By: 17 March 2022
New AI and medical scanning study could end the need for invasive surgery to diagnose endometriosis

A new study of women with endometriosis is underway exploring the potential of Artificial Intelligence (AI) and cutting-edge medical scanning techniques in diagnosing women earlier without going through invasive surgery.

The study named DEFEND (Developing an US-MRI-biomarker fusion model for Endometriosis) is now recruiting patients. Researchers are aiming to recruit up to 100 patients at King’s Fertility clinic who are experiencing the condition.

Endometriosis is a common condition affecting one in 10 women in the UK of childbearing age.[i] Women living with endometriosis may have to put up with significant pelvic and abdominal pain during menstruation, painful intercourse and spontaneous pain outside menstruation. In some cases, it can lead to infertility issues and 30-50% of women with infertility also being diagnosed with endometriosis.[ii] With March being Endometriosis Action Month, improving diagnosis and care is needed through tangible and collective efforts.[iii]

An All Party Parliamentary Group (APPG) Report on Endometriosis revealed that women with endometriosis are facing serious delays to diagnosis with nearly two thirds visiting their GP over 10 times, a quarter visiting doctors in hospitals 10 times or more and over half ending up in A&E due to their pain.[iv] According to the report, it takes 8 years on average from onset of symptoms to receiving a diagnosis, the same length of time as it did a decade ago, highlighting an urgent need for investment in research to drive down this time and ensure appropriate access to care when women need it.3

In addition, The Department for Health and Social Care’s (DHSC) ’Women’s Health – Let’s talk about it’ survey showed that 63% of respondents identified gynaecological conditions as a top priority for action and further research.[v]

Endometriosis is a clinical condition, initially diagnosed based on a collection of symptoms.  However, a definitive diagnosis of endometriosis involves invasive surgery called laparoscopy (keyhole surgery into the abdomen under general anaesthetic).  Ultrasound and magnetic resonance imaging (MRI) can be used to help diagnose the condition. However, ultrasound is not always reliable for all types of endometriosis and MRI is reserved for severe disease where bowel and bladder symptoms are present, added to which there is no single standardised MRI scanning protocol currently provides all the information required for a definitive diagnosis.[vi],[vii]

The DEFEND study will explore the effectiveness of using 2D and 3D ultrasound and MRI scanning. It aims to create a database of ultrasound and MRI images, along with clinical symptoms and medical history, for women with a diagnosis or symptoms of endometriosis.[viii] These images will be analysed to enable the potential development of computer algorithms to read images and harness the power of AI to better diagnose and manage women with endometriosis in the future.4

Chief Investigator Dr Ippokratis Sarris, Consultant in Reproductive Medicine and Director of King’s Fertility, said: “I’m really excited that the first Defend patients have been recruited. Endometriosis is such a poorly diagnosed condition, affecting 10-33% of women of reproductive age in the UK. Despite endometriosis being such a common condition, delayed diagnosis is a significant problem as symptoms often overlap with other conditions and can lead to infertility. Currently, diagnosis using conventional imaging techniques isn’t always possible, with key-hole surgery (known as a laparoscopy) being the present ‘gold standard’ way to reach a definitive diagnosis. However, this is an invasive procedure, meaning that it can either lead to years of uncertainty for those that have endometriosis but are reluctant to have surgery, or unnecessary surgery for those that are eventually shown not to have endometriosis. Our sincere hope is that this study will bring about the development of an algorithm to enable diagnosis through minimally invasive approaches; this would be a huge step forward for women with endometriosis symptoms.”

Jacqui Nix, a 35 year old patient at King’s Fertility clinic living with endometriosis said: “Living with endometriosis has meant living with debilitating pain, in particular 8-day long excruciatingly painful periods, since I was 15 years old. I have had four surgical laparoscopies to help manage my endometriosis. It’s an honour to be part of a study like this to help other women be diagnosed much sooner, using less invasive techniques. If they knew back then what they know now, my life could be incredibly different. My hope is that in the future women do not have to go through invasive surgeries to have endometriosis diagnosed and receive the support and care they need.”

The study is funded by the National Consortium of Intelligent Medical Imaging (NCIMI) and is being executed by a partnership between King’s FertilityPerspectum and GE Healthcare.

Dr Mark Beggs, Chief Operating Officer of NCIMI, said: “This is an important study for NCIMI to be funding because it is working to address an unmet need for many women in the UK who are living with this painful condition. We believe the latest technology in medical scanning, together with developing powerful new algorithms, could be key to unlocking more efficient diagnosis. We’ve brought together a powerful partnership of industry and medical experts to carry out this research. Conditions that only affect women can often get overlooked. We are determined for that not to be the case with endometriosis.”

Recruitment of patients to take part in the DEFEND study started in January 2022 and will continue for at least 6 months. Participants will be recruited from those patients on the surgical waiting list at the Gynaecology and Reproductive Medicine Clinic at King’s Fertility.



[i] Rogers PA, D’Hooghe TM, Fazleabas A, et al. Priorities for endometriosis research: recommendations from an international consensus workshop. Reprod Sci 2009;16(4):335-46. Available at:

[ii] Meuleman C, Vandenabeele B, Fieuws S, Spiessens C, Timmerman D, D’Hooghe T. High prevalence of endometriosis in infertile women with normal ovulation and normospermic partners. Fertil Steril 2009;92(1):68-74. (

[iii] Endometriosis Action Month. Endometriosis UK. Available at: Last accessed February 2022

[iv] Endometriosis in the UK: time for change. All Party Parliamentary Group (APPG) on Endometriosis Inquiry Report 2020. Available at: Last accessed February 2022

[vi] Nisenblat V, Bossuyt PMM, Farquhar C, Johnson N, Hull ML. Imaging modalities for the non‐invasive diagnosis of endometriosis. Cochrane Database of Systematic Reviews 2016, Issue 2. Art. No.: CD009591. DOI: 10.1002/14651858.CD009591.pub2.

[vii] Bazot M, Bharwani N, Huchon C, Kinkel K, Cunha TM, Guerra A, Manganaro L, Buñesch L, Kido A, Togashi K, Thomassin-Naggara I, Rockall AG. European society of urogenital radiology (ESUR) guidelines: MR imaging of pelvic endometriosis. Eur Radiol. 2017 Jul;27(7):2765-2775. doi: 10.1007/s00330-016-4673-z. Epub 2016 Dec 5. PMID: 27921160; PMCID: PMC5486785.

[viii] DEFEND study protocol. Data on file.

[ix] Vercellini P, Viganò P, Somigliana E, Fedele L. Endometriosis: pathogenesis and treatment. Nat Rev Endocrinol. 2014 May;10(5):261-75. doi: 10.1038/nrendo.2013.255. Epub 2013 Dec 24. PMID: 24366116.

[x] Rogers PA, D’Hooghe TM, Fazleabas A, et al. Priorities for endometriosis research: recommendations from an international consensus workshop. Reprod Sci 2009;16(4):335-46. Available at:

[xi] Simoens S, Dunselman G, Dirksen C, et al. The burden of endometriosis: costs and quality of life of women with endometriosis and treated in referral centres. Hum Reprod 2012;27(5):1292-9. Available at: Last accessed March 2022

[xii] NHS Endometriosis Overview. Available at: Last accessed March 2022

[xiii] Meuleman C, Vandenabeele B, Fieuws S, Spiessens C, Timmerman D, D’Hooghe T. High prevalence of endometriosis in infertile women with normal ovulation and normospermic partners. Fertil Steril. 2009 Jul;92(1):68-74. doi: 10.1016/j.fertnstert.2008.04.056. Epub 2008 Aug 5. PMID: 18684448.