BLOGs
Artificial Intelligence is the discipline focused on computer systems development with the capacity to afford tasks requiring human intelligence.
This technology is undergoing a significant transformation into medical assistance by improving:
- Diagnosis accuracy
- Diseases early detection, and
- Improving patient outcomes
Artificial Intelligence is providing the option to optimize medical access to remote territories, accelerating the arrival of sanitary attention.
MEDICAL PERSPECTIVES IN Artificial Intelligence
Much data is generated from daily medical routines, for example, patient data, laboratory results, images, research outcomes among others. Artificial Intelligence can analyze this burden of information and possibly identify critical undetectable pathways of clinical importance.
For example, in terms of diagnosis, Artificial Intelligence could ameliorate the scenario of possibilities based on evidence, helping to improve the level of uncertainty.
ADDRESSING Artificial Intelligence VALUE ALIGNMENT
According to Dignum V., a Responsible Artificial Intelligence expert, this technology should ensure human values alignment, such as:
- Patient autonomy: respect for decisions and preferences of patients
- Privacy: patient confidentiality protection
- Fairness: AI recommendations should be unbiased and equitable
- Trust: AI system must be reliable and transparent
RECENT EVIDENCE
Very recently, Lin C. and peers published the results of a randomized controlled trial focused on the rapid identification of ST-elevation myocardial infarction (STEMI) powered by Artificial Intelligence.
Main objective: to evaluate Artificial Intelligence-enabled electrocardiogram in the STEMI detection to reduce treatment delays.
Relevant details: open-label, cluster randomized controlled trial, 43234 patients, mean age: 60 years.
Intervention group: AI-ECG-assisted STEMI detection vs. standard of care.
Primary endpoint: door-to-balloon time. ECG-to-balloon time was also evaluated.
RELEVANT RESULTS
The AI-ECG group had a positive predictive value of 89,5% (95% confidence interval, 85,3% to 93,6%) and a negative predictive value of 99,9% (95% confidence interval, 99,9% to 100,0%).
Door-to-balloon time in the intervention group was 82 minutes compared with 96 min in the control group.
In conclusion, AI-ECG-assisted triage in STEMI patients deceased the door-to-balloon time and ECG-to-balloon as well.
This trial, as a proof of concept, demonstrates how Artificial Intelligence can improve clinical practice.
MESSAGE TO TAKEAWAY
The Artificial Intelligence era is revolutionizing many different fields, including healthcare. Its application requires the intervention of many specialists, apart from physicians, such as data scientists and Artificial Intelligence engineers.
However, as mentioned previously, an appropriate Artificial Intelligence value alignment, with particular emphasis on ethical principles, is required.
Finally, before recommending widespread implementation, it is highly requested that this tool be evaluated in different clinical settings under appropriate designs.
