Trial Review

Registering a new trial?

To achieve prospective registration, we recommend submitting your trial for registration at the same time as ethics submission.

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers
Trial registered on ANZCTR


Registration number
ACTRN12623000646640
Ethics application status
Approved
Date submitted
22/03/2022
Date registered
15/06/2023
Date last updated
15/06/2023
Date data sharing statement initially provided
15/06/2023
Type of registration
Retrospectively registered

Titles & IDs
Public title
IMAGENDO: Diagnosing Endometriosis with Imaging and Artificial Intelligence
Scientific title
Non-invasive endometriosis diagnosis in women using machine learning
Secondary ID [1] 305284 0
PHRDI000014
Universal Trial Number (UTN)
Trial acronym
IMAGENDO
Linked study record

Health condition
Health condition(s) or problem(s) studied:
endometriosis 323585 0
Condition category
Condition code
Reproductive Health and Childbirth 321131 321131 0 0
Other reproductive health and childbirth disorders

Intervention/exposure
Study type
Observational
Patient registry
False
Target follow-up duration
Target follow-up type
Description of intervention(s) / exposure
There are two study arms:
Stage 1 - retrospective study - Women who have had an MRI or MRI and TV-US in the last 5 years will be contacted directly by mail by administration staff from our partner radiology and ultrasound clinics (including Benson Radiology, Specialist Imaging Partners, OmniGynaecare, O &G ) to obtain consent for these images and to follow up on any operation notes to confirm diagnosis. It is anticipated that it will take approximately 25 – 35 mins for the participant to complete their baseline data collection. Baseline data collection includes date of birth, height and weight, any previous surgery, any previous diagnostic imaging, Treating Specialist Doctor / Gynaecologist, Surgeon name and Hospital. Images and operation notes are de identified by admin staff from the clinic before being sent to the computer analysts. Participants will only need to read information and provide consent, and supply baseline information including date of birth, height and weight, any previous surgery, any previous diagnostic imaging, Treating Specialist Doctor / Gynaecologist, Surgeon name and Hospital. Once imaging and operation notes are received there will be no further observation. Images and medical history details will be entered into a machine learning algorithm designed specifically for this study. We will then compare the results from the algorithm to the documented diagnosis.

Stage 2: After being identified as eligible for the study, if they haven’t already had one, women will be invited to undertake a transvaginal endometriosis ultrasound scan and/or an MRI at one of the imaging partners. These scans take less than an hour to complete. They will then attend a follow up interview with their Gynaecologist at least one week prior to their surgery, for 15 minutes who will explain the findings of the scans. They will also attend a review appointment within one month after their surgery with their Gynaecologist. Operation notes will be accessed by the study team after surgery.
When consenting for the study, women will also be asked if they are willing to have their contact details including their name, date of birth, address, phone number, email address and treating specialist doctor entered on a secure electronic database and be contacted about future research questions that might arise from this project.
It is anticipated that it will take approximately 25 – 35 mins for the participant to complete their baseline data collection. Baseline data collection includes date of birth, height and weight, any previous surgery, any previous diagnostic imaging, Treating Specialist Doctor / Gynaecologist, Surgeon name and Hospital. The observation period will end at the follow up review appointment post surgery.
Intervention code [1] 321689 0
Diagnosis / Prognosis
Comparator / control treatment
Laporascopic surgery will be used as a comparison for both Stage 1 and Stage 2. Surgical notes will be obtained directly from surgical clinics. No active involvement from participants will be required post surgery. The oveerall duration of observation for patients will be from imaging diagnosis until surgical notes are obtained.
Control group
Active

Outcomes
Primary outcome [1] 330837 0
Diagnostic Test Accuracy for Stage 1:
Machine Learning algorithm (Transvaginal Ultrasound & MRI) compared with laparoscopic surgery for endometriosis - Composite sensitivities, specificities, diagnostic accuracy, PPV, NPV and the AUC will be calculated
Timepoint [1] 330837 0
6 months after participant enrolment in study
Primary outcome [2] 334123 0
Diagnostic Test Accuracy for Stage 2:
Machine Learning algorithm (Transvaginal Ultrasound & MRI) compared with laparoscopic surgery for endometriosis - Composite Sensitivities, specificities, diagnostic accuracy, PPV, NPV and the AUC will be calculated
Timepoint [2] 334123 0
2 years after participant enrolment in study
Secondary outcome [1] 419422 0
Time elapsed between imaging and surgery - sugical data will be requested directly from surgeons after patient consent.
Timepoint [1] 419422 0
2 years after patient enrolment

Eligibility
Key inclusion criteria
Women with symptoms of endometriosis including:
o Period paid
o Other chronic pelvic pain
o Fatigue
o Dysmenorrhoea,
o Dyspareunia,
o Difficulty conceiving
• Women do not need to have regular menstrual cycles, and can be taking oral contraceptive or have a Mirena in place
Minimum age
18 Years
Maximum age
45 Years
Sex
Females
Can healthy volunteers participate?
No
Key exclusion criteria
• Women with cancer
• Women with bowel conditions such as Crohn’s Disease or Ulcerative Colitis
• Postmenopausal women
• Women with an intellectual disability/inability to give informed consent

Study design
Purpose
Screening
Duration
Cross-sectional
Selection
Defined population
Timing
Both
Statistical methods / analysis
As Machine learning is dependant on obtaining as mcuh data as possible, we have not coalbulated a sample size.
Composite Sensitivities, specificities, diagnostic accuracy, PPV, NPV and the AUC will be calculated

Recruitment
Recruitment status
Recruiting
Date of first participant enrolment
Anticipated
Actual
5/11/2020
Date of last participant enrolment
Anticipated
30/06/2024
Actual
Date of last data collection
Anticipated
30/06/2025
Actual
Sample size
Target
1000
Accrual to date
350
Final
Recruitment in Australia
Recruitment state(s)
ACT,NSW,NT,QLD,SA,TAS,WA,VIC
Recruitment outside Australia
Country [1] 25313 0
Canada
State/province [1] 25313 0
Ontario

Funding & Sponsors
Funding source category [1] 309656 0
Charities/Societies/Foundations
Name [1] 309656 0
Australian Gynaecological Endoscopy Society (AGES)
Country [1] 309656 0
Australia
Funding source category [2] 311071 0
Charities/Societies/Foundations
Name [2] 311071 0
Endometriosis Australia
Country [2] 311071 0
Australia
Funding source category [3] 311072 0
Government body
Name [3] 311072 0
Australian Government (MRFF 2020 Primary Health Care Research Data Infrastructure Grant)
Country [3] 311072 0
Australia
Funding source category [4] 314019 0
Charities/Societies/Foundations
Name [4] 314019 0
Australasian Society of Ultrasound in Medicine
Country [4] 314019 0
Australia
Primary sponsor type
University
Name
Robinson Research Institute, University of Adelaide
Address
Norwich House
Ground Floor, 55 King William Rd
North Adelaide SA, 5006
Country
Australia
Secondary sponsor category [1] 312405 0
University
Name [1] 312405 0
Australian Institute of Machine Learning, University of Adelaide
Address [1] 312405 0
Corner Frome Road and, North Terrace, Adelaide SA 5000
Country [1] 312405 0
Australia

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 309423 0
University of Adelaide Human Research Ethics Committee
Ethics committee address [1] 309423 0
Office of Research Ethics, Compliance and Integrity
The University of Adelaide
Level 4, Rundle Mall Plaze
50 Rundle Mall
Adelaide SA 5000 Australia
Ethics committee country [1] 309423 0
Australia
Date submitted for ethics approval [1] 309423 0
15/08/2019
Approval date [1] 309423 0
23/04/2020
Ethics approval number [1] 309423 0
H-2020-051

Summary
Brief summary
Endometriosis is a chronic, inflammatory condition which can lead to chronic pelvic pain and infertility. There is no cure for this condition and the gold standard for diagnosis is laparoscopy (keyhole surgery) which is costly, has long wait times and is associated with risks. This study (Imagendo) will use artificial intelligence to create a diagnostic algorithm by analysing ultrasound and MRI endometriosis scans, providing general practitioners with an earlier, easily accessed, non-invasive, diagnosis of endometriosis.
Trial website
www.imagendo.org.au
Trial related presentations / publications
Public notes
Presentations: 2023: Australian Society for Ultrasound in Medicine (ASUM): Sydney, Avery J et al. (Invited speaker) Enhancing the detection of Pouch of Douglas obliteration for endometriosis diagnosis with Artificial Intelligence, using magnetic resonance imaging and unpaired endometriosis ultrasounds.
2023: 12th Congress of the Asia Pacific Initiative on Reproduction (ASPIRE 2023) Adelaide. Avery J et al (Oral) EXTRAPOLATING ENDOMETRIOSIS DIAGNOSIS USING IMAGING AND MACHINE LEARNING: THE IMAGENDO PROJECT
2023: 20th IEEE International Symposium on Biomedical Imaging (ISBI): Colombia: Zhang Y, et al. (Oral) “Distilling Missing Modality Knowledge from Ultrasound for Endometriosis Diagnosis with Magnetic Resonance Images”. Winner: best Oral Presentation
2023: World Congress on Endometriosis (WCE), Zhang Y, Avery JC (Presenter), et al Oral: A multimodal AI analysis of endometriosis imaging markers”.
Posters: Deslandes A, et al A quantitative grading system for the assessment TVUS image quality
2023: ARGANZ Adelaide, White S, et al (Poster) Development and validation of a machine learning system for automated routine 2-dimensional morphometric measurements on female pelvic MRI
2023: Computer Vision and Pattern Recognition Conference (CVPR 2023) Wang H, et al. (Oral) "Multi-modal Learning with Missing Modality via Shared-Specific Feature Modeling".
45th IEEE Engineering in Medicine and Biology Society, (EMBC 23) Sydeney, Butler D, et al.(Oral) “The Effectiveness of Self-supervised Pre-training for Multi-modal Endometriosis Classification”
2022: RANZCOG Meeting Gold Coast, Avery J et al IMAGENDO – non-invasive diagnosis of endometriosis using machine learning
11th Congress of the Asia Pacific Initiative on Reproduction ASPIRE: Warner R, Avery J et al (Oral) Associations between environmental exposures in the middle east area of operations and reproductive outcomes in Australian defence force veterans
2022: RCOG – BSGI Meeting Hull L, Avery J et al. Oral and Poster IMAGENDO: Combining ultrasound and magnetic resonance imaging using artificial intelligence to reduce diagnostic delay. (Virtual) London.
2022: Australian Society for Ultrasound in Medicine (ASUM) Meeting Adelaide Avery J. (Poster) Imagendo® – Non-Invasive diagnosis of endometriosis using machine learning. (Oral)
2022: European Conference on Computer Vision. Uncertainty-aware Multi-modal Learning via Cross-modal Random Network Prediction Wang, H. Avery J, et al. Uncertainty-Aware Multi-modal Learning via Cross-Modal Random Network Prediction.
2021: World Congress on Endometriosis (WCE) Maicas G, Avery J et al (Oral) Artificial Intelligence for Sliding Sign Detection to Diagnose Endometriosis.
2020: ISUOG Virtual World Congress Leonardi M, Avery J, et al (Poster) Machine learning to diagnose rectouterine pouch obliteration with the sliding sign on transvaginal ultrasound. 2020 UOG 56(S1)

Publications:
Butler D, Wang H, Zhang Y, To MS, Avery JC, Hull ML, Carneiro G. The Effectiveness of Self-supervised Pre-training for Multi-modal Endometriosis Classification, Proceedings of the 45th IEEE Engineering in Medicine and Biology Society, 2023 In Press.

H Wang, Y Chen, C Ma, AVERY J, L Hull, G Carneiro
Multi-Modal Learning With Missing Modality via Shared-Specific Feature Modelling. - Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp15878-15887 2023

Zhang Y, Wang H, Avery J, et al. “Distilling Missing Modality Knowledge from Ultrasound for Endometriosis Diagnosis with Magnetic Resonance Images”. Proceedings of 20th IEEE International Symposium on Biomedical Imaging (ISBI). In Press. https://ieeexplore.ieee.org/xpl/conhome/1000080/all-proceedings

Wang H, Zhang J, Chen Y, Ma C, AVERY J, Hull L, Carneiro G, Uncertainty-Aware Multi-modal Learning via Cross-Modal Random Network Prediction. Computer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part XXXVII, Pp 200-217

Leonardi M, Macais G, Avery J, Panuccio C. Carneiro G, Hull ML, Condous G. Machine learning to diagnose rectouterine pouch obliteration with the sliding sign on transvaginal ultrasound. ISUOG Virtual World Congress 2020 Ultrasound in Obstetrics and Gynecology 56(S1) doi: 10.1002/uog.23327

Maicas G, Leonardi M, Avery J, Panuccio C, Carneiro G, Hull ML, Condous G. Deep learning to diagnose pouch of Douglas obliteration with ultrasound sliding sign. Reprod Fertil. 2021 Aug 25;2(4):236-243. doi: 10.1530/RAF-21-0031. eCollection 2021 Dec.

Contacts
Principal investigator
Name 114110 0
Prof Mary Louise Hull
Address 114110 0
Robinson Research Institute
Ground Floor, 55 King William Rd
North Adelaide SA 5006
Country 114110 0
Australia
Phone 114110 0
+61403933312
Fax 114110 0
Email 114110 0
[email protected]
Contact person for public queries
Name 114111 0
Jodie C Avery
Address 114111 0
Robinson Research Institute
Ground Floor, 55 King William Rd
North Adelaide SA 5006
Country 114111 0
Australia
Phone 114111 0
+61410519941
Fax 114111 0
Email 114111 0
[email protected]
Contact person for scientific queries
Name 114112 0
Jodie C Avery
Address 114112 0
Robinson Research Institute
Ground Floor, 55 King William Rd
North Adelaide SA 5006
Country 114112 0
Australia
Phone 114112 0
+61410519941
Fax 114112 0
Email 114112 0
[email protected]

Data sharing statement
Will individual participant data (IPD) for this trial be available (including data dictionaries)?
No
No/undecided IPD sharing reason/comment


What supporting documents are/will be available?

No Supporting Document Provided



Results publications and other study-related documents

Documents added manually
No documents have been uploaded by study researchers.

Documents added automatically
No additional documents have been identified.