DZone

Machine Learning (ML) research in the healthcare field has been ongoing for decades, but almost exclusively in the lab rather than in the doctor’s office. The problem of implementing ML in patient-facing settings largely stems from two barriers:

• Regulation: in a highly regulated industry like healthcare, there currently exist very few guidelines on the use of ML. This is beginning to change, as just last month the US Food and Drug Administration announced a new regulatory framework designed to promote the use of AI-based technologies.
• Litigation: ML “mistakes” do not fit comfortably in either the “doctor’s negligence” or “defective product” lawsuits we typically see in healthcare today. Determining compensation for patients injured by the use of ML may require new laws, or else a compensation scheme similar to vaccine complications.

As a result, we’re most likely to see the implementation of ML solutions in clinical systems first. For example:

Source: DZone