When it comes to digital transformation, the healthcare industry is caught between a rock and a hard place. The potential of innovations in AI, data analytics and digitisation all have the promise to streamline administration, reduce patient waiting times, and ultimately boost patient outcomes – but the critical need to safeguard patient data limits the proliferation of these technologies.
Boris Bialek, Field CTO of Industry Solutions at MongoDB, shares his expert insight around how new encryption techniques at the application development stage can turn data from an inhibitor to an enabler of these innovations.
What potential does AI, data analytics and digitisation have in the healthcare sector?
While AI algorithms have been applied in many fields, (specifically in medical image diagnostics and clinical decision support systems), with the emergence of GenAI the opportunity for impact has increased exponentially. AI, data analytics, and digitisation have the power to transform the healthcare sector by improving diagnostic accuracy, enhancing treatment outcomes, reducing operational costs, increasing healthcare accessibility, alleviating administrative burden, and accelerating medical research and advancements.
Despite the optimism AI can offer, it is critical to remember that healthcare scenarios often involve complex and nuanced factors that may not be fully captured by AI algorithms alone, and healthcare decisions may have life-or-death consequences.
As a consequence, all the AI approaches to healthcare decisions require the human-in-the-loop approach, where the inclusion of humans in the decision-making process ensures safety, accountability, ethical compliance, and a holistic approach to patient care, making it an indispensable component of AI-driven healthcare systems.
How can these innovations streamline administration, reduce patient waiting times, and ultimately boost patient outcomes?
The integration of AI, data analytics, and digitisation into healthcare can streamline administration, reduce patient waiting times, and ultimately boost patient outcomes in several ways. According to some reports, 80% of doctors spend more than 50% of their time on data entry and administrative tasks, limiting their efforts to provide optimal patient care. With the automation capabilities of AI, healthcare professionals can be alleviated of administrative tasks such as summarising patient charts, personalising reports to doctor’s transcripts and performing data entry. Instead, AI can assist healthcare providers in critically reducing their admin work, and increasing their life changing work for patients.
AI could also help with predictive analytics for patient flow, workflow optimisation and the efficient utilisation of resources, such as equipment, personnel, and facilities. Also, digitisation can help to automatically prioritise patients based on the severity of their condition, including triage. AI driven appointment scheduling systems can then optimise the allocation of time slots based on patient preferences and healthcare provider availability.
AI-powered decision support systems can also assist healthcare professionals in making quicker and more accurate decisions, leading to faster diagnosis and treatment, and ultimately improving patient outcomes. Patient progress can even be tracked in real-time, with AI-enabled monitoring systems, triggering early intervention when needed.
What limit does security put on these technologies?
Security is a critical concern for all aspects of healthcare — and there are several limitations and challenges related to security that can impact the adoption and use of these technologies. Across many industries, data privacy has become a major concern, with the protection of sensitive patient data being crucial to prevent data breaches and maintain patients’ trust. We must take into consideration the availability, correct use and traceability of the patient consent for any patient data used in any algorithm through explicit permissions from the patient themselves. Healthcare systems are also often prime targets for cyberattacks, which risk the disruption of operations and compromisation of patient data.
Healthcare networks are composed of many different players, including health providers, payers, technology providers, research organisations and institutional bodies — and every new actor and application could themselves potentially introduce a new security risk. That’s why secure data exchange between all systems is mission-critical. It is also crucial to be aware of the potential biases within AI algorithms, which could lead to disparities in diagnosis and treatment, threatening patient security.
How can new encryption techniques at the application development stage turn data from an inhibitor to an enabler of these innovations?
Encryption is a fundamental security measure used to protect sensitive data in various applications, including healthcare systems. It involves converting plaintext data into ciphertext using cryptographic algorithms and encryption keys. In healthcare, where patient privacy is paramount, encryption is crucial to safeguarding sensitive medical records and personal information.
Today’s cloud applications communicate with servers through encrypted channels (in-flight encryption), and all the data stored in the cloud is also encrypted through in-rest encryption.
At MongoDB, we provide two key features, to lift security to the highest possible level. Client-Side Field-Level Encryption (CS-FLE) is an advanced encryption technique that provides a high level of security for sensitive data stored in MongoDB databases. Unlike traditional server-side encryption, which encrypts the entire database, CS-FLE allows for encryption at a more granular level—namely, at the field level. Here, data is encrypted at the origin, at the application level, using unique encryption keys. In addition, our Queryable Encryption capabilities combine the benefits of encryption with the ability to perform specific queries on encrypted data. In healthcare, this can be a game-changer as it allows for secure data retrieval without compromising data privacy.
Tell us about your plans for 2024?
In 2024, we are staying committed to vector search for AI workloads, enabling semantic search and generative AI. Notably, a Retool State of AI survey found, our Atlas Vector Search received the highest developer net promoter score among all vector databases in the market. We're excited to build on this success in 2024, enabling more AI workloads on top of MongoDB.
Alongside this, customers are continuing to feel pressure to modernise their data infrastructure, aware that legacy platforms are holding them back from building modern applications designed for an AI future. We want to be part of this journey, so we launched Relational Migrator earlier this year to help customers successfully migrate data from their legacy relational databases to MongoDB. Now, we're looking beyond just data migration to help with the full lifecycle of application modernization. We will also continue to focus on allowing customers to run-anywhere, on-premises and in multiple clouds, all while helping to manage cost. Cost management is unsurprisingly a core focus for our customers, and we’re dedicated to increasing developer productivity, and supporting a variety of use cases, to support our customers with their goals.
Healthcare Digital is a BizClik brand