An ever-growing amount of healthcare and life sciences data is generated from a variety of sources, including medical practices, laboratories, pharmaceutical companies, payers, electronic medical record and imaging systems, and the proliferation of IoT devices. Failure to manage all this data properly results in what Michael K. Gianninopoulos calls “data sludge.”
“We are rapidly approaching a point where the volume of data and the number of points of ingestion becomes unmanageable by one or more people,” said Giannopoulos, Chief Information Security Officer and Chief Technology Officer for Healthcare and Life Sciences. ‘ said. For Dell Technologies. The result, he said, is an “amorphous chunk of data” that needs to be sorted for relevance.
Artificial intelligence and machine learning technologies are poised to help healthcare organizations capture and keep control of all this data. These advanced tools do more than just synthesize datasets and glean actionable insights. You can also quickly analyze patient data, prioritize it, and protect it appropriately.
“We are not replacing humans in any shape or form,” emphasized Yiannopoulos, who is also Dell’s federal medical director. Rather, if models were developed with input from clinical stakeholders as well as data scientists, these tools would enhance human capabilities and improve care delivery, he said.
AI can help healthcare IT teams classify data and apply the most cost-effective security guidelines based on the classification rather than applying the same protections regardless of the data’s relevance. For example, electronic medical records and imaging data fall under Tier 1 and require a more robust cybersecurity infrastructure than Tier 2 level user access data logs.
Giannopoulos stressed that this low-cost approach also means that all data remains safe. “We are not reducing the security levels of the various data tiers, but we are streamlining the data levels and their impact on operations and his RTO. [recovery time objectives] It could cost less,” he said.
Generative AI, like OpenAI’s ChatGPT, is currently gaining traction for its potential to extract large amounts of data into digestible bits. Giannopoulos believes these large-scale language models will ultimately help healthcare organizations determine data protection levels and prevent ransomware, data breaches, and other cyberthreats.
He envisions healthcare systems using internal data to develop private, generative AI systems to implement and automate appropriate security guidelines. Currently available products such as his ChatGPT, which rely on the extensive Internet, are known to return inaccurate responses and, in Giannopoulos’ words, even patient It is known to cause enough “hallucinations” to construct studies and results of
“Public general search generation AI will be very different from what data scientists and technical experts develop for healthcare as part of healthcare delivery,” he said.
Radiology is focused on finding abnormalities in images and is currently leading the field in AI-assisted scanning. Gianninopoulos, who has spent his entire career in healthcare, believes digital pathology is next for his AI. This is because digital pathology is also image-based, facilitating the construction of rule sets that warn of anomalies.
For healthcare systems that have not yet embraced these new technologies but are interested in the potential for leveraging AI and advanced analytics to protect and validate patient data and reduce “data sludge,” he says 1 I offer one important piece of advice. And remember… your North Star is always patient. ”
