Life Image made a name for itself in the image sharing space. In recent years the company has sought to play a vital role in the evolving landscape of AI applications for healthcare.
In an interview Janak Joshi, Chief Technology Officer and Head of Strategy with Life Image, talked about some of the milestones Life Image has reached and how it is collaborating with technology vendors to address some of the data challenges of AI adoption in healthcare.
Joshi will be taking part in MedCity’s CONVERGE conference on innovation in cancer treatment June 19 in Philadelphia. He will be on a panel discussing the data problems that make widespread adoption of AI in healthcare complicated.
Medical image sharing is a popular segment of the digital health sector these days. But what was the landscape like when your company launched 11 years ago?
Eleven years ago, our ability to digitize medical images at scale in a consistent way was non-existent. Further, these images were siloed and not a part of your core medical record, despite its critical role for many point-of-care decisions. Medical images were not easily accessible or readily available during routine clinical and treatment decisions. There was no widely adopted, interoperable, vendor-agnostic provider network or a standards-based process or technology that allowed medical images to flow across health systems. And these data silos of medical images proved even more difficult for patients to break down when they need to access their medical images throughout their individual patient journey. This negatively affected care coordination and patient outcomes. A burden was placed on patients to courier their uninterpretable images, which were given to them by imaging modality manufacturers, not their medical providers. The lack of digitized medical images, including pathology, radiology, and cardiology significantly increased the cost of healthcare as patients underwent repeat and unnecessary additional testing. There was a general lack of collaboration and transparency among manufacturers.
Life Image recognized a critical need to solve this chronic and widespread problem in healthcare. Recognizing that digitizing and exchanging medical images is complex and expensive, Life Image was one of the first in the industry to solve transactional image exchange while building a vendor-agnostic network to scale this complex problem, not only nationally, but globally.
A Deep-dive Into Specialty Pharma
A specialty drug is a class of prescription medications used to treat complex, chronic or rare medical conditions. Although this classification was originally intended to define the treatment of rare, also termed “orphan” diseases, affecting fewer than 200,000 people in the US, more recently, specialty drugs have emerged as the cornerstone of treatment for chronic and complex diseases such as cancer, autoimmune conditions, diabetes, hepatitis C, and HIV/AIDS.
Janak Joshi
It can be challenging for health tech companies to develop software to support clinical workflows without [unintentionally] making physicians lives more difficult. How did your company solve that problem?
Historically, the replicability and reproducibility of clinical workflows during software install has been a complex and expensive service-based offering by large health IT vendors.
Life Image has spent the greater part of the past 10 years investing in a highly configurable and personalized network framework that accommodates any workflow from any specialty and across vendor systems enabling interoperability of those workflows at scale, using common standards. This network framework has allowed Life Image to accommodate existing clinical workflows for clinics, hospitals, and tertiary care centers without adding additional operating burden by introducing new workflows, training requirements, or new areas of potential variability.
Attend MedCity CONVERGE to hear from healthcare innovators like Janak Joshi and other experts. Use promo code MCN50 to save $50. Register now!
One area of interest you have identified is neurology. Can you offer a couple of examples of how your software can support clinicians and patients in this area?
Patient care is one of the most important areas our network impacts. For example, Life Image is responsible for managing the largest stroke network in the United States today. Our network manages care coordination and care delivery for thousands of stroke patients across the country. The ability to treat stroke patients is highly dependent on the networks capability to move critical clinical information during the patient transfer process as well as the ability to integrate directly inside the neurologists’ workflow without wasting valuable cycle-time for a potential intervention. This is only achievable with a broad network that integrates multi-vendors and workflows across institutions.
We are stepping into the next generation of innovation in neurology by servicing AI and machine learning companies that are working to solve time and accuracy issues in diagnostics. There is a fundamental gap that AI and machine learning companies experience as they work to scale their algorithms insofar as they have been unable to expand at scale and integrate in the clinicians’ workflow. By allowing access to our network, Life Image can service AI companies as a distribution network that influences physician behavior and treatment decisions to improve outcomes.
Another example of the Life Image network supporting clinicians is with access to real-world evidence. Our network delivers novel data sets to help train AI models and help identify imaging biomarkers during the drug development process across heterogeneous data sets that represent a varied patient cohort across the United States and globally. This enables Life Image to mature the quality of evidence needed to investigate the correct intervention and treatment pathways, improving patient care and outcomes.
What are some of the milestones in your company’s development that you’ve hit in recent years?
Over the past 11 years, we have grown into a network company that advances the industry through data democratization and access to clinical information. As an organization, we bring novel, heterogeneous datasets together for AI, pharmaceutical, and academic medical research to discover novel therapies. Life Image is the first in the industry with the ability to anonymize, link, curate, and label these novel datasets across oncology, neurology, multiple sclerosis, and musculoskeletal. We have scaled the network internationally, now servicing thousands of clinics and physicians across Europe, the Middle East, and Southeast Asia, a proof point that network expansion is critical beyond the United States and an important need for every major economy.
In the past couple of years you have started working with AI technology vendors and health systems developing their own AI tools. What are some of the data challenges you/these companies have encountered? How do you address them?
Almost all AI companies have trained their models on publicly available data sets that either represent a single homogeneous vendor, or data that has been collected from a single institution. AI algorithms are typically best adopted and most sensitive when they are trained on heterogeneous data sets representing multiple care pathways and manufacturers. Further, these datasets are generally outdated and for training certain types of models, no longer relevant. Access to volumes of broad data is limited without access to a network. Life Image is bringing a heterogeneous dataset to this industry to help accelerate and improve the sensitivity of the algorithms’ output and improve the validity of the arguments presented by these AI companies.
You have described medical imaging data as the “missing link” in AI. Can you explain what you mean by that?
Medical imaging biomarkers are represented as a critical path in most drug development pipelines, specifically oncology, cardiology, cystic fibrosis, multiple sclerosis, and musculoskeletal associated therapies. Medical imaging is critical to treatment decisions — for example, investigating for tumor growth against therapy responsiveness. It is important for the medical community to recognize the importance of including medical imaging data across the patient’s journey.
What are some examples of the AI applications you are collaborating on?
Life Image partners with many technology companies, creating strategic partnerships that add value to the global network. We recently partnered with Mendel.ai to advance the use of their AI algorithm, which supports life sciences and research facilities by analyzing clinical data to identify patient cohorts for a given protocol or clinical trial.
As a Google Cloud partner, we bring novel opportunities to care networks and accelerate understanding of the clinical utility of medical evidence across a network of physicians, medical imaging providers, research, and most importantly patients.
Looking ahead, do you have any interest in expanding the kinds of companies with which you are collaborating? Which segments do you find the most compelling?
Our collaboration with Google as a marquee partner continues to expand across various different function areas, specifically related to improving how we continue to expand our network globally and how we service AI and machine learning companies with better data sets. As our partnership expands, we continue to leverage Google’s infrastructure capabilities for scaling growth.
A second compelling segment Life Image continues to expand is helping pharmaceutical companies access, govern, curate, and link data sets across a heterogeneous network nationally and internationally and throughout their product life cycle.
Lastly, one of the most compelling and important areas we strive to positively impact is oncology. Our reach within oncology continues to develop, whether it’s through care delivery, AI led diagnostics or drug development. We recognize the significant unmet clinical need in oncology in the United States and strive to influence care coordination to ultimately positively impacting patient lives. We do this through both network access to data and strategic partnerships.
Photo: Getty Images
