Women's Health Tech and the Data Gap Holding Back Medical Innovation

An Accidental Pioneer in Women's Health Tech

When Marissa Fayer launched her professional career, her sights were set on aerospace engineering — not radiology software or reproductive health advocacy. Yet today, she leads DeepLook Medical, a healthtech company specialising in imaging and radiology technology designed specifically to visualise soft-tumor lesions in dense tissue. It is a niche with enormous clinical stakes, and one that has historically suffered from both underfunding and underattention. "I never intended to be in women's health, but I luckily fell into it," Fayer told SiliconRepublic.com. For the engineers, data architects, and policy professionals working at the intersection of technology and regulation, her trajectory offers a useful lens through which to examine a systemic failure in how medical data has been collected — and who it has historically excluded.

Fayer is not just a company executive. She is also the founder of the global non-profit HerHealthHQ and a published author who speaks regularly at major medical industry events. Her path into women's health began roughly four to five years into her engineering career, when she was recruited by one of the largest women's health companies at a time when the field was still largely uncharted territory. "I was innovating women's health technologies, I was traveling and moving around the world, I was working in the M&A space for women's health and at that point, I was hooked," she said. That combination of technical engineering background and commercial experience in mergers and acquisitions gave Fayer a rare vantage point: she could see both the technological deficits and the structural investment gaps that were holding the sector back.

Why the Clinical Research Data Gap Matters to Tech and Policy Professionals

For those working in data governance, AI model development, or healthcare compliance, the structural problem Fayer describes is immediately recognisable: when training data is skewed, outputs are skewed. In medicine, that skew has historically meant using male physiological data as the default standard — with real-world consequences for diagnosis accuracy, drug efficacy, and device calibration in female patients.

The numbers are stark. According to Fayer, women were not formally required to be included in clinical research in the United States until 1993, when new National Institutes of Health (NIH) regulations mandated their inclusion. Before that, women of childbearing age were routinely excluded from drug trials citing risk concerns — concerns that were not extended to male participants despite comparable or different risk profiles. The result: decades of medical research built almost entirely on male physiological baselines. As Fayer explained, "Women were only allowed to be included in research since 1993 and they still only make up less than 40 percent of all research study participants." For a demographic that constitutes 51 percent of the global population, that representational gap carries compounding effects on innovation velocity and clinical accuracy.

This is not merely a medical ethics concern — it is a data quality problem with direct implications for AI-driven diagnostics, imaging analysis, and algorithmic triage tools. Medical AI systems trained predominantly on male-derived data risk building systemic bias directly into their decision-making architectures. Research published by the New England Journal of Medicine has highlighted how diagnostic algorithms can underperform for underrepresented patient demographics, including women, when training datasets fail to reflect population diversity. For developers building on electronic health record (EHR) data or imaging repositories, this is a foundational issue: garbage in, garbage out — regardless of how sophisticated the model architecture.

From Taboo Topics to Technical Infrastructure: What's Actually Changing

Fayer is cautiously optimistic about the current state of the women's health space. She describes a pivot from awareness campaigns toward what she calls "capacity building and growth mode" — a transition that will resonate with any product or policy professional who has watched an emerging tech category mature from advocacy to infrastructure. "It's most exciting that we are moving past awareness and advocacy into capacity building and growth mode. That is when real infrastructure and focus really starts to grow and amplify the space," she noted.

Practically, this shift is visible in several ways. Conditions such as endometriosis, polycystic ovary syndrome (PCOS), and uterine fibroids — long treated as niche or secondary concerns — are increasingly entering mainstream clinical and public discourse. When high-profile public figures speak openly about these conditions, the "niche" label loses its grip. For technologists and entrepreneurs, this cultural shift creates a window: conditions that were previously too stigmatised to attract serious R&D investment are now viable product categories. Fayer made this point directly: "None of those are small areas of focus, but when we speak about them loudly, when celebrities start speaking about it publicly, the previous perceived niche is no longer relevant."

The broader femtech market reflects this momentum. According to analysis from McKinsey & Company, closing the women's health gap represents a multi-trillion-dollar economic opportunity globally, with significant scope for digital health tools, AI-assisted diagnostics, and data-driven treatment personalisation. Meanwhile, the European regulatory environment — including the EU Medical Device Regulation (MDR) and ongoing AI Act provisions — is beginning to incorporate diversity and representativeness requirements into clinical evidence standards, a development that has direct implications for healthtech companies seeking market access in Europe.

Investment, Policy, and the Multi-Capital Problem in Women's Health Tech

When Fayer discusses investment in women's health, she uses a framing that will be familiar to anyone who has worked in policy advocacy or startup ecosystems: the problem is not just a lack of venture capital, but a lack of multiple forms of capital simultaneously — financial, political, social, and intellectual. "It doesn't always have to be investment capital, but that significantly helps. It is time, multiple types of capital sources and changes in mindsets both personal and societal that start to overcome the gaps," she explained.

This multi-capital framing maps closely to how privacy professionals and digital sovereignty advocates think about systemic change in technology policy. The GDPR, for example, did not succeed through funding alone — it required sustained policy pressure, civil society advocacy, legal expertise, and ultimately a shift in how companies architectured their data systems. Women's health research reform faces a structurally similar challenge: changing entrenched norms across research institutions, regulatory bodies, funding agencies, and commercial developers simultaneously.

At the regulatory level, the NIH's 1993 mandate was a critical inflection point, but implementation has been uneven. A review published via the National Institutes of Health (NIH/PubMed) found persistent gaps in sex-disaggregated data reporting even after the policy was enacted — meaning that even when women were enrolled in studies, their data was frequently not analysed or reported separately from male participants. For AI developers building on clinical datasets, this distinction matters enormously: enrollment inclusion without analytical disaggregation still produces biased training data.

Why Closing the Women's Health Data Gap Is Everyone's Problem — Not Just Women's

One of Fayer's most practically important arguments is also one of the most frequently misunderstood in this space: the responsibility for improving clinical research equity does not rest solely with women or women-led organisations. She is explicit that inclusion in medical research is not a charitable act — it is a quality standard. "She said inclusion isn't an act of charity or goodwill, but rather a commitment to higher quality research that aids society and overall prosperity."

This framing has direct implications for IT decision-makers, data architects, and compliance professionals. When organisations procure or deploy AI diagnostic tools, the quality of those tools is directly dependent on whether the training data was representative. Deploying a diagnostic AI trained predominantly on male-derived imaging data in a mixed-patient clinical environment is not just an equity issue — it is a performance and liability issue. European healthcare institutions operating under GDPR and forthcoming EU AI Act obligations will increasingly need to scrutinise not just how patient data is protected, but how representative the training datasets of the tools they deploy actually are.

The European Commission's AI Act explicitly addresses high-risk AI systems in healthcare, requiring providers to demonstrate that systems are tested against diverse demographic groups. For healthtech vendors targeting the European market, this is not a soft recommendation — it is a compliance requirement. Fayer's call for systemic investment in research equity aligns directly with where European digital health regulation is heading.

"Women have the

Originally reported by Silicon Republic. Summarised and curated by European Purpose.