Can AI Really Transform Women’s Health?

We are on the precipice of huge transformational technology shifts that will ultimately revolutionize industries and if used correctly, augment the power of human creativity, generate tremendous global economic growth and, as Naval Ravikant describes it in The Almanack, “democratize consumption.”

Artificial Intelligence is essentially the science of making machines think like humans – the machine has the ability to perform the cognitive functions we associate with human minds. According to Walters Kluwers, while traditional AI typically focuses on recognizing patterns and making predictions, generative AI (GenAI), a subset of AI, focuses on generating new content, such as text, images, audio and video based on patterns and information it has learned from existing data, often in response to prompts. AI is increasingly being integrated into a wide range of industries, transforming operations, enhancing efficiencies, and creating new opportunities for innovation.

This technology’s potential impact on healthcare is especially profound as it is poised to redefine and reframe various aspects of healthcare, and could, potentially, address the long-standing gaps and disparities in the newly coined women’s healthcare industry.

In today’s market despite making up 51% of the global population, products and services which are specifically tailored to women’s health needs are still trailing behind. A McKinsey report highlights that women spend an average of 9 years in poor health – 25% more time than men. This not only affects their productivity but also reduces their earning potential.

AI is potentially able to address pertinent problems within the healthcare sector whilst simultaneously allowing for the faster development of products and services catered specifically for women.

  • Healthcare research is notoriously lengthy. Historically it has excluded women and overlooked problems specific to them. Prior to 1993 women were rarely included in clinical trials as researchers had concerns about how hormones and reproductive systems might invalidate study results. The AAMC found that due to this research gap, the medical field still doesn’t know how well many drugs and devices used today work for women. It is important to include women in these studies as women have different physiological, metabolic, hormonal, and cellular differences that can influence how diseases present and the effectiveness of pharmaceuticals and medical devices. Consequently, a Berkeley study found that the failure to include women in medical studies has contributed to women experiencing adverse effects from medications at twice the rate of men. AI can help accelerate the research process and allow for the faster development of products catered to women’s specific needs. By handling large data sets from multiple sources, AI enables researchers to expedite their work and reduce the process lengths. A study in the NCBI states that the automation of punctilious tasks such as designing trials, data collection and analysis allows for researchers to move more swiftly through their tasks. Moreover, AI’s ability to combine diverse data sources makes it more efficient in understanding and incorporating gender differences in research, leading to more effective solutions for women. GenAI can also be used to generate drug simulations and variations to identify drugs which may have been overlooked. This is essential in creating new drugs to treat conditions which were previously untreatable. GenAI can also draw on its data repertoire to fill in gaps in incomplete information effectively enabling researchers to move forwards. In so, it can be used to extrapolate information regarding how new drugs may affect women and how efficient they may be for women.
  • In clinical settings, AI can serve as a “virtual collaborator”, through Clinical Support Systems, and truly help healthcare providers in diagnosing and treating patients more accurately. Marsh McLennan states that medical professionals tend to under-diagnose or under-treat diseases that predominantly affect women or in situations where women have different symptoms for common diseases. In fact, a study found that women and minoritized individuals are 30% more likely to be misdiagnosed than white men. AI excels at analyzing complex and diverse information. It is therefore able to recognize subtle health problems and recommend precise interventions. Large language models (LLMs) have the capacity to process and understand terminology used in patient records and physicians notes as well as analyze lab results and medical imaging reports. LLMs can also infer meaning from incomplete information by drawing on the vast data they’ve been trained on to make sense of all the data. In doing so, it is able to create a holistic view of the patient’s health. This is particularly useful when applied to complex cases where a patient may have multiple conditions or symptoms which are related. By identifying prominent patterns, AI enables healthcare professionals to diagnose various conditions more accurately and much earlier. These powerful systems are also able to predict how certain individuals may react to certain therapies and can craft plans which optimize efficiencies. This capability enhances the customization of treatment plans, ensuring that therapies are more effective and tailored to individual needs. Great examples of start-ups revolutionizing this space are Predicta Med, which leverages AI to better diagnose and treat autoimmune diseases, and Glass Health, which aims to support clinicians by developing different diagnoses and drafting clinical plans.
  • AI’s role in fertility treatment is another area with significant potential. Today, unfortunately infertility affects around 1 in 6 couples. Consequently, IVF is becoming more widely used as a means for conceiving but in reality only has a 45% success rate (which exponentially goes down when women are over 35 years old) as per Forbes. AI can be leveraged in a multitude of ways to help address the increasing needs of couples in this space. For example, training AI systems to deliver recommendations on follicle-stimulating hormone dosages based on a woman’s personal data would enable to optimize the number of healthy eggs during the retrieval process. AI can also assist embryologists in labs to grade and rank embryos to determine the highest likelihood of success. Given that embryologists spend around 20% of their time documenting embryos, AI would enable them to focus more on their core work whilst augmenting the efficiency of IVF. Moreover, AI systems have great prediction capabilities which enable them to predict the likelihood of a successful IVF cycle. By integrating a multitude of factors from age, hormonal level, genetic data and previous treatment outcomes, AI is not only able to make a personalized prediction but is also able to develop tailored treatment regimens to optimize cycles for best outcomes. Companies making waves in the space are ImVitro, which leverages AI to streamline admin tasks and is a clinical decision support system for IVF clinics, as well as Gaia and Midi Health, which both offer AI-supported insurances for IVF patients, and Gameto, an AI-powered women’s reproductive health treatment solution developer.
  • AI can also play a crucial role in cancer diagnostics. The Guardian states that the global cancer cases are set to rise by more than 75% by 2050. Cancer Research found that today, close to half of all cancer cases in the UK are diagnosed at a late stage (3 or 4). However, the research found that if diagnosed and treated early, lung, breast and bowel cancers can have between 60-99% rates of survival as opposed to under 1% to 30% if diagnosed and treated at a later stage. Therefore, it is essential to be able to predict and diagnose cancers as early as possible to ensure patients are experiencing the highest rates of survival. In this realm, AI tools can be leveraged to analyze images and incorporate multiple data sources to detect cancer sooner and more precisely. In terms of image analysis and interpretation, the National Cancer Institute found that AI can be integrated to bring more logic to a highly subjective task. AI softwares are trained on thousands of images in order to be able to recognize patterns of healthy or cancerous cells. AI tools are therefore able to identify cancerous cells and/or highlight suspicious areas that may need further investigation, rendering workflows more efficient by enabling doctors to identify cancers faster, earlier and at a more treatable stage. Moreover, by integrating data from multiple sources, from genetic information, medical history and symptoms, AI tools are able to find complex patterns and relationships to highlight individuals which may be at higher risks, allowing for targeted screening and preventative measures. Innovative companies in this space include Vara, offering an AI support system for breast cancer screening, Deeplook which leverages AI in image analysis to diagnose cancers, and Panakeia, which uses AI to make biomarker profiling faster and find faster diagnosis and treatments.

Whilst AI systems are indeed full of promise, there are genuine ethical concerns. Algorithms still have their inherent faults. Because AI models depend on the data that is fed into it, a lack of representative data can lead to biassed models which deliver skewed outputs and health assessments. Therefore, while this technology is transformative, we must ensure that these new LLM models are trained with complete, diverse data sets which include women from all ethnicities in order for algorithms to ensure equal treatment.

Healthcare is a $12 Trillion global industry. Binarks found that in 2022, the global AI healthcare market was valued at $16 billion and is expected to grow 40.2% annually to reach $173.5 billion by 2029. McKinsey found that this exponential growth is primarily due to the extraordinary 5-10% savings (representing up to $360 billion for the healthcare industry in the US alone) that could be catalyzed. By making internal processes more efficient, healthcare companies are therefore able to service new customers, augment their overall offering and create new revenue streams. Moreover, savings on the patient side can be significant which in turn attracts more clients towards healthcare companies. For example, Fortune states it can cost as much as $70,000 to have a baby through IVF. Software which can help more patients get pregnant faster will ultimately help lower the cost of pregnancies as patients will not need to go through multiple rounds and will have more efficient treatment plans. Furthermore, AI can help support healthcare professionals as it renders medical staff more productive by helping them better diagnose and treat patients as well as taking the admin burden off their shoulders.

Venture Capital understands the potential AI represents in the healthcare industry. In 2023, SVB found that 23% of total US VC healthcare investment went towards companies leveraging AI, with pre-money valuations for healthcare companies leveraging AI coming in higher at seed and series A than those which did not leverage AI . Valuations within the AI healthcare sector can be scattered but tend to stay above average. For example, in December 2023, Verily Life Sciences (an Alphabet company) had a revenue multiple of 12.5x, Freenome 19.3x and Atomic AI 50x (Finro).

Similarly, we have seen many corporations looking to invest in, partner with or acquire AI healthcare companies to enhance their AI capabilities/offering or extend their market coverage. For example, Nvidia has partnered and invested in an array of AI healthcare companies to help grow its own offerings. From Abridge (AI-powered clinical documentation tools) to ClinicAI (early detection of gastrointestinal cancers and diseases) and Artisight (hospital automation platform). Microsoft recently acquired Nuance (AI solutions for healthcare providers to reduce burnout and improve diagnosis) and Alphabet has invested in the likes of Rad AI (AI radiology solutions) and Dyno Therapeutics (AI for novel gene therapy vector discovery). This wave of big tech companies investing into healthcare started in 2022 when Amazon bought One Medical for $3.9 billion, according to Bloomberg, in order to expand its reach into primary care, outbidding CVS Health (one of the leading health solutions companies in the US). Amazon has been integrating AI into One Medical through some of its AI services which include automating document processing as well as diagnosing illnesses and recommending appropriate treatments as per Radius.

As the General Partner of Goddess Gaia Ventures, we have dedicated this fund to investing in start-ups which are creating solutions for health issues which solely, disproportionately or differently affect women. We’ve made investing in AI enabled solutions a key focus of our fund as we are convinced that women deserve precision medicine through a women’s lens, venture capital is the perfect modality by which to accelerate innovation and ultimately AI could help close the 169 year women’s health gap faster and more efficiently as found per McKinsey and World Economic Forum.

The integration of AI in healthcare is continuously evolving, offering the potential to significantly improve patient care, reduce costs and accelerate medical research. Our current healthcare systems are clearly crumbling – we need to move from a volume-based business of understanding budgets, business models, value chains and waiting lists into a more holistic model based on preventative medicine– I believe AI has its role to play in this brave new world. Number 42 ☺

#ArtificialIntelligence #GenderGap #Innovation #VentureCapital #DigitalHealth