NEW YORK (GenomeWeb) – Dell said recently that it has extended its Cloud Clinical Archive, its solution for storing and managing medical imaging data, to include genomics data.
The company said during this year's Healthcare Information and Management Systems Society meeting that it will begin allowing customers to store BAM and variant call files, for example, in the Cloud Clinical Archive starting in April. The archive currently holds more than 11 billion medical images and around 159 million clinical studies from multiple healthcare providers. The company intends to make the same hosting and management capabilities it used for images available to existing Clinical Archive customers who are looking for a secure location to store their data.
"There's a whole technical catch-all around storing these images which is not dissimilar really from what is required in the genomics space," Carrick Carpenter, global solutions director for Dell's Healthcare Cloud, told GenomeWeb. "We've built out IP ... that can handle very large images and millions and millions of them at a time, [but] beyond that we can guarantee availability of images across data centers, and we have rulesets in place for pre-fetching [images] when required [and] minimizing latency."
Opening up the Clinical Archive to genomic data provides Dell customers such as the Translational Genomics Research Institute with a low-cost, scalable alternative to local storage infrastructure. "So instead of buying large storage arrays, we are offering [customers] an alternative to store [data] with us," he said. "They can keep the images they need for their current workflows and pipelines local and then at the point the research is done, they can archive them off into our environment." The company charges a fixed rate per file for storage that depends on the size of the file and how much files the customer needs to store.
Longer term, Dell's goal is to enable clinicians to combine insights extracted from medical images with the output from sequencing instruments to inform patient care, Carrick said. Dell's vision is "to be able to link the output from our imaging analytics to the genomic identifiers of a patient to determine if there is not only a consulting preventative care program we can put in place, but also a genetic marker that we need to be able to add into a patient record."
Late last year, Dell put one more of those building blocks in place when it collaborated with Zebra Medical Vision to provide more fine-grained analytics of medical images. Specifically, Dell announced a multi-year agreement with Zebra Medical to integrate its imaging analytics and research platforms with the Cloud Clinical Archive.
Zebra's research platform provides access to anonymized and indexed patient imaging studies and reports along with associated clinical information. Its analytics tools provides algorithms that help clinicians assess things like calcium buildup in the heart or bone density; identify patient populations that are at risk for ailments such as osteoporosis, cardiac disease, and liver disease; and potentially enroll them earlier in preventative care and wellness programs. For example, a clinician could identify patients at risk of breaking a hip because of low bone density using the tools that Zebra provides.
However, a patient might also have genetic factors that predispose them to bone breaks, for example. Alerting clinicians to the presence of predictive markers early on enables them take steps to minimize those risks, Carpenter said. Another example comes from pediatric cancer cases. Some children treated with chemotherapy develop heart issues post-treatment that could be due to the treatment does or be the result of a genetic predisposition to heart disease. Knowing up front that a patient has markers that predispose them to heart trouble could alter the treatments and dosages that they receive. "That's really where this announcement is coming into play," Carpenter said. "By marrying these data sources, we think we are going to be able to do ... imaging and genomic tie-outs to add more value."
Dell is collaborating with customers such as TGen and commercial partners such as Appistry to store genomic information "at any phase after it exits the sequencer and comes into our environment," Carpenter said. The company has enjoyed a long-term relationship with TGen and has provided both grant funding and infrastructure to support pediatric cancer research projects and trials run by the institute and its collaborators.
Dell has worked with Appistry in the past to provide a bundled hardware/software solution. The product is a Dell Workstation or server that comes pre-loaded with GenomePilot, an Appistry solution that provides preconfigured genome analysis tools and pipelines. GenomePilot helps clinical laboratories identify variants from raw sequence from exome, whole genome, or gene panels, as well as data from tumor/normal and tumor-only cancer studies. Dell also signed an agreement with the Innovation Institute, a healthcare incubator, in 2015 to run pilot programs focused on things like extracting actionable information from healthcare data and better engaging with patients.
Dell is also partnering with Saffron, an Intel company, to incorporate cognitive compute capabilities developed by Saffron to help Clinical Archive users glean additional insights from their data. However, its efforts are still very much in the early stages, Carrick stressed. "We are in the crawl stage and we are moving to walk. We are not near running yet," he said.
However, "we believe firmly that this is where the market is going," he continued. "We think there is tremendous value in the images that institutions have already paid for and are sitting nascently in their or our data centers. This is an emerging area." Much of the interest in Zebra's products has come from risk-based health organizations while fee-based organizations are less interested, according to Carpenter.
Interest in integrating genomic data with medical images, however, might take longer to catch on in the healthcare community. "I think we are a little bit ahead of the curve on the genomics side," Carpenter said. "There is interest [but] people are still trying to understand what to do with it."
Research organizations have the most interest. Organizations like TGen have embraced the marriage of imaging and genomic data. However, hospitals are still trying to figure out the return on investment for genomics. "There are a few hospitals that have as a standard of care adopted genomic analysis but there are so few right now. I think the provider space especially is going to have to move towards genomics being part of the standard of care before this really takes off," he said. "But given that we have ... specific use cases we want to get started now."
Dell's activities in the genomics space also include hardware designed with clinical use cases in mind. Last year, the company announced that it had upgraded its hardware platform for genome analysis. The so-called High Performance Computing System for Genomic Data Analysis can process around four times as many genomes per day as the first version of the system and cuts down on analysis times for samples at 10x coverage by 40 percent.