Technology
Addressing AI bias in FemTech
While digital advancements are exciting and hold much promise, they also pose significant issues, especially concerning bias in AI models. Tackling these challenges will require comprehensive strategies for detecting and mitigating bias, alongside robust regulatory frameworks, says Tim Bubb.
The FemTech sector, which focuses on female health and wellbeing, has been experiencing significant growth and enjoying new enthusiasm for investment in the last decade. In fact, in 2023, the sector witnessed an overall investment of $1.14 billion across 120 deals.1
This growth,
however, masks significant inequities in the way that funds and investments are distributed within the overall MedTech industry. In fact, less than 2.5 % of public-funded research in the UK is dedicated to reproductive health, even though it causes health issues for a third of women.2 Similarly, female-founded FemTech startups have raised $4.6 million per business on average, compared to $9.2 million by all male FemTech management teams.3
The rise of AI in MedTech Artificial Intelligence (AI) is a swiftly advancing technology transforming virtually every industry, and it is already impacting the MedTech sector. AI’s ability to analyse unstructured and structured data, using Machine Learning (ML) techniques, and ability to detect underlying patterns and associations, enables it to offer novel insights and breakthroughs in health. Whether this data is in text form (such as
notes), video or imagery, AI/ML can help save endless man-hours and help provide fact- based interpretations and analysis that could otherwise take human researchers years to complete. For example, AI/ML can rapidly analyse radiology images, histological data,
posture, eye movement, speech speed, pitch and sound and a whole range of other types of input. The specific inputs required will always depend on the intended use of the medical device, and the medical conditions it is associated with. An AI/ML enabled medical device can then provide structured medical information back to clinicians and patients, inferred using the training data used to develop the model. AI can be employed in many different applications, such as genetic testing to deliver personalised health recommendations or to recognise patterns between genetic data and routine health data that indicate early signs
Regulation needs to address the important issue of data privacy and security as this is closely connected to women’s rights and trust in providing sensitive health data, while ensuring that existing data sets are broadened to be more inclusive to consider differences between male and female health experiences and outcomes.
of underlying health conditions, facilitating preventive measures or early treatments to enhance outcomes. As a specific example, an AI algorithm trained to analyse mammograms, achieved a 9.4% decrease in false-negative breast cancer detection compared with human radiologists, as well as a 5.7% reduction in false-positive diagnoses.4
It is clear that the
application of AI to FemTech is beginning to yield some exciting results.
When AI meets FemTech However, the deployment of AI in FemTech is not without challenges. For starters, there are important security and data privacy concerns that could stem from the mismanagement, misuse, and misappropriation of intimate data on issues such as abortion and fertility.5
Recent
research showed that of 25 period trackers, 84% allowed the sharing of personal and sensitive health data beyond the developer’s system, with third parties.6 In Europe, medical device software
incorporating AI or machine learning typically falls under Class IIa or higher risk classification and requires formal regulatory assessment
August 2024 I
www.clinicalservicesjournal.com 31
Angelov -
stock.adobe.com
t
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60
