AI Based Real Time Detection of Endometriosis Lesions

Development of AI-Based Approaches for Automated Real-Time Detection of Endometriosis Lesions Using Endoscopic Image and Video Data

Development of AI-based approaches for automated real-time detection of endometriosis lesions using endoscopic image and video material.

Study Overview

• Status: Completed
• Conditions: Endometriosis

Detailed Description

In the field of endometriosis, artificial intelligence (AI) has been used for diagnoses or even predictions of endometriosis before confirmation through laparoscopy. AI's significant potential in minimally invasive surgery lies in automatic image analysis, aiding in the detection of structures or anomalies based on image data. This offers the potential to detect endometriosis lesions during laparoscopy regardless of the indication. Training and creating such AI models are done using machine learning algorithms based on annotated data. These training data consist of image data with pixel-level annotations of the content that the model should detect. Deep learning (DL) algorithms have proven effective in image analysis, relying on neural networks to autonomously fill them with the most critical decision criteria for correct analysis of the image content. The trained AI model can then be applied to unknown data, providing the probability of detecting a structure for each pixel. Possible visual outputs of the model include outlining the detected content or segmenting, assigning predefined content to each pixel. The quality of the model depends crucially on a sufficiently large number and quality of training data. Quality includes correct annotation of data to prevent the model from learning errors. Diversifying image data by including negative examples in the training and test datasets is equally important. The F1-score is used as a measure of the model's quality, combining precision (P) with recall (R) to a value between 0 and 1, based on an annotated test dataset.

The goal is to achieve a high F1-score through the selection of training data and an appropriate DL algorithm. Parameters like image preparation optimization or DL algorithm parameters such as selecting different neural networks can improve the F1-score. The number of required training data for a good AI model depends on the complexity of the question and the number of contents to be detected, as the model can only recognize learned content. It is possible to iteratively adjust the selection of training data for different questions based on the achieved F1-scores after each training and testing. If necessary, the number of training data can be increased, and problematic image data, such as missing annotations, can be identified and corrected based on the results.

• Study Type: Observational
• Enrollment (Actual): 26

Contacts and Locations

Study Locations

• Germany
  • Tübingen, Germany, 72076
    • University Hospital Tuebingen, Department of Women's Health

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

Included are patients who present themselves at the University Women's Clinic as part of the outpatient clinic in the Endometriosis Center. Women with a suspected diagnosis or confirmed diagnosis of endometriosis who have an indication for laparoscopic assessment are included.

Description

Inclusion Criteria:

• Age ≥ 18 years
• Written consent after explanation
• Indication for surgical treatment of endometriosis

Exclusion Criteria:

• Expected lack of patient compliance or inability of the patient to understand the purpose of the clinical trial
• Absence of patient consent
• Malignancies

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort
Suspected endometriosis

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Development and validation of an AI model for real-time automated detection of endometriosis lesions	Based on laparoscopic image and video data.; Evaluation of model accuracy using the F1-score, with a target value of ≥ 0.7.	During time-span of study (approx. 1 year)
Quality of video anonymization	Assessment of the effectiveness of the "InOut" AI model v0.2 in identifying and removing non-relevant image data.; Quality assurance through manual review of anonymized data. The videos are correlated with the following anonymized metadata, which are also transferred to KS; Age group of the patient (18-25; 25-30; 35-40,…); Weight class of the patient (BMI <17.5; 17.5-19; >19-25; >25-30; >30); Type of surgery (laparoscopy with or without treatment of endometriosis); Total duration of the operation; Complications during the operation (yes/no); Endoscopic devices used, especially the camera; Existing pathological findings related to endometriosis	During time-span of study (approx. 1 year)
Creation of a high-quality annotated image dataset for AI training	Target: 80-90% of selected images should contain endometriosis lesions, with the remaining being negative samples.; Annotation performed by medical professionals Clinically trained personnel at the University Hospital Tübingen (UKT) select 300 varied JPEG images from each anonymized video for annotation. The aim is to include 80%-90% of images displaying endometriosis lesions, with the remainder depicting other tissue abnormalities or no lesions.	During time-span of study (approx. 1 year)

Collaborators and Investigators

• Sponsor: University Hospital Tuebingen

Study record dates

Study Major Dates

• Study Start (Actual): October 10, 2023
• Primary Completion (Actual): January 28, 2025
• Study Completion (Actual): June 30, 2025

Study Registration Dates

• First Submitted: January 17, 2025
• First Submitted That Met QC Criteria: March 6, 2025
• First Posted (Actual): March 12, 2025

Study Record Updates

• Last Update Posted (Actual): January 29, 2026
• Last Update Submitted That Met QC Criteria: January 27, 2026
• Last Verified: January 1, 2026

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Urogenital Diseases; Genital Diseases; Female Urogenital Diseases; Female Urogenital Diseases and Pregnancy Complications; Genital Diseases, Female; Endometriosis
• Other Study ID Numbers: MT_STORZ federated learning

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No