Artificial Intelligence (AI) in Modern Medicine
$1,811,800,000,000.00. This is the estimated value of the AI market in 2030. The market clearly sees huge potential for artificial intelligence, and entrepreneurs want to invest in solutions that use AI. At the same time, however, it is worth remembering that investors are not particularly enthusiastic about areas without clear legal regulations. Why? Because this increases the risk of investment. At this moment, there are no provisions within the European Union that explicitly regulate AI-related legal issues.
However, the lack of clear rules does not mean that officials do not recognize the needs of the market. Legislative work is currently underway in the European Parliament on the “Artificial Intelligence Act”. Action is also being taken at national level: in September 2020, the Committee of the Council of Ministers for the Affairs of the European Union Digitization adopted the “Artificial intelligence Development Policy”, which sets out objectives for Poland in the short, medium and long term, dividing them into six areas and establishing a task force to monitor the implementation of development policy and coordinate actions.
Although the current regulations could be more conducive to the development of artificial intelligence products, this does not mean that such solutions are no longer being developed today.
Regardless of the purpose of AI and the type of algorithms used, it is essential to obtain the right amount of good data to develop the most effective algorithms to achieve the best possible result.
In the case of medical data, it is important to emphasize that we will never have access to millions of data records, so one possible solution is to use machine-learned AI (supervised learning). In this particular type of machine learning, we provide not only data but also information on what such data contains. This allows much faster, efficient algorithm output with less data. For example, our CMAP platform (Comarch Medical Algorithms Platform) is based on 3,000 30-day studies, resulting in 90,000 24-hour ECG waves. In addition, the platform is still evolving, based on an analysis of 26,000 Holter studies, which add as many as 62,000 24-hour ECG waves registrations.
Comarch also uses artificial intelligence algorithms for:
- Sleep polygraph studies
- Skin lesions analysis
- Medical records analysis
- Automatic classification of documents
- Analysis of the results of laboratory tests
- Converting PDF files into semantic data
- Studies to detect potential diseases
- Medical history analysis for diagnosis profiling
Our work focuses on four objectives:
- Improving diagnosis and supporting staff in clinical decision-making
- Saving time of medical professionals
- Support for clinical studies and research and development
- Enabling faster detection of epidemic risks
In addition, we do not forget the most important people in the treatment process - patients. Artificial intelligence allows patients to access many new opportunities in treatment and prevention. With faster detection of patient abnormalities, less invasive treatment is often possible. AI solutions can also enhance the patient’s sense of security.
The patient’s sense of security is directly influenced by 24-hour monitoring of vital signs, which would not be possible on a larger scale without artificial intelligence. Automatic monitoring of a large patient population also allows clinicians to focus more attention on patients who are at the highest risk of complications.
AI-based solutions can also take on some of the administrative responsibilities of healthcare professionals. This allows professionals to spend more time treating patients. This is equivalent to accepting more people and reducing the waiting time for receiving treatment.
Another example of the use of AI algorithms is to implement them in the patient’s first contact path with the facility. This ensures that data is pre-processed, analyzed and sent straight to healthcare professionals. The data processed in this way allows professionals to take more personalized action on a particular patient, and this in turn also positively affects patient comfort and confidence.
Interestingly, the current topic of discussion in the medical environment is the ability of artificial intelligence to make autonomous decisions. It should be pointed out that, despite the controversy, this is probably the inevitable future in many medical fields.
About the author:
Mateusz Bruski
Product Manager
COMARCH SA
Graduate of the AGH University of Science and Technology, faculty of Biomedical Engineering. I gained my professional experience in the laboratory and medical industries. My passion outside my job is carpentry, and I like to spend my leisure time with my family and my dog. Please visit my LinkedIn profile.