Validation and Optimisation of Ultrasound Diagnosis of Adenomyosis

Recruiting

• Conditions: Adenomyosis

• Registration Number: NCT06795711
• Lead Sponsor: IRCCS Azienda Ospedaliero-Universitaria di Bologna

Brief Summary

Defining ultrasound criteria for normal uterine biometry and assessing the prevalence of repeat abortions in patients with abnormalities of the uterine cavity

Detailed Description

Adenomyosis is a gynaecological disorder with a high prevalence in women of childbearing age and is characterised by the presence of glands and endometrial stroma within the myometrium, associated or not with hypertrophy and hyperplasia of the surrounding myometrium. Adenomyosis may cause pelvic pain and/or abnormal uterine bleeding. Transvaginal ultrasound may be considered the main non-invasive diagnostic modality for the diagnosis of adenomyosis. The aim is to optimise the ultrasound diagnosis of uterine pathology and in particular of adenomyosis by defining uterine biometric parameters (longitudinal, transverse and anteroposterior diameters and their ratios; uterine volume) allowing patients to be divided into 3 groups:

* Uterus affected by adenomyosis (group A): adenomyosis is a gynaecological condition with high prevalence in women of childbearing age and is characterised by the presence of endometrial tissue (innermost layer of the uterus) within the uterine muscle. Adenomyosis can cause abdominal pain and abnormal uterine bleeding.

* Uterus affected by fibromatosis (group B): uterine fibromatosis is a gynaecological condition characterised by the appearance of numerous fibroids in the uterus. It is a very frequent condition in the general population and its frequency increases as the age of the patients increases.

* Normal uterus (group C). Transvaginal ultrasound, although a reference diagnostic tool, still remains an operator-dependent examination to date: our secondary objective is to build models that can simplify diagnosis through the use of artificial intelligence. The aim is to create various artificial intelligence software that can 'learn to make a diagnosis'. This method has already been applied in radiology, proving capable of discriminating between benign and malignant tumours from images from different diagnostic methods with performance similar to that of experienced radiologists.

Study Timeline

• Study Start: 2022-04-04
• Status Verified: 2024-11
• Study First Submitted: 2024-12-03
• Primary Completion: 2024-12-31
• Study First Submitted QC: 2025-01-21
• Last Update Submitted: 2025-01-21
• Study First Posted: 2025-01-28
• Last Update Posted: 2025-01-28
• Study Completion: 2025-03-31

Recruitment & Eligibility

• Status: RECRUITING
• Sex: Female
• Target Recruitment: 465

Inclusion Criteria

• age between 18 and 60;
• obtaining informed consent

Exclusion Criteria

• Hysterectomised patients;
• Virgo patients (hymenal integrity);
• Patients reporting intolerance to transvaginal ultrasound;
• Gynaecological oncology;
• Recent pregnancy or childbirth (within 6 months);
• Menopausal patients

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Definition of uterine biometric parameters	After enrollment on first visit	Definition of uterine biometric parameters for the diagnosis of adenomyotic uterus (group A), fibromatous uterus (group B) and normal uterus (group C) by means of transvaginal ultrasound, performed as per the care procedure. Evaluation of the diagnostic capacity of 'globular uterus' for the diagnosis of adenomyosis as an additional parameter to those already known in the literature with possible subsequent identification of a biometric cut-off
Diagnostic capacity of 'globular uterus' for the diagnosis of adenomyosis	After enrollment on first visit	Evaluation of the diagnostic capacity of 'globular uterus' for the diagnosis of adenomyosis as an additional parameter to those already known in the literature with possible subsequent identification of a biometric cut-off

Secondary Outcome Measures

Name	Time	Method
Evaluation of diagnostic accuracy of deep learning validated	After enrollment on first visit	Evaluation of diagnostic accuracy of deep learning validated for ultrasound diagnosis of adenomyosis
Construction of deep learning models on uterine ultrasound images	After enrollment on first visit	Construction of deep learning models trained, validated and tested on uterine ultrasound images for the ultrasound diagnosis of adenomyosis and evaluation of their diagnostic accuracy
Identification of the frequency of finding ultrasound signs of adenomyosis in the cervix	After enrollment on first visit	In patients with a diagnosis of adenomyosis made on the basis of ultrasound features at the level of the uterine body and fundus
Evaluation of diagnostic accuracy	After enrollment on first visit	Evaluation of the diagnostic accuracy of trainees when experienced (identifying experienced operators as doctors in specialised training in Gynaecology and Obstetrics for at least four years, with an experience of at least 500 gynaecological ultrasound cases) and moderately experienced (identifying moderately experienced operators as doctors in specialised training in Gynaecology and Obstetrics for at least two years, with an experience of at least 200 gynaecological ultrasound cases

Trial Locations

Locations (1)

IRCCS Azienda Ospedaliero-Universitaria di Bologna; 🇮🇹 Bologna, Italy