AI in healthcare: a help or hindrance?

Meenu Vithanage

The healthcare industry is undergoing a technological revolution. From robotic-assisted surgeries to AI-driven diagnostic tools, digitisation and artificial intelligence are reshaping the way medical professionals diagnose disease, manage patient data, and deliver treatment.

From electronic health records to AI-driven diagnostics, such advancements supposedly improve efficiency, accuracy, and patient outcomes. Hospitals worldwide are embracing the innovations with the goal of streamlining operations and enhancing patient care. However, despite these implementations' promises, staff face many challenges due to technology's constant evolution. Many healthcare professionals find themselves struggling to keep up with frequent updates, complex systems, and new protocols that require constant learning and adaptation.

This was the main conclusion I reached, whilst studying the survey that I conducted, during my work experience in the Histopathology Lab. The survey aimed to study staff experiences with the implementation of LIMS software, which is the new data management system in the hospital. It was made clear that the staff are constantly being forced to adapt to advanced technology, which, without proper support, can become a burden to their workload. This could be fixed by implementing effective training programs, having IT technicians on site or introducing more user-friendly interfaces.

For the most part, AI is improving the efficiency of diagnosis for each patient. However, it means, especially at this early stage of digitisation, it is consuming a substantial portion of doctors' valuable time in administrative tasks.

In addition to this, there is a major data security threat as more patient data is moved online, meaning hospitals must also invest in robust security measures to protect sensitive information from cyberattacks. An example of such a data breach in healthcare is the Shields Health Care Group Data Breach Litigation in the US in 2022; which potentially exposed the private information of patients to an unknown actor.

Despite these challenges, there are many benefits to a digitised healthcare system. With electronic health records, doctors and specialists have instant access to patients' medical history ensuring better-informed decision making which reduces the risk of misdiagnosis. This also means there is seamless communication between specialists leading to better coordinated care.

One of the most promising applications of AI in healthcare is in clinical trials and drug development. Recently, the European Medicines Agency (EMA) approved an AI tool for use in fatty liver disease trials, marking a significant step forward in how medical research is conducted. Fatty liver disease is a growing global health issue that refers to a range of conditions caused by a build-up of fat in the liver. The tool, trained on over 100,000 annotations from 59 pathologists assessing more than 5,000 liver biopsies, has demonstrated the ability to determine disease activity with less variability than traditional methods that rely on a consensus of three pathologists. This means AI could eventually reduce human error in diagnostics, making medicine more precise and efficient.

As beneficial as AI could be, there are many aspects of its use that could cause the movement to spiral out of control. The most prominent concern being the over-reliance of such technology which could lead to rare conditions being overlooked. Although the rare case is least common, medicine is a field in which lives are at stake and mistakes are fatal. In addition to this, medicine is not an exact science, it requires interpreting data from each patient's life differently and with empathy. This is not something that AI is capable of simply learning at this current point in time. If doctors trust AI too blindly, they may miss out on diagnosing less common but equally serious diseases. Thus, AI should support the practice of medicine, not replace human expertise. 

As AI continues to evolve, it is clear that healthcare will experience significant developments across most fields in medicine, especially in diagnostic fields. By analyzing a copious number of data sets, AI could be very skilled in optimising patient outcomes based on genetic profiling, lifestyle aspects, and others. Thus, AI could be implemented to reduce administrative tasks for doctors and other health professionals, which would lead to enhanced efficiency.

The digital transformation of healthcare is revolutionising medicine, from AI-assisted diagnostics to fully digitalised patient records. While the benefits are clear (improved accuracy, efficiency, and accessibility), the challenges must be addressed to ensure a smooth and ethical transition.