a tremendous pressure on hospital emergency staff. In the emergency setting of AIS a plethora of information must be considered by a specialized physician including imaging, clinical presentation and electronic health record data before safely committing the optimal therapy. For certain stroke profiles clot-busting (attempting to unblock the artery) with pharmaceuticals or surgery has optimal prognosis, whereas for others this implies a high risk of an intracerebral haemorrhage. Specialized stroke units are, however, very scarce in EU and worldwide. Indeed, only 5% of hospitals have facilities meeting the standards of comprehensive stroke centres, which reflects the fact that only about 30% of AIS patients receive specialized stroke unit care across Europe. Annually, 15 million people worldwide suffer a AIS leading to 5 million deaths and 5 million people permanently disabled.

The iTrust solution by COMBAT Stroke enables any hospital to instantaneously provide state-of-the-art management of emergency patients with AIS. iTrust is a clinical decision support system utilizing advanced medical imaging and clinical data from thousands of patients together with new technology to identify complex risk profiles in large unstructured data using dedicated deep learning technologies developed at COMBAT Stroke.

Today only 7% of patients with AIS are treated with clot-busting drugs and it is estimated, that even for ideal candidates up to 40% are not treated. The iTrust solution not only reduces time to treatment and guides the optimal medical intervention, but also promotes treatment in general through automated management. Moreover, most AIS patients arrive too late at the hospital to be considered candidates for therapy under today’s guidelines (4.5 hours). The iTrust solution has the potential to extend the treatment window by objectively assessing each patient’s individual potential for favourable outcome even with late treatment. iTrust can therefore substantially increase the number of patients receiving active therapy.

Combat Stroke forecasts to achieve cumulated revenues of over €41M and EBITDA of €30M within five years after market rollout. The feasibility study represents a key step towards the implementation of our overall innovation project, the primary goal of which is to validate the solution in patient cohorts, obtain CE-marking and finalize EU and US market strategy and business model. The project will also involve technology refinements in ease-of-use, such as user experience (UX) design, software/user interface. Through a methodical market survey, the demand for iTrust was clearly validated. Our study has shown that the above ambition is realistic if a focused route-to-market strategy is followed.
During the feasibility study, we have: • Completed a detailed market investigation; • Defined our ideal business model to be a shared revenue model with scanner vendors to ensure swift monetization of the iTrust solution. The market has taken a turn towards these ‘medical app-stores’, which all major vendors will be pushing quite hard within the next 12 months, including Siemens (Syngo.via) Philips (Intellispace), and GE (GE healthcloud). Combat Stroke is in the fortunate situation that we are in contact/negotiations with all these vendors (and others), and will be part of the first wave of third-party applications on the platforms; • Documented a comprehensive quality and regulatory strategy for iTrust, which implementation is ongoing and to be finalized before the planned demonstrations in Phase 2; • Further developed and enhanced our business plan based on a thorough market survey performed during the feasibility study.

Roughly, the main results from the action are:

• Non-contrast Computed Tomography (CT) is the most used imaging technique in the acute setting of suspected stroke with respondents making use of the method in 86% of all AIS cases. Combat Stroke must focus attention in CT in the short-term to reach a wider market penetration with the iTrust software.
• Magnetic Resonance Imaging (MRI) is not frequently used in the acute setting of stroke with respondents accusing an average frequency of use (in % of cases) of 34% for diffusion weighted and 23% for perfusion weighted.
• Nearly 94% of respondents believe that the iTrust solution will reduce patients’ complications (further infarction) by accelerating the decision on the most appropriate treatment.
• There is a consensus surpassing 87% of respondents that the iTrust solution will increase confidence in the treatment decision taken by the attending physician.
• Importantly, most (75%) respondents believe that iTrust solution will open for thrombolytic treatment decision in hospitals without stroke specialists.
• The acknowledged most valuable (scored at 4.5 out of 5) proposition from the iTrust solution to KOLs is “a software that is easy to use, with minimal interaction”. Therefore, in this project we will pay close attention to improve the usability and ease of use of the software before market launch.
• Equally valuable (4.3/5) is the capacity of the iTrust solution in using CT images as input information, reinforcing the strategic importance of enabling CT-based workflows for iTrust.
• The survey conformed the unique selling points for the iTrust:
o A software that shows the location and extent of tissue predicted to infarct as a function of different treatment strategies; with nearly 70% of respondents recognizing the feature as real innovation.
o A software that uses both epidemiological and clinical markers as the underpinnings of a predictive algorithm; with over 70% of respondents acknowledging the disruptiveness of the feature.
The acknowledged more appropriate commercialization strategy for the iTrust software should be via a yearly subscription fee; whose preference was stressed by over 85% of respondents.
The results of the analysis overall confirm the suppositions set forth in the SME phase 1 application, namely that:
1. The classification for the iTrust product in the EU is CE class I.
2. The classification for the iTrust product in the US is FDA class IIa, under the product code LLZ.
In the current workflow – whether using a support software or not – the triaging of patients uses only a single fixed value (the patient’s own data) to determine the best treatment. Hence, state-of-the-art does not make use of the vast potential in continuously learning from previous and new patients, and falls short in integrating the large number of possible scan images. Another shortcoming of current methodology is the reliance of ‘hand-crafted’ imaging predictors. Whereas medical images contain highly complex information, manual identification of information-carrying features (such as high or low image intensities in specific brain regions) is sub-optimal with proportional loss in predictive ability. The iTrust solution consists of a radically different approach: Our solution uses both epidemiological and clinical markers as the underpinnings of a predictive algorithm. While in training, iTrust’s algorithms automatically extract potentially complex image features, which are informative of the outcome. Because of the automatic identification of predictors through Convolutional Neural Network, our solution is a giant leap forward in extracting complex/valuable information when compared to handcrafted predictors.

In the hospital setting, iTrust’s predictions will assist the physician in choosing the individually most advantageous intervention and ensure the patient receives the state-of-the-art treatment. It is estimated that up to 7 million EUR could be saved per year per each additional 2% of patients treated timely with the state-of-the-art. On top of those socio-economic benefits, iTrust provides an attractive business case with sound and realistic economic benefits for a start-up company such as Combat Stroke that strives to reach sustainable revenue through organic growth. In addition, the solution can be the propeller for a range of additional, positive societal benefits such as job creation – with an expected 24 additional employees at Combat Stroke five years after market launch.

The commercialization of the iTrust solution will have the following socio-economic impact | New job positions at Combat Stroke):

2020 – € 750,000 | 2
2021 – € 2,250,000 | 5
2022 – € 6,500,000 | 10
2023 – € 11,750,000 | 16
2024 – € 20,000,000 | 24
Prediction of likely brain damage for different pathways in an AIS via iTrust.