Opinion article for Jornal Económico | Artificial Intelligence in Healthcare: Opportunity or Threat?

Gonçalo Carvalho, Healthcare Director at TDGI S.A., wrote an insightful opinion article for Jornal Económico on “Artificial Intelligence in Healthcare: Opportunity or Threat?.”
Advancements in Artificial Intelligence (AI) are increasingly capturing interest due to their potential to automate numerous tasks. One area with significant potential for AI application is healthcare, enabling the processing of vast volumes of information in a short time and automating various processes.
Automating tasks through AI will free up healthcare professionals to focus on indispensable activities, a crucial advantage in a sector plagued by a chronic shortage of professionals. The World Health Organization (WHO) estimates a shortage of approximately 4.3 million healthcare professionals, which is projected to worsen, reaching a deficit of 10 million professionals by 2030.
Radiology is one field experiencing great progress in AI. According to a study by the American Radiology Association, the presence of AI in Radiology has grown from 0% to 30% between 2015 and 2020. For instance, Deep Learning Reconstruction (DLR) enhances the quality of CT images with significant noise, minimizing radiation doses while maintaining image quality, resulting in benefits for patients.
Telemedicine is another area with enormous potential for AI utilization. The development of wearables, devices that monitor biometric parameters in real-time, has energized its application. The combination of the Internet of Things (IoT) and cloud computing allows real-time data availability and the creation of historical records. Consequently, AI tools can analyze this vast amount of information and alert physicians to parameter deviations.
The enthusiasm surrounding AI is largely due to the success of ChatGPT, which amassed 100 million users in just two months (TikTok took nine months to achieve the same milestone). ChatGPT, a Large Language Model (LLM), is a neural network designed to process and generate data, particularly text. The potential applications of ChatGPT in healthcare are numerous, including support for triage and clinical record completion. Google has also developed an LLM for healthcare, Med-PaLM 2, which achieved an 85.4% success rate when tested with medical exam questions.
While AI capabilities create opportunities, they also pose new risks. One of the major risks associated with AI in healthcare relates to the privacy of clinical data. These risks, along with others, have prompted the need for regulation. The Food and Drug Administration (FDA) recently published guidelines stipulating that certain AI tools should be considered medical devices. At the European level, discussions are underway regarding the AI Act, a legislative package aimed at regulating AI use and updating the 2017 Medical Devices Regulation (MDR), as AI presence in Software as a Medical Device (SaMD) was still in its early stages at the time.
In summary, the potential of AI in the healthcare sector is vast, with many applications already being utilized on a large scale. This technology has the capacity to free healthcare professionals from repetitive tasks and partially address the workforce shortage. However, it is crucial to regulate AI to mitigate its risks and ensure it serves as a tool to enhance human capabilities rather than replace them.
Read the full article (original) in Portuguese here.
Source: Jornal Económico