Crafting a Centralized Sandbox for Healthcare Data

Crafting a Centralized Sandbox for Healthcare Data

By John Axerio-Cilies, CTO and Founder, Arterys

John Axerio-Cilies, CTO and Founder, Arterys

Benefits of Cloud Computing 

The benefits of cloud computing are substantial, but care must be taken to analyze the pros/cons of any decision before diving into cloud vs. on premises computing. Factors such as utilization, data size, privacy and security, as well as accessibility can sway decisions in either direction. Within cloud computing, we must differentiate the benefits between private and public cloud infrastructure; there are pros/cons to both. The benefits of a public cloud are that you don’t pay for what you don’t use. You have the full flexibility of dynamic horizontal and vertical scaling. In a private cloud setting it is often difficult to predict how much hardware is required. Many organizations either underestimate (meaning that more servers need to be provisioned at a later point in time, which adds to the cost) or overestimate hardware requirements (leading to low utilization).

“They don’t care how you do it, they just need results, take the time to plan over the next 6 months to 1 year”  

When it comes to providing data to hospital administrators, application developers, researchers or scientists the cloud provides a centralized sandbox where all healthcare data is in one place. When this data is located in one place it makes it very easy to compute various analytics on this data. Creating dashboards for administrators that provide real-time insight into usage and logging is extremely valuable in providing a reliable service. Of course this can be achieved with an on premises server based application, but complexity arises when healthcare providers are buying up hospitals within the region or even in completely different parts of the world. Support and training can be tailored to specific individuals with much deeper insights into user preferences such that custom workflows can be created for specific users. With cloud computing there is no longer a need for maintaining and servicing local hardware. No power, cooling, and system management facilities are required. In the event of a natural disaster, multiple availability zones scattered around the world can easily direct web traffic to non impacted zones allowing for greater than 99.9999 percent uptime. 

We recognized back in 2010 that there was going to be a revolution in cloud computing specifically as it related to the healthcare industry. We took a chance and decided to build out a bleeding edge cloud platform that was able to support several clinical modules (aka workflows). Not a day goes by where we regret this decision. I’ve seen several companies unsuccessfully try to transition from workstation based software to cloud based software. The cost and effort involved in re-architecting the system to support a cloud based platform is daunting. Our full software stack is built from the ground-up to support thousands of simultaneous users accessing data from all regions of the world. 

Effective and Proactive Use of Data 

At Arterys, data is king. We use data to drive all aspects of the business. We would not be where we are today if none of our customers were willing to share data. Since we collect all kinds of multi-modal data (session data, image data, patient data, network data, and user data) we need specific algorithms to mine this data effectively and find insights that help drive the business forward. That’s where our machine learning team comes in. We have a dedicated team of data scientists that is their full-time job to make use of this data and find analytics that help either internal or external customers. To give a concrete example, our engineers use deep learning to automate the identification and segmentation of medical images; this reduces the time radiologists spend annotating these datasets, which in-turn allows them to diagnose more patients (higher revenue and reduced cost to the hospitals).  

Role of CIO in Recent years 

Over the past couple years CIOs and CTOs are focusing a lot more attention on security and privacy in the healthcare sector. Recent high profile breaches are highlighting how crucial these two aspects of the business are in making people feel comfortable sharing data seamlessly across a network. Organizations see the opportunity in the cloud but are hesitant to dive right in. That’s where it is critical to architect a platform where security and privacy are the top priority.  

Advice to Fellow CIOs 

In terms of unique lessons I’ve learned from leading our IT organization I would say that it’s easy to get caught up looking in the rear-view mirror at the competition or trying to use the latest and greatest technology buzzword. The truth is that the focus should be in front of you and beyond the horizon. Ask yourself, what am I missing? What are the unknown unknowns? The focus needs to always be driving value to your customers. They don’t care how you do it, they just need results, take the time to plan over the next 6 months to 1 year. It’s too easy to get distracted with day to day operations; carve out some time on a weekly basis to forecast and innovate or else you’ll miss the iceberg right in front of you that you’re about to hit. 

 IoT in Pharma and Life Science Industry 

The internet of things for the healthcare industry means the seamless movement of data back and forth between devices and the cloud. To start, the transition to IoT will be passive in a sense where the data flowing back and forth will primarily be used for monitoring, servicing, and improving performance through analytics. There will be a transition over the next few years where devices will be able to sense and be controlled by other remote users or devices thereby allowing any individual device to be as knowledgeable as its creator. We are seeing this change today with MRI machines; scanners used to be disconnected from the network. Now these machines are getting smarter by exporting usage data as well as image data to centralized databases for advanced analytics. In addition, these machines can be controlled remotely in the case where specific local domain expertise is missing to accurately diagnose a patient.     

Weekly Brief

Read Also

The Pandemic and patient experience Innovations

The Pandemic and patient experience Innovations

Kathy DellaMura, VP of Clinical Services, Hartford HealthCare
The Future Of Healthcare Is Digital, Telemedicine Destined For Greatness

The Future Of Healthcare Is Digital, Telemedicine Destined For...

Mitchell Fong, Director of Telehealth for Renown Health in Reno, Nev.
Medical Spine Care: Have We Finally Arrived?

Medical Spine Care: Have We Finally Arrived?

Michael Geraci, Medical Director Comprehensive Spine Program, Catholic Health
Artificial Intelligence and the Lost Art of Auscultation

Artificial Intelligence and the Lost Art of Auscultation

Edward Kersh, Medical Director, Sutter Care
Tools and Capabilities Required for Value-Based Care

Tools and Capabilities Required for Value-Based Care

Mark Weisman, Chief Medical Information Officer, TidalHealth System