In 2019, GE Healthcare started Edison ^[X] – a startup collaboration programme centered on the Edison platform, intended to work with startups addressing some of the toughest healthcare challenges in the country. Recently, the company hosted a virtual Demo Day for the first cohort of startups, even as the groundwork for the second cohort is well underway.

Nritya Ganesh, Program Director, Edison^[X], GE Healthcare, South Asia talks to INDIai on leading a healthcare programme like Edison ^[X] in India, building a collaboration-based ecosystem for innovation and how AI is a huge component of this endeavour.

Tell us about Edison^[X] and why GE Healthcare decided to start this platform 

At the core of the Edison^[x] program is the idea of leveraging collaboration to compound the capabilities of GE Healthcare’s Edison platform. Edison is an advanced intelligence offering used by GE Healthcare’s internal developers and strategic partners to develop new healthcare applications, services, and AI algorithms. It comprises of applications, smart devices, and the Edison platform. The aim of the Edison[x] program is to bring in competent startups who can build their solutions using this platform resulting in mutual benefit to all three stakeholders – GE, the start-up in question, healthcare practitioners and most importantly, the individual consumer. The aim of the program is to work with the start-up community to accelerate product development and reach new markets while building trust within the start-up ecosystem.

How does Edison^[x] help start-ups reach the next level of competence? 

The Edison[x] program helps its start-ups at multiple points in the engagement lifecycle. The inputs can mainly be bucketed under the following heads:

Product development

Senior scientists and product managers from the GE Healthcare R&D and Project Management functions are assigned to the selected start-ups at the start of the program. These scientists engage with the start-up’s product development research team on a regular basis to fine-tune algorithms and help them integrate their solution on the Edison platform.

Market access

We help the start-ups connect with our large customer base to help them validate their solution. We are also engaging with the start-ups on possible commercial collaboration and giving them access to newer markets.

Global expansion 

Depending on the relevance of the solution, the GE Healthcare India teams help the start-up reach other GE Healthcare locations worldwide, with a view to commercialize the solution in the said geography.

Edison^[x] thrives on the concept of collaboration; how challenging is this in India? 

The Edison[x] program was launched in India due two reasons: Presence of our largest, integrated, multi-disciplinary R&D centre outside of the USA in Bangalore and a strong and mature start-up ecosystem. Our experience through Cohort 1 has confirmed our belief that launching the program from India is a win-win for both sides. Key to the success of any collaboration is coming together of the complimentary strengths and trust in each other’s capabilities. GE brings to the table global know-how, scale, customer insights and decades of MedTech domain expertise. Start-ups on their side bring agility, deep technology knowhow, differential thinking and a can-do attitude. We saw this combination work very well in Edison ^[X].

The program is leveraging health tech start-ups with core AI capabilities - can you tell us why you think AI in healthcare is important, especially to address a crisis like COVID? 

Technology can play a critical role in driving access and affordability in healthcare, improve accuracy and speed of diagnosis and reduce the dependence on scarce resources like a specialist doctor with significant clinical experience. It is a known fact that the accuracy of interpretation of results of radiological scans and consequent diagnosis correlates strongly with the amount of experience of the clinician reading the scans. This makes it very hard to train new doctors to speed up this process and bring them up to equivalent accuracy levels. This dependence on a handful of senior specialists has resulted in turnaround times of the order of 48 hours from scan to diagnosis, which can be very problematic from the perspective of Covid-19 pandemic containment, as asymptomatic patients could meet and infect a large number of people in the period between scan and diagnosis.

In the context of Covid-19, there is the additional angle of “touchless” care to protect frontline workers. Again, AI solutions deployed at the site of screening can help with reducing human contact with healthcare workers while also reducing the turnaround time for diagnosis which is critical to disease containment as mentioned earlier.

The reducing price of storage and computing power as well as the shift to digital media over the years now makes it possible to bring in automation in the form of AI and Machine Learning algorithms to process clinical data and generate accurate and repeatable diagnoses. This makes a strong case for the inclusion of AI algorithms into clinical workflows.

Caption: AI-guided radiology startup Synapsica's CEO Meenakshi Singh presents during Demo Day.

How are you augmenting start-ups with GE's inhouse expertise in medical device tech & AI? 

Help with fine-tuning algorithms

The technical mentors in the Edison[x] program have helped start-ups identify specific use-cases where maximum value is likely to be derived from their solution. Further, the mentors have worked with the start-ups’ developers on tweaking the AI algorithm by making training data available to them. This could be de-identified/ anonymized real data or phantom/ synthetic training data that is generated by an AI engine to mimic typical real-world data while also incorporating typical natural variation.

Help with integration of the solution with the Edison platform

The next step towards deployment comes with integration of the start-up’s solution on the Edison platform and the Edison Health Link boxes deployed at individual customer sites. The GE technical teams worked with the start-ups on this integration, which allows the solution to also be bundled along with other GE offerings.

Help with deployment in virtual sandbox

The final step involves pilot deployment at customer locations where testing of the solution on real world data, back-to-back with traditional processes, gives the start-up invaluable feedback on the on-ground realities of the effectiveness of their solution.

You have successfully completed one cohort, and the second one is about to begin soon... What would you say are your key takeaways so far? What do you expect to achieve in the 2nd cohort? 

We are excited by our journey so far and believe there are success stories, with tangible business outcome for start-ups and GE Healthcare. We do intend to intensify our engagement with the Indian start-up ecosystem.

In the second cohort we will be looking at a few themes primarily,

• Advanced Visualization and Reporting
• Virtual Hospitals
• AI for Operational and Clinical efficiency

We will also be looking at more mature start-ups who have a soundproof concept and are reasonably close to applying for regulatory approval, in order to accelerate the process of deployment in the market.

Some of the start-ups in your first cohort were developing solutions in line with the demands put forth by COVID19. Your thoughts on the tech/solutions being developed?

In response to the Covid-19 pandemic, two start-ups from the first cohort – Predible Health and 5C Network collaborated to produce solutions to automate the reading of CT Scans from suspected patients resulting in earlier diagnosis. Predible Health already had a solution called LungIQ which was being used to diagnose liver and lung cancer but could be quickly modified to look for signs of Covid-19

Till date, the collaboration has looked at over 800 probable cases with a positive incidence rate between 40 and 50%, and a turnaround time of about 40 minutes.

Going forward, what are the main areas GE Healthcare would like to see innovation in healthcare?

At GE Healthcare our vision is to be ‘the leading innovator enabling precision health’. Our innovations are focused around making that vision a reality. Given that, there are two themes of innovation that the Edison[x] program will focus on:

The first is Advanced imaging, visualization, and reporting. This would involve building solutions for CT Scanning and X-Rays, leveraging Edison capabilities for Computer Aided Detection and Diagnosis. This could also be extended to diagnosis of certain cancers.

The other theme is that of Virtual Hospitals – coordinating care between various healthcare providers, building capability in remote patient health monitoring, and better integration of data and workflows.

Both these themes have operational underpinnings of the use of AI for building operational and clinical efficiency for better patient outcomes and financial viability, while also ensuring safety for frontline health workers. The second cohort of Edison^[X] will broadly centre around these two themes.