Neoplasms

Respiratory Tract (Lung and Bronchial) Diseases

All Conditions

Lung Neoplasms

Lung Diseases

Respiratory Tract Diseases

Respiratory Tract Neoplasms

Thoracic Neoplasms

All patients will undergo EBUS-TBNA as per routine care, except for the one difference where the procedures will be video-recorded so that they can be used for computer analysis at a later time. Static images will be obtained from EBUS videos in order to perform segmentation. Segmentation will be conducted by both an experienced endoscopist and NeuralSeg.

Endobronchial Ultrasound

Wael C Hanna

This study aims to determine if a deep neural artificial intelligence (AI) network (NeuralSeg) can learn how to assign the Canada Lymph Node Score to lymph nodes examined by endobronchial ultrasound transbronchial needle aspiration(EBUS-TBNA), using the technique of segmentation. Images will be created from 300 lymph nodes videos from a prospective library and will be used as a derivation set to develop the algorithm. An additional100 lymph node images will be prospectively collected to validate if NeuralSeg can correctly apply the score.

The Use of Artificial Intelligence to Predict Cancerous Lymph Nodes for Lung Cancer Staging During Ultrasound Imaging

Development and Validation of a Computer-aided Algorithm Using Artificial Intelligence and Deep Neural Networks for the Segmentation of Ultrasonographic Features of Lymph Nodes During Endobronchial Ultrasound

Objective: to determine whether a deep neural AI network (NeuralSeg) can learn how to assign the Canada Lymph Node Score to lymph nodes examined by EBUS, using the technique of segmentation on an existing (derivation) set of lymph node videos

Objective: to determine whether NeuralSeg can correctly apply the Canada Lymph Node Score to a new (validation) set of lymph node videos that it has never seen before

Objective: to compare the accuracy and reliability of the segmentation performed by NeuralSeg to the segmentation performed by an experienced endoscopic surgeon using DICE-SORENSEN coefficients.

Objective: to determine whether NeuralSeg can accurately predict malignancy in lymph node when compared to biopsy results of the lymph node that was examined.

St. Joseph's Healthcare Hamilton

Dr. Waël Hanna, MDCM, MBA, FRCSC

This is a single-center observational feasibility study to develop an automatic segmentation and registration algorithm based on ultrasound imaging of the pelvic arteries. The duration of this study will be approximately 1 year. Patients scheduled for a laparotomy at the NKI are eligible for inclusion. Patients are informed about the study before the planned surgery and, after being provided with the necessary information regarding participation in the study, will be asked for informed consent. The ultrasound acquisitions for this study will be performed intra-operatively with CE marked equipment for intra-operative ultrasound. There is no impact on the standard surgical procedure or decision making of the surgery. After surgery, no further participation or cooperation of the patient is required.

A patient-specific 3D model will be created using an available pre-operative CT scan. Anatomical target points are selected on this virtual model before the start of the surgery. Prior to surgery, a patient-reference electromagnetic (EM) sensor will be placed between the patient and the matrass on the operating table to account for patient movement during acquisition. Intra-operatively, an initial point registration of the 3D model with the electromagnetic tracking system (EMTS) will be performed based on ultrasound (US) imaging of the arterial bifurcations by the surgeon. Then, the surgeon will acquire multiple US sweeps of the pelvic arteries (abdominal aorta and left and right iliac arteries). For validation purposes, the pre-operatively defined anatomical target points will be visualized on US imaging and by pinpointing the EM tracked pointer. All tracking and US data will be recorded and stored for post-operative analysis.

Intra-operative ultrasound measurement

Marijn Hiep, MSc

Theo Ruers, Prof. Dr.

Wout Heerink, Dr.

Harald Groen, Dr.

Theo Ruers, prof. dr.

In this study we aim to develop an automatic pelvic artery segmentation algorithm, which is required for future clinical implementation of US registration for surgical navigation. Various registration methods will be evaluated with the data of this study to obtain most optimal results. If automatic segmentation and registration is successful, the final accuracy of the developed US registration method for tumor tracking should be evaluated in future studies in patients eligible for surgical navigation. Eventually, we aim to replace the CBCT-scan with an automatic tracked US registration pipeline for a more efficient and accurate registration procedure, which could improve the applicability and accuracy of surgical navigation and patient outcomes.

Image-guided navigation surgery allows for full utilization of pre-operative imaging during surgery and has the potential of reducing both irradical resections and morbidity. To use navigation, a registration procedure is required to correlate pre-operative imaging with the patient's position on the operating room (OR). Currently, registration is done by Cone-Beam CT (CBCT) scanning on the OR prior to navigation surgery. However, the main limitation of the CBCT method is that it cannot compensate for per-operative changes such as bed rotation, retractor placement and tissue displacement due to the surgery. Alternatively, by using intra-operative tracked ultrasound and vessel-based patient registration, changing conditions during surgery can better be dealt with. This improved patient registration method could lead to an increased navigation accuracy and improved clinical usability and outcomes.

The main difference between CBCT and proposed ultrasound registration is that CBCT is based on bones, while the ultrasound is based on vessels. Bones can be very easily imaged on the CBCT and therefore used for bone-bone registration with pre-operative CT-scans. However, vessels are more difficult to acquire, especially with ultrasound, and an automatic registration process with pre-operative imaging is needed for efficient clinical usability. For this, the vessels need to be extracted from the tracked ultrasound images to create a 3D representation that can be registered. Therefore, an algorithm needs to be developed that can automatically segment the pelvic vessels from ultrasound images.

One group to evaluate the feasibility and accuracy of the proposed methods.

Vessel-based Ultrasound Registration

Intra-operative Pelvic Vessel Acquisition, Towards Ultrasound Registration for Surgical Navigation

An automatic segmentation algorithm for real-time intra-operative pelvic vessel segmentation from tracked US images will be developed. The accuracy of this network will be evaluated using the Dice similarity coefficient, ranging from 0 to 1 where a higher score means a better outcome.

To evaluate the accuracy of different patient registration methods of intra-operative US imaging with pre-operative CT imaging, such as 3D model or centerline registration.

Evaluation of the intraoperative usability of the tracked US device and visualization of pelvic vessels by the surgeons will be evaluated using the system usability scale (SUS). This scale ranges from 0-100 where a higher score means a better outcome.

The Netherlands Cancer Institute

Carcinoma, Non-Small-Cell Lung

Carcinoma

Carcinoma, Bronchogenic

Bronchial Neoplasms

an automated deep learning detection and segmentation software for non-small cell lung cancer (NSCLC) that can automatically detect and segment tumors on CT scans and thus reduce the human variation.

Automatic detection and segmentation of NSCLC tumors

Accurate segmentation of lung tumor is essential for treatment planning, as well as for monitoring response to therapy. It is well-known that segmentation of the lung tumour by different radiologists gives different results (inter-observer variance). Moreover, if the same radiologist is asked to repeat the segmentation after several weeks, these two segmentations are not identical (intra-observer variance). In this study we aim to develop an automated pipeline that can produce swift, accurate and reproducible lung tumor segmentations.

In this study, we aim to develop and test an automated deep learning detection and segmentation software for non-small cell lung cancer (NSCLC) that can automatically detect and segment tumors on CT scans and thus reduce the human variation. We will assess the level of agreement between a group of radiologists, performing manual versus semi-automatic tumour segmentation. To do so, we will provide radiologists with two sets of CT scans. The first set will be segmented manually; the second one will be segmented using the automated software program.

Subsequently, we will use the inter- and intra-observer variance from the clinical study in a simulation or modeling study. We also compare the time needed and the consistency in segmentations by the software to medical doctors performance.

Reliability and Agreement study:

Primary tumours of 25 lung cancer patients will be delineated by 6 segmentation experts.

1. Assess agreement between automatic segmentation and radiologists' segmentation The primary tumours of 25 patients will be manually segmented by the radiologists and automatically by the the tool. The time needed to perform this task and the reproducibility of the segmentation will be recorded. The degree of overlap between the ROs and the automatic contour will be assessed pairwise using the Dice coefficient.
2. Delination of tumours by the experts, assisted by the software tool For another 25 patients, the experts will be provided with an automatic delineation, performed by the tool. They have the possibility to adjust and validate it. The time needed will be recorded. The difference between the mean overlap fraction in the first situation (manual delineation of experts) and the second situation (delineation of experts+ software tool) will be assessed, using a multi-observer Dice coefficient.
3. Assessment of intra-observer variance The experts will repeat the segmentation of the lung tumours after 2 weeks. They will repeat the manual segmentation (n=25) and the semi-automatic segmentation (n=25). This will make it possible to assess the intra-observer variance in both situations.
4. Qualitative assessment of the experts' preferences using an in-house developed visualization toolbox.

Fully Automated Pipeline for the Detection and Segmentation of Non-Small Cell Lung Cancer (NSCLC) on CT Images

Fully Automated Pipeline for the Detection and Segmentation of Non-Small Cell Lung Cancer (NSCLC) on CT Images: Quantitative and Qualitative Evaluation

Currently, there is no plan to make it public

Centre Hospitalier Universitaire de Liege

University Hospital RWTH Aachen University, Aachen, Germany.

Affiliated Zhongshan Hospital of Dalian University, 6 Jiefang street, Dalian 116001, China

University of California, San Francisco

Maastricht University

Digestive System Diseases

Colonic Neoplasms

Colorectal Neoplasms

Intestinal Neoplasms

Gastrointestinal Neoplasms

Digestive System Neoplasms

Gastrointestinal Diseases

Colonic Diseases

Intestinal Diseases

Herbal and Botanical

All Drugs and Chemicals

Cola

Andrea Laghi, MD

The goal of this clinical trial is to develop an artificial intelligence-based model to assess radiogenomics signature of colon tumor in patients with stage II-III colon cancer. The main question it aims to answer is:

• Can artificial intelligence-based algorithm of radiomics features combined with clinical factors, biochemical biomarkers, and genomic data recognise tumor behaviour, aggressiveness, and prognosis, identifying a radiogenomics signature of the tumor?

Participants will

* undergo a preoperative contrast-enhanced CT examination;
* undergo surgical excision of colon cancer
* undergo adjuvant therapy if deemed necessary based on current guidelines

The ATTRACT trial consist of a retrospective and a prospective part. In the retrospective part of the trial, radiology, pathology, and genomics data of 300 patients with stage II-III colon cancer will be used to identify the genetic and radiomics features of colon tumors and the clinical endpoints as the outcomes of the predictive model.

Tumors will be manually segmented on CT images and used for the AI (artificial intelligence)-model generation. Pathological annotations will be associated to the corresponding anonymised profiles. Immunohistochemistry will be used to classify the samples in the 4 molecular subtypes according.RNA-seq profiles will be also generated from tissue samples through targeted transcriptomics using custom NGS (next-generation sequencing) panels specifically designed to evaluate gene expression and assess Tumor Mutational Burden (TMB). Raw data will be processed and modelled using Topological Pathway Analysis to stratify patients according to the relevant molecular features and define molecular annotations that will be used to train the model for the identification of specific clinically relevant groups. Raw data together with radiological data will be used to generate and train the AI-models for the automated segmentation and the extraction of the radiogenomics signature.

Radiomics features will be extracted from manually segmented tumors. Standard PyRadiomics tools as well custom-made tools will be used. Feature robustness will be guaranteed by selecting only those with high inter-observer statistical correlation. Two families of AI models will be generated, one family dedicated to segmentation, and the other dedicated to radiogenomics-based phenotyping according to the clinical, molecular biology and pathological data available. The two families will be fused for the creation of the ATTRACT AI-model. For the generation of these models, specific convolutional neural network (CNN) architectures based on deep learning (DL) and Artificial Intelligence like UNet and MaskNet will be applied. The training will be performed using the manual ROI (region of interest) segmentations as ground truth. For the generation of radiogenomics analysis models, radiomics and genomic features will be combined using different multivariate algorithms. The classificatory will be trained to recognise the cancer subtypes and clinical endpoints.

In the prospective part of the trial, patients with stage II-III colon cancer will be recruited and will undergo a preoperative contrast-enhanced CT examination. The recruitment rate will be 70 patients per year, for a total of 210 patients. After pre-operative CT, surgery will be performed according to international standard protocols. Eventually adjuvant therapy will be considered following current guidelines. Pathological sample of the prospective enrollment will be analyzed. First, with RNA-seq data, TMB (of coding genes) and clinical data, patients will be clustered by making use of two different techniques Markov Cluster algorithm (MCL) and t-SNE (t-distributed stochastic neighbor embedding), Multi-Layer Network clustering. Patients will be represented as node of a network, edges between nodes will be weighted and thresholded according to the Jaccard Similarity. The similarity will be computed on top of Gene Expression, TMB, and perturbation information coming from Topological Pathway Analysis. Results of clustering will be matched with those coming from Immunohistochemistry. Clinical follow-up data (i.e. outcome of the therapy etc...) will be, once available, also plugged into the workflow to enforce the learning. Extracted knowledge will be used to annotate the dataset used to train and validate the radiomics classification. Gross specimen will be analyzed in order to extract different transcriptomics molecular subtypes (CMS1, CMS2, CMS3, CMS4) in accordance to the Colorectal Cancer (CRC) Subtyping Consortium (CRCSC) assessing the presence or absence of core subtype patterns among existing gene expression-based CRC subtyping algorithms. The accordance between pathological molecular profile and ctDNA analysis during protocol will be related to radiomics classification in order to provide a new whole-diagnostic model of approach in CRC treatment and surveillance.

Prospective data will be used to validate the AI models. For the segmentation models, the Dice Coefficient will be used as an indicator to measure the degree of overlap between the automated and the expert segmentation. For the radiogenomics model, performances will be evaluated using accuracy, integral under the receiver operator curve (ROC-AUC) and clinical decision curve. The investigators will also take into consideration, in order to select the best AI models, the response to the variation of the input characteristics and will produce saliency maps where the features of the input image that mostly contributed to the classification are highlighted.

Clinical evaluation with will be performed every 6 months for 2 years, including regular serum CEA (carcinoembryonic antigen) tests and Whole-Body CT every 6-12 months in patients who are at higher risk of recurrence in the first 3 years following ESMO (European Society for Medical Oncology) guidelines .Disease-free survival (DFS) and relapse-free survival (RFS) will be calculated.

ArTificial inTelligence-based RAdiogenomics in Colon Tumors

ATTRACT - ArTificial inTelligence-based RAdiogenomics in Colon Tumors

Associazione Italiana per la Ricerca sul Cancro

University of Roma La Sapienza

Full Professor

Urinary Tract, Sexual Organs, and Pregnancy Conditions

Prostatic Neoplasms

Genital Neoplasms, Male

Urogenital Neoplasms

Genital Diseases

Genital Diseases, Male

Urogenital Diseases

Prostatic Diseases

Male Urogenital Diseases

Vincent BOURBONNE, MD, PhD

Vincent Bourbonne, MD, PhD

Following the major technological and scientific advances in external radiotherapy in recent decades, thanks to the use of three-dimensional conformal techniques combined with intensity modulation, image-guided radiotherapy has enabled radiotherapists to increase doses without increasing sequelae and complications, giving rise to the term "dose escalation".

Following multiple dose-escalation clinical trials showing better biological control of PSA, the results of the latest phase 3 FLAME trial incorporated the notion of intraprostatic boost in relation to the primary prostate lesion, considered to be the preferred site of neoplastic recurrence in prostate cancer.

This leads to the first question, which concerns the identification of the dominant lesion and its precise delimitation. This last point is subject to variation between operators. A retrospective cohort from the Finistère region will therefore be used to develop a number of study points relating to :

inter-operator contour variability

* Factors influencing contour
* Impact of contour variability on dosimetry
* Automatic segmentation

Following the major technological and scientific advances in external radiotherapy in recent decades, thanks to the use of three-dimensional conformal techniques combined with intensity modulation, image-guided radiotherapy has enabled radiotherapists to increase doses without increasing sequelae and complications, giving rise to the term "dose escalation".

Following multiple dose-escalation clinical trials showing better biological control of PSA, the results of the latest phase 3 FLAME trial incorporated the notion of intraprostatic boost in relation to the primary prostate lesion, considered to be the preferred site of neoplastic recurrence in prostate cancer.

One of the issues raised by such a study is the methodology used to contour the tumour lesion, an issue which concerns the whole field of radiotherapy. The reference imaging technique for diagnosing prostate cancer, and more specifically the dominant tumour lesion, is multiparametric Magnetic Resonance Imaging. This leads to the first question, which concerns the identification of the dominant lesion and its precise delimitation. This last point is subject to variation between operators. A retrospective cohort from the Finistère region will therefore be used to develop a number of study points relating to :

inter-operator contour variability

* Factors influencing contour
* Impact of contour variability on dosimetry
* Automatic segmentation

Inter-observer Variability in the Segmentation of Prostate Tumour Lesions Using Multiparametric MRI (VARIOP)

Inter-observer Variability in the Segmentation of Prostate Tumour Lesions Using Multiparametric MRI

All collected data that underlie results in a publication

Contour variability using a deep-learning automated segmentation technique, with and without implementation of factors identified as impacting contour in manual technique.

The main criterion for judging inter-observer variability is the spatial overlap of contours on mpMRI, represented by the DICE similarity coefficient, which ranges from 0 to 1, where 1 represents zero variability between contours of the same lesion.

Subgroup analysis of radio-clinical-histological factors likely to influence inter-observer variability.

The influence of variability on dosimetry in IMRT/VMAT conformal radiotherapy.

Study title	NCT ID	Conditions	Interventions	Eligibility	Organization
The Use of Artificial Intelligence to Predict Cancerous Lymph Nodes for Lung Cancer Staging During Ultrasound Imaging	NCT03849040	Lung Diseases Lung Neoplasm	Endobronchial U...	18 Years -	St. Joseph's Healthcare Hamilton
Vessel-based Ultrasound Registration	NCT05637346	Abdominal Cance...	Intra-operative...	18 Years -	The Netherlands Cancer Institute
Fully Automated Pipeline for the Detection and Segmentation of Non-Small Cell Lung Cancer (NSCLC) on CT Images	NCT04164186	Detection Segmentation	Automatic detec...	-	Maastricht University
The Use of Artificial Intelligence to Predict Cancerous Lymph Nodes for Lung Cancer Staging During Ultrasound Imaging	NCT03849040	Lung Diseases Lung Neoplasm	Endobronchial U...	18 Years -	St. Joseph's Healthcare Hamilton
ArTificial inTelligence-based RAdiogenomics in Colon Tumors	NCT06108310	Colon Cancer St... Colon Cancer St...		18 Years -	University of Roma La Sapienza
Inter-observer Variability in the Segmentation of Prostate Tumour Lesions Using Multiparametric MRI (VARIOP)	NCT05996289	Prostate Cancer		18 Years -	University Hospital, Brest
Vessel-based Ultrasound Registration	NCT05637346	Abdominal Cance...	Intra-operative...	18 Years -	The Netherlands Cancer Institute
Inter-observer Variability in the Segmentation of Prostate Tumour Lesions Using Multiparametric MRI (VARIOP)	NCT05996289	Prostate Cancer		18 Years -	University Hospital, Brest
Fully Automated Pipeline for the Detection and Segmentation of Non-Small Cell Lung Cancer (NSCLC) on CT Images	NCT04164186	Detection Segmentation	Automatic detec...	-	Maastricht University

Spots Global Cancer Trial Database for segmentation

Every month we try and update this database with for segmentation cancer trials from around the world to help patients and their families find trials that might be right for them.
We offer this 100% free of charge and do not endorse any of the trials listed here, we hope it helps you or a loved one.