Neoplasms

Skin and Connective Tissue Diseases

All Conditions

Breast Neoplasms

Breast Diseases

Skin Diseases

No interventions are administered. Data is retrospectively collected in an anonymized way after ethical approval at each site.

No intervention.

ODELIA is a project that aims to improve breast cancer detection in magnetic resonance imaging by utilizing artificial intelligence and swarm learning (MRI). The project will create an open-source swarm learning software framework that will be used to train AI models for breast cancer detection. These models' performance will be compared to that of conventional AI models, and the results will be used to assess the effectiveness of swarm learning in improving the accuracy and robustness of AI models. The project will use retrospective, anonymized breast MRI datasets with manual ground truth labels for cancer presence. The study is not associated with any patient treatment or intervention. The project's goal is to provide evidence of the clinical benefits of swarm learning in the context of breast cancer screening, such as accelerated development, improved performance, and robust generalizability.

Artificial Intelligence (AI) is set to revolutionize healthcare as its diagnostic performance approaches that of clinical experts. In particular, in cancer screening, AI could help patients to make better-informed decisions and reduce medical error. However, this requires large datasets whose collection faces severe practical, ethical and legal obstacles. These obstacles could potentially be overcome with swarm learning (SL) where partners jointly train AI models without sharing any data. Yet, access to SL technology is currently limited because no studies have implemented SL in a true multinational setup, no freely usable implementation of SL is available, researchers \& healthcare providers have no experience with setting up SL networks and policymakers are currently unaware of the broader implications of SL.

ODELIA will aim to solve these issues: ODELIA will build an open-source software framework for SL, providing an assembly line for the streamlined development of AI solutions in a preclinical setting. To serve as a blueprint for future SL-based AI systems, ODELIA partners collaborate as a consortium to develop AI models for the detection of breast cancer in magnetic resonance imaging (MRI). The size of ODELIA's distributed database will be substantial and ODELIA's AI models could reach expert-level performance for breast cancer screening.

Thereby, ODELIA will could not just deliver a useful medical application, but provide evidence to summarize the clinical benefit of SL in terms of accelerated development, increased performance and robust generalizability.

To achieve this, ODELIA partners will collect retrospective, anonymized breast MRI datasets with manual ground truth labels for the presence of cancer, and will train AI models conventionelly and via SL. The performance of these technical approaches will be compared. The aim of the study is to test the methodology of Swarm Learning and the performance of AI algorithms developed within ODELIA on retrospective data. There will be no effect on treatment of patients as all evaluations will be done retrospectively. No patient treatment or any intervention is associated with the study.

Female patients only, as defined per the European Society of Breast Imaging (EUSOBI) screening guidelins of 2023

A Retrospective Analysis of Magnetic Resonance Imaging Data for Breast Cancer Screening in the Open Consortium for Decentralized Medical Artificial Intelligence

Data are collected at multiple study centers and the decision whether or not to release IPD is made at each study center.

Diagnostic performance for breast cancer detection (Sensitivity and specificity) compared to the gold stnandard method of expert-based assessment of breast MRI, may be summarized in a receiver operating characteristic curve for multiple threshold values, comparing multiple technical approaches, including swarm-learning based AI models and local AI models.

European Institute for Biomedical Imaging Research (EIBIR), Austria

University Hospital, Aachen

Vall d'Hebron Institute of Oncology

Mitera Hospital

Radboud University Medical Center

UMC Utrecht

Ribera Salud Hospitals, Spain

Fraunhofer Institute for Digital Medicine (MEVIS), Germany

University Hospital, Zürich

Cambridge University Hospitals NHS Foundation Trust

Technische Universität Dresden

Urinary Tract, Sexual Organs, and Pregnancy Conditions

Rare Diseases

Uterine Cervical Neoplasms

Uterine Cervical Dysplasia

Uterine Cervical Diseases

Uterine Diseases

Genital Diseases, Female

Genital Diseases

Urogenital Diseases

Female Urogenital Diseases

Pregnancy Complications

Female Urogenital Diseases and Pregnancy Complications

Precancerous Conditions

Cervical Intraepithelial Neoplasia

Imaging of study participants will be performed during the colposcopy examination. The mobile colposcope and the confocal imaging probe, together referred to as the Multimodal Mobile Colposcope (MMC), will be used to image and/or record videos the cervix. Cervical biopsies will be performed using biopsy forceps per standard protocols.

Women referred for colposcopy due to abnormal cervical screening or follow-up for a history of dysplasia will be recruited into the experimental group of the study. Imaging of study participants will take place during the colposcopy examination.

Multimodal Mobile Colposcope (MMC)

Júlio César P Resende, MD, PhD

Mila P Salcedo, MD, PhD

Júlio César P Resende, MD,PhD

Cervical cancer remains the first or second leading cause of cancer death among women in many low-and middle-income countries. Cervical cancer prevention programs in low-resource settings are hampered by a lack of personnel with appropriate clinical expertise, lack of pathology services, and lack of associated infrastructure. There is an urgent need for appropriate diagnostic tools to enable accurate screening and diagnosis in low-resource settings. The purpose of this study is to develop and validate a low-cost Multimodal Mobile Colposcope (MMC) for global cervical cancer prevention programs. This new device will combine the imaging capabilities of a mobile colposcope with the microscopic imaging capabilities of a fiber-optic confocal imaging probe.

Low-cost Imaging Technology for Global Prevention of Cervical Cancer

Low-cost Mobile Colposcopy and Confocal Imaging for Global Prevention of Cervical Cancer

Image data will be used as a training set to develop and refine image analysis algorithms. Multimodal Mobile Colposcope will be compared with colposcopic impression considering clinical performance (sensitivity, specificity, positive and negative predictive values, and area under the Receiver Operating Characteristic (ROC) curve)

William Marsh Rice University

M.D. Anderson Cancer Center

Irmandade Santa Casa de Misericórdia de Porto Alegre

Barretos Cancer Hospital

Diseases and Abnormalities at or Before Birth

Hemangioma

Hemangioma, Capillary

Port-Wine Stain

Neoplasms, Vascular Tissue

Neoplasms by Histologic Type

Congenital Abnormalities

Skin Abnormalities

Anti-Arrhythmia Agents

Antihypertensive Agents

Vasodilator Agents

All Drugs and Chemicals

Propranolol

Timolol

Maleic acid

Dosage of 1 mg/kg per day divided 2 times daily. Blood pressure, heart rate and oxygen saturation are monitored after propranolol initiation. Treatment is continued with a gradual increase to 2 mg/kg per day divided 2 times daily. Treatment is continued until the hemangioma no longer changes in the characteristics judged by the physicians, including, color, size, temperature and deformability.

One drop of of timolol maleate 0.5% gel is topically applied and massaged into the hemangioma twice per day . This dosage provides an estimated 0.5 mg of timolol per day. The treatment is continued until it is considered to be no longer effective as judged by the physicians.

No oral or topical treatment will be given as recommended by the treating physicians and elected by the parents. Patients will be evaluated periodically to determine what changes in treatment are warranted. If this occurs, the patients will be included in the appropriate study group.

Marty O Visscher, PhD

Denise Adams, MD

Adrienne Hammill, MD, PhD

A combined set of quantitative skin imaging methods will quantitatively describe the natural ontogeny and the response to standard treatments over time in patients with infantile hemangiomas.

Infantile hemangiomas (IH) are relatively common benign vascular neoplasms, which appear shortly after birth and grow rapidly in the following weeks and months. The proliferative phase involves a rapidly dividing endothelial cell proliferation. The tumors generally stabilize and then involute over time. They are usually clinically insignificant; however, a small percentage of them can be life threatening secondary to location, associated syndrome, and physical findings such as ulceration. Treatment options have included steroids, interferon, vincristine and more recently propranolol. There are few published prospective research studies on the effectiveness of these treatment options. Because of several factors, it has been difficult to objectively measure response of these lesions. The purpose is to determine (1) whether multiple quantitative skin imaging modalities can quantitatively describe the clinically relevant features of infantile hemangiomas, including color (red, blue), lightness, size (height, volume), biomechanical properties, temperature and perfusion and (2) the natural ontogeny and response to treatment over time. The treatments are oral propranolol, topical timolol and untreated (observation).

Ontogeny of Infantile Hemangiomas With Skin Imaging Modalities

Ontogeny and Quantitative Multimodal Skin Imaging of Infantile Hemangiomas

IH height and volume will be obtained from surface scans with the Artec scanner with a 0.5 mm resolution, 0.1 mm accuracy, at 60 cm. Three dimensional scans will be created with the Artec Studio software and features quantified with the three dimensional software. Soft tissue landmarks will create the region of interest for all evaluations. Height and volume will be calculated for the infantile hemangioma versus the control.

Temperature information will be acquired with a thermal imaging camera with analysis software. The thermal features of the surrounding uninvolved control skin are removed by applying thresholds. Static infrared thermography captures the steady-state condition and spatial distribution. Dynamic infrared thermography applies a stress (cooling) to stimulate the thermal response within the tissue.

High resolution color Images will be taken taken with a digital camera, Micro 60 mm lens with a wireless close up flash system at 30cm and perpendicular to the site. The images are color corrected and processed into three distinct images : red color, blue-yellow color, and lightness (white - back).

Color and thermal images are co-registered and analyzed with customized software. The system creates quantitative outputs and maps for visualization of IH regions of thermal and color activity. Threshold values for the lightness, red color, blue color and thermal images are selected based upon histogram distributions of the uninvolved skin as the values above the mean and multiple standard deviations The pixels beyond the thresholds are used to quantify and map the features due to the hemangioma itself.

Tissue mechanical properties (i.e., elasticity, elastic deformation, laxity, stiffness, etc.) will be measured to assess tissue deformability. The instrument measures biomechanical properties of the skin by applying a negative pressure, in the range between 20 and 500 mbar. These properies are compared to an uninvolved contralateral control site.

Study title	NCT ID	Conditions	Interventions	Eligibility	Organization
A Retrospective Analysis of Magnetic Resonance Imaging Data for Breast Cancer Screening in the Open Consortium for Decentralized Medical Artificial Intelligence	NCT05698056	Breast Cancer	No intervention...	18 Years - 90 Years	Technische Universität Dresden
Low-cost Imaging Technology for Global Prevention of Cervical Cancer	NCT05078528	Cervical Intrae...	Multimodal Mobi...	25 Years -	Barretos Cancer Hospital
Ontogeny of Infantile Hemangiomas With Skin Imaging Modalities	NCT02061735	Infantile Heman...		1 Month - 5 Years	Children's Hospital Medical Center, Cincinnati
Ontogeny of Infantile Hemangiomas With Skin Imaging Modalities	NCT02061735	Infantile Heman...		1 Month - 5 Years	Children's Hospital Medical Center, Cincinnati

Spots Global Cancer Trial Database for image analysis