Modern infrastructure provides a shot in the arm for better medical care
Omar Akar, regional vice president, Middle East & Emerging Africa, Pure Storage writes on harnessing AI, data growth and data protection is vital for enhanced patient care.
Healthcare data has become fundamental to ensuring that medical teams can improve patient care and treatment. However, managing this unstructured data, which is stored in widely distributed silos and rapidly growing, is an increasing challenge. Legacy technology slows down clinician decisions and leads to longer waiting times for patients to access important information and diagnoses. Hospitals might have dozens of different applications to store, study, analyse and diagnose patients and clinicians need a simple, effective and fast way to navigate this mountain of data.
In addition, healthcare professionals face a challenge of how to secure this private information and protect against the growing threat of cyberattacks. Investing in a modern data strategy is one way for healthcare professionals to improve patient outcomes and treatment.
Improving access to medical information to increase productivity
Clinicians need to have visibility of relevant patient data to inform their decisions as to the optimal care treatment pathway: This means access to all the written data in medical notes as well as unstructured data including medical imaging, variants from clinical sequencing, pathology data or other lab results. An example of this type of data is Picture Archiving and Communication systems (PACs), which manage the workflow of images produced by imaging modalities (CT scanners, MRI scanners, X-rays). PACs are used by radiologists to deliver a patient diagnosis by examining and analysing these images. They can deliver between 8,000 and 12,000 diagnoses per year and in order to complete this, radiologists must have continuous access to imaging studies. Historically, when medical data is stored in separate silos this has not been possible which leads to a delay in treatment.
In a hospital, modern storage is essential to enable specialists to be more productive. It needs to deliver multi-dimensional performance as radiologists, pathologists and other clinicians need access to multiple types of medical information — from CT studies that contain thousands of very small images to pathology studies that contain fewer but much larger images. Traditional storage systems have difficulties delivering this multi-dimensional performance, in part due to silos of information, but also due to exponential data growth and their inability to handle this large volume.
Last year IDC predicted an average 270GB of healthcare data per person in the world. Siloed clinical data prevents clinicians from identifying correlations in treatment or disease, which in turn can lead to misdiagnosis or delays in treatment whilst the clinician finds the right information. Consolidating the data into a data “pool” instead, eliminates bottlenecks in accessing the right information, instead making it always available and performant, ready for consumption by any specialist, researcher or AI-algorithm to improve speed of diagnosis.
Artificial Intelligence (AI) enables faster results for clinicians
Medical professionals are already incorporating AI into their decision making, to support diagnoses and boost productivity. In medical imaging, dozens of AI algorithms have already been approved for clinical use and given a CE-mark to show they meet regulatory requirements. AI algorithms can support clinicians by streamlining their work, for example, AI can examine the 2,000 images generated by an MRI scan and distinguish those that contain anomalies or medical problems which must be examined by a health professional. While the final diagnosis remains fundamentally a human responsibility, AI in medical imaging acts as a virtual radiologist and can greatly improve the quality and productivity of diagnoses by prioritising examination results, images or finding similarities between studies for comparison. To do this, storage infrastructure needs to be scalable, performant and simple to use in order to simplify adoption of AI tools within modern PACS workloads.
Protecting against ransomware
The threat of cyber and ransomware attacks cannot be ignored when it comes to hospitals and medical facilities, because of the critical and private nature of the data involved and because the consequences of downtime due to a cyberattack are dramatic in terms of human life. The longer clinical systems are taken off line, the more risk there is to patient care.
As ransomware becomes more sophisticated, it is evading anti virus programs for longer and embedding itself in clinical files over weeks and months. It is an issue which requires solid protection against, as well as a plan for restore when a breach does occur. Medical facilities need solutions which provide twofold protection — taking immutable snapshots of data so hackers can’t delete or encrypt information and also providing the ability to restore at speed and at scale. Some solutions provide up to 270TB per hour data-recovery performance which will get medical facilities back online very rapidly if the worst does happen.
Better patient treatment with fast access to data
Complex workloads, made up of ever-increasing quantities of unstructured data, need a modern approach to consistently support the required performance and availability. The data management strategy and infrastructure of some healthcare institutions are not keeping pace with what’s necessary for medical practitioners and clinicians to be effective.
Healthcare in the future must be supported by solutions capable of bringing together all patient data with a single point of access to treat the patient as a whole, instead of examining a single medical problem at a time. It is about bringing together medication, imaging and AI expertise to improve detection and knowledge of diseases, and to provide a better patient experience with faster diagnosis and treatment. By breaking down silos of data and adopting a unified fast file and object storage platform, clinicians will be in a strong position to access and analyse data to better support improved decision making for modern healthcare needs.
Simplicity and continuous availability of critical application data are key, as the healthcare sector simply cannot afford downtime or fail to exploit the value of its vital data. Focusing on the effective use of healthcare data ultimately leads to improved real-time diagnostic capability for clinicians, faster patient treatment and improved patient outcomes.