Machine learning, with its ability to combine and learn from data in order to provide increasingly more accurate answers is being used in a wide range of medical technologies. From the development of targeted drugs, to smart wearables, to the use of virtual reality (VR), AI can help medical professionals and patients alike to provide and receive more precise, better tailored and faster diagnosis and treatment.
The big worry that's often cited, centres around the privacy of all the data collected and used, considering its highly sensitive and personal nature. What if a patient's medical records are leaked, or health-insurance companies get hold of them? Are patients being informed correctly of how their data is being used? Who can sell data of this nature, and to what ends? International standards can help mitigate these risks, by providing frameworks for the security and privacy aspects of data collection, storage and use.
A much less talked-about issue is the “gender data gap”, the fact that data used as a basis for the development of a myriad of products, from medication to cars, is disproportionately collected from men. Or when data is also collected from women, it is often not sex-disaggregated, meaning it is not separated and analyzed accordingly. This leads to products that are not only not adapted to half the world’s population, but can be downright dangerous and lethal for women.
The answer, of course, is to ensure that when data is collected it represents all segments of the population and is analyzed correspondingly. Equally, women need to be represented at all levels of the conception, design and roll-out of products, and this includes standardization. In May of this year, IEC signed the UNECE Gender Responsive Standards Declaration. Ensuring that standards in all fields are gender responsive and that women are included in all stages of the standardization process will require a strong will and commitment on the part of standards organizations, but one that is urgently needed.