Using AI to Protect NC Healthcare Systems
An editor has nominated this article for deletion. You are welcome to participate in the deletion discussion, which will decide whether or not to retain it. |
Cybersecurity attacks are defined as threats to computer systems or networks with the intention to steal, damage, or alter information. These attacks come in different forms, ranging from the installation of malicious software, phishing emails, or ransomware just to name a few. As the healthcare industry grows more reliant on IT systems, medical technologies, and smart equipment, there is a need for increases in security to prevent vulnerability (Kioskli, K.; Fotis, T.; Mouratidis, H., 2021). The healthcare sector in North Carolina has proven itself to be unprotected and at risk of data breaches, with patient records and personal information in danger of being accessed unlawfully.
Just within the past decade, multiple North Carolina healthcare systems have experienced attacks on their networks, resulting in multiple patient data security incidents. On February 3rd, Duke Health confirmed that a pro-Russian hacking group interfered with their website, threatening hospitals. This group is known for causing Denial of Service (DoS) attacks by overloading a website or network, to slow down a service, cause poor performance and inaccessibility. On April 2nd, 2023, the University of North Carolina (UNC) Health released a notice discussing an incident in which patient data was disrupted in the medical record system Epic, a healthcare software company that provides electronic health records (EHR) systems. Within the past year, on February 1st, 2024 a UNC School of Medicine user experienced a social engineering attack by clicking on a phishing hyperlink. This allowed an unauthorized individual to access the user’s university email account, which then enabled viewing of various data types including patient names, date of birth, diagnosis and treatment information, Social Security numbers, and health insurance identification numbers.
The article “The Landscape of Cybersecurity Vulnerabilities and Challenges in Healthcare” by Kioskli, Fotis, and Mouratidis calls for oversight and expansion of security standards in healthcare. The abstract discusses how current standards are “often contradicting and confusing, making these measures ineffective,” which only emphasizes the need for improvements in the healthcare industry. They present possible solutions as well as some of the challenges in implementing different practices, along with recommendations and next steps. One of their propositions includes the use of the Internet of Things (IoT) devices which can be regulated using Artificial Intelligence.
The use of Artificial Intelligence (AI) has been increasing in recent years due to technological advancements that have allowed it to become prevalent in many industries and field studies such as healthcare, finance, retail, and many more. One important use of AI is its role in the cybersecurity industry in preventing or detecting malicious threats before it can compromise sensitive data thus endangering people in the long term. This is especially noticeable in the healthcare industry where cloud computing and the Internet of Things have been integrated into hospital systems making the sharing and usage of medical information much easier and at the same time more susceptible to cyber attacks (Ganesh et al. 2021). In this paper, we will look at existing studies on how AI is used in the security of healthcare system’s data and how it's evolving to meet new threats.
The article “Intelligent AI-based Healthcare Cyber Security System using Multi-Source Transfer Learning Method” by Ganesh Gopal Devarajan et al. talks about existing cyber security methods currently employed in healthcare institutions around the world and proposes a new system to better the security of the healthcare sector. This paper was published in 2023 by The Association for Computing Machinery, allowing it to discuss more modern issues related to the security of the healthcare industry and the need for better security. An example is how the paper talks about various security systems developed by various researchers that are limited in one way or another as described in the paper, “These issues mainly occur in communication overhead, latency, and privacy concerns between cloud computing and healthcare entities regarding the aggregation of sensitive data.” (Ganesh et al. 2023). This has led the team behind the paper to propose a new system that leverages AI in order to better connect IoT devices with each other with a central unit that can process data in real time and simultaneously be able to protect the data in transit using blockchain technology to address the issues stated above and highlights how a new system is needed not only to deal with attacks but also in the efficiency of data sharing.
Impact in NC
[edit]As discussed earlier, North Carolina Healthcare systems are vulnerable to security and data breaches and are targeted by international cybersecurity attackers through phishing and Denial of Service (DoS) attacks. According to the news article “Health Care Providers Battling Cyber Attackers” by Mark Tosczak, published in 2018 states that “more than 385,000 patient records in North Carolina have been exposed in more than 40 cybersecurity incidents.” Various incidents discuss how malware encrypts data, making it impossible for some doctors to access patient records resulting in appointment cancellations and time spent in restoring records through backups. Specifically on October 17, 2017, FirstHealth in Pinehurst, North Carolina was attacked by ransomware known as WannaCry. (Tosczak, 2018) WannaCry has infected more than 300,000 devices, and in this case, corrupted patient data, withholding it until a ransom was paid. This forced the organization to shut down its computer network for 1-2 weeks after identifying the threat. (Tosczak, 2018) This led to the following question: How can AI be integrated into North Carolina healthcare systems to enhance cybersecurity?
The integration and application of AI in healthcare facilities has become a notable movement in North Carolina according to a research article written by Tardif-Douglin of UNC at Chapel Hill. In the article “A Compass for North Carolina Health Care Workers Navigating the Adoption of Artificial Intelligence,” we see that North Carolina Hospital Systems such as Atrium Leath Wake Forest Baptist has already used AI to help as a diagnosis tool while Duke University uses AI in the identifying stages of sepsis. Observing successful uses of AI in NC medical services showcases that artificial intelligence has the potential to grow healthcare.
A local research group that has recognized AI’s potential includes the newly launched Center for Artificial Intelligence and Public Health (CAIPH) at the University of North Carolina at Chapel Hill. The CAIPH program understands that “There are vital issues in public health that could benefit from the support of AI,” according to Dean Nancy Messonnier of UNC Gillings School of Global Public Health. The Center of AI and Public Health is led by Dr. Michael Korsorok who also published a research article called “The Future of Precision Health is Data-driven Decision Support” alongside Nikki Freeman of Duke University. This article emphasizes how our healthcare systems are largely dependent on patient data and trends, which support the decision-making process of health professionals. Through the development of medical device wearables and secure data collection, our healthcare system can increase and expand its support tools increasingly. These basic principles can be enhanced through the integration of Artificial Intelligence and the promotion of cybersecurity when handling patient data within North Carolina.
Another research group with a similar interest is The Secure Computing Institute (SCI) at NC State University. It focuses on cybersecurity and its application in the healthcare industry through Internet of Things (IoT) devices and cloud and software security amongst others. NCSU can actively engage in cybersecurity research within the medical sector by developing solutions to healthcare-related threats. SCI at NCSU focuses on initiatives that relate to patient data privacy, healthcare IT security, and medical device security. Not only does it utilize resources provided within the Research Triangle Park, but this organization also advocates and spreads cybersecurity awareness. Additionally, the cybersecurity program at NC State University provides customized onsite training programs for healthcare professionals and regional managers. To summarize, North Carolina has experienced serious cybersecurity threats within the healthcare industry which is a cause for concern. However, through local research groups and support organizations, there are possible solutions to further secure patient data, electronic health records, hospital applications, and web services.
Majors involved
[edit]In today’s world, the healthcare industry faces significant challenges in being able to protect patients' data from cyber-attacks/threats. AI is being integrated into cybersecurity strategies to help combat this sector. At NCSU, engineering students from Computer Science and Chemical engineering are contributing to this cause.
The Computer Science department at NCSU focuses on AI algorithms and security protocols designed to detect and fix cyber threats. Courses like CSC 411 (Introduction to Artificial Intelligence) and CSC 236 (Computer Security) allow students the knowledge to create advanced learning models to identify data patterns. An important factor in preventing access to healthcare systems. Teaching ethical hacking and cybersecurity frameworks helps students develop protection mechanisms like encryption, multi-factor authentication, and intrusion detection systems. The textbook mentions how computer scientists are the ones responsible for designing computer and software systems for people and businesses of various fields. Labs and projects ensure Computer Science students can apply these theories to protect healthcare data against attacks.
Chemical Engineering students can apply their knowledge in data analysis and systems used to optimize AI. Courses like CHE 225 (Chemical Process Principles) and CHE 446 (Process Control), focus on optimizing physical systems. This translates into Chemical Engineering students working on optimizing AI systems to make sure large datasets like patient records or prescription information are processed securely. The ability to manage data flow will be very important in protecting healthcare information.
AI-derived solutions in the healthcare space for cybersecurity gain immensely from the contributions of Computer Science and Chemical Engineering students. Interdisciplinary engineering courses support the students with the knowledge and tools to address this challenge. Aiding in the protection of patient data from cyber attacks.
Possible solutions
[edit]Detecting Unusual Activity, AI can be used to monitor how data is accessed and shared. If someone tries to access too much information at once, or if data is being shared from an unexpected place, AI can detect that and send an alert. This can be a combination of IDS (Intrusion Detection System) and IPS (Intrusion Prevention System) which can provide the capabilities stated above. These solutions can also be customized to fit the needs of the user or organization such as enabling signature-based detection which allows the IDS to detect malware based on signatures that are associated with certain attacks or anomaly-based detection which flags any malicious or suspicious patterns that it has never seen, allowing to catch unknown threats more easily. It can also be scaled to do this job on one machine or a network segment depending on the need. IPS is similar to IDS in that it can detect but also take action to prevent any malware coming in and be scaled to either one machine or a network segment. Adding AI to these solutions allows it to be more flexible and not be dependent on a human to change the IDS or IPS to changing environments and threats with mitigated risk of errors compared to a human.
Secure Cloud Storage, Many healthcare providers store patient data in the cloud. AI can make cloud storage more secure by managing and encrypting data to keep it safe from hackers. AI is also used in other aspects of cloud storage such as managing the flow of data coming in from various devices, generating harder encryption keys through various machine learning processes, automating encryption processes in general, and much more. Engineers can feed AI various encryption algorithms that are complex in order for the AI to generate its own encryption algorithms and even outside sources such as sound for the AI to generate it into data and further the encryption process. From there white hat hackers can test the strength of the encryption to gauge its effectiveness before implementing it into the real network. They also can simulate various attack scenarios to see how fast the AI can detect and respond to threats to see how well it handles old and new threats in today’s digital environment. If the algorithms and database the AI was given are good, it should allow the AI to be proficient at its job and ease the burden of cybersecurity professionals to focus their time on new threats, new avenues of attack, and staying ahead of hackers in general and relay that information back into the AI to improve its effectiveness.
References
[edit](1)Kioskli, K.; Fotis, T.; Mouratidis, H. The Landscape of Cybersecurity Vulnerabilities and Challenges in Healthcare: Security Standards and Paradigm Shift Recommendations. The 16th International Conference on Availability, Reliability and Security 2021. https://doi.org/10.1145/3465481.3470033.
(2) Chakraborty, C.; Nagarajan, S. M.; Devarajan, G. G.; Ramana, T. V.; Mohanty, R. Intelligent AI-based Healthcare Cyber Security System using Multi-Source Transfer Learning Method. ACM Transactions on Sensor Networks 2023. https://doi-org.prox.lib.ncsu.edu/10.1145/3597210
(3) Korobenko, D.; Nikiforova, A.; Sharma, R. Towards a Privacy and Security-Aware Framework for Ethical AI: Guiding the Development and Assessment of AI Systems. dg.o '24: Proceedings of the 25th Annual International Conference on Digital Government Research, 2024. Pg 740-753. https://dl-acm-org.prox.lib.ncsu.edu/doi/10.1145/3657054.3657141
(4) Newza, A. I.; Sikder, A. K.; Rahman, M. A.; Uluagac, A. S. A Survey on Security and Privacy Issues in Modern Healthcare Systems: Attacks and Defenses. ACM Transactions on Computing for Healthcare, 2021. Volume 2, Issue 3, Pg 1-44. https://doi-org.prox.lib.ncsu.edu/10.1145/3453176
(5) Duvet, A. A.; Rahi, P.; Ganesh, N. S. G.; Basa, S. S.; Smart framework for enhancing security of IoT linked devices in Healthcare Systems. ICIMMI '23: Proceedings of the 5th International Conference on Information Management & Machine Intelligence, 2024. Pg 1-8. https://dl-acm-org.prox.lib.ncsu.edu/doi/10.1145/3647444.3647902
(6) Wilson, E.; Yang, K.; Beasley, A. A Compass for North Carolina Health Care Workers Navigating the Adoption of Artificial Intelligence. N. C. Med. J. 2023, 84 (5), 348–350. https://doi.org/10.18043/ncm.84.5.348. https://ncmedicaljournal.com/article/120571
(7) Hoban, R. Health Care Providers Battling Cybersecurity Attackers. N. C. Health News, February 21, 2018. https://www.northcarolinahealthnews.org/2018/02/21/health-care-providers-battling-cybersecurity-attackers/
(8) Robbins, H.Q.&.M. C. (2024). First Year Guide to Engineering: Challenges, Successes, & Opportunities (5th ed.). McDonald Publishing DBA Tavenner Publishing. https://ncsu.vitalsource.com/books/9781642202830
Secure Computing Institute. North Carolina State University. https://sci.ncsu.edu (accessed October 13, 2024).
(9) Zhao, Z.; Xie, L. Developing a Cybersecurity Strategy for Health Care Systems. Stat. Anal. Data Min. Appl. 2018, 11 (1), 42-51. https://doi.org/10.1002/sam.11475. https://onlinelibrary.wiley.com/doi/abs/10.1002/sam.11475