In her multi-award winning book, 'Invisible Women: Exposing Data Bias in a World Designed for Men', Caroline Criado Perez discusses the systemic lack of data relating to women and describes the so called "gender data gap". In the medical and healthcare space, this has led to innovation largely based on men and male data and innovations in female-specific conditions are lacking.

When artificial intelligence (AI) plays a part in diagnostics and treatment, diverse and unbiased data sets are crucial to ensure these healthcare solutions work for everybody. As new developments in technology emerge, there is potential for plugging the data gaps in general health, improving research on female-specific conditions and improving the future of women's health.

Historically, women were often excluded from medical trials on the basis that their cyclical hormone fluctuations introduced too many variables. In the US, in the 1970s and 1980s, females of reproductive age were excluded from medical trials almost completely and even today, medical trials often include a disproportionate number of men to women. Medical clinical trial data are therefore largely male data, with sex-specific interactions of certain drugs being unaccounted for. In some cases, drugs and dosages of drugs developed based on largely male data have later led to severe adverse effects in females and the drugs later being removed from the market.

In fact, women are 60% more likely to have an adverse reaction to prescription drugs. Women's health is also underfunded compared to general health and male-specific conditions, leading to many female-specific conditions lacking quality research and data. Accordingly, there has been little innovation in women's health until recent years.

Rapid advancements in technology have led to new ways of collecting and analysing medical data, including data about the female body. The rising waves of Femtech, wearable tech and digital healthcare combined with advancements in AI and big data are leading to solutions finally focussing on women's health.

These solutions are leading to rapid generation of women's health data, filling some of the previous gaps, and in some cases generating data sets for the first time for areas of women's health which have not been researched in the past, from being able to identify reproductive system anomalies such as polycystic ovarian syndrome or a potentially life-threatening ectopic pregnancy to a fitness training app which designs a training programme tailored on female physiological data unique to each individual.

Wearable tech for women has led to data being collected on a continuous or regular basis, so called "longitudinal data", which can give a more in-depth view of bodily physiology compared to the brief snapshot obtainable through a single test taken at a doctor's surgery, which can lead to reduced diagnosis time.

Interestingly, these solutions targeted at women's health also have potential for making advancements in general health. For example, AVA was involved in trials determining whether the data collected for measuring a woman's fertility using their tracking bracelet (such as temperature, breathing rate and heart rate) could also be used for early detection of COVID-19 symptoms. Although the technology in the women's health space has begun to improve our knowledge about women's health, we still need more female health data.

For AI powered healthcare solutions, it is of critical importance to plug the data gaps in female health data. AI systems are tuned based on the demographics represented within the datasets and any biases within the data can be amplified by AI. Where healthcare-related AI systems are still operating on data sets which include a larger proportion of males than females, females will continue to be negatively impacted compared to males.

In one example, a lack of female medical research data meant that AI systems diagnosing heart attacks based on certain physiological indicators are 50% more likely to misdiagnose women than men, because the physiological indicators of heart attack for women are underrepresented in the data.

Targeted female data collection is also important to continue to innovate in the women's health space for largely female-specific conditions. Innovative AI solutions may help to reduce the cost and time in diagnosis of certain conditions.

These may be of particular benefit in developing countries. For example, in countries which do not have national screening programmes, technologies such as handheld devices using AI to detect early stage breast cancer offer a potential low-cost solution to improving breast cancer detection and treatment worldwide. Such devices also reduce the need for invasive and painful mammogram procedures which can be a factor in some females opting not to attend their screening appointments.

Clearly, a combination of AI, wearable tech and Femtech has vast potential for improving women's health, clarifying the problems that currently exist and unveiling potential new solutions. But the momentum in researching women's health must continue to increase, with continued importance placed on collection of diverse data sets in general healthcare.

At Venner Shipley, we are passionate about diversity and equality and therefore have a natural interest in technology which improves the lives of women. With experts in medical devices, wearable tech, AI and pharmaceuticals, our team is perfectly placed to advise on patent protection for women's health tech solutions.

Originally published 27 November 2020

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.