Artificial Intelligence (AI), Lung Cancer Screening Program
Conditions
Keywords
CT scan low-dose, Artificial Intelligence, High-risk sucjects, Lung cancer
Brief summary
The study is an interventional pilot study. The study is designed to be monocentric and it presents additional procedues.
Detailed description
Interventional pilot study, single-center with additional procedures, such as completion of EORTC-QLQ-LC29, EORTC-QLQ-C30 questionnaires, motivational test, Fagestrom test, anamnestic questionnaire, spirometry, measurement of carbon monoxide, Low-dose spiral computed tomography without contrast medium, peripheral venous blood sampling for a volume of 20 ml. The study has the main objective of traininig and validate a reliable and unbiased Artificial Intelligence (AI) algorithm that detects the presence of nodules and differentiates between malignant or benign tumor types. The study considers patients with suspected diagnosis or with a dignosis of lung cancer, smokers and former smokers over 50 years of age at high risk of lung cancer and subjects enrolled in previous screening cohorts at this Institute.
Interventions
DIAGNOSTIC_TESTComputed tomography (CT) scan low dose
The radiological investigation will be done with multi-detector-row (64 or more) computed tomography (CT) scanners at low-dose protocol. The low-dose spiral CT consists of a CT study of the chest, without the need for injection of contrast medium, characterized by less radio exposure than the standard CT of the chest with high sensitivity in detecting pulmonary nodules.
PROCEDUREBlood sampling
Peripheral venous blood sampling (20 ml)
PROCEDURETissue sampling (lung)
sampling of tumor and healthy tissue during surgery
OTHERSpirometry
Spirometry measurement using spirometer
OTHERQuestionnaires
Compilation of epidemiological questionnaire, quality of life questionnaires
Study guarantee valid support for quitting smoking, which for a smoker is a more effective intervention to reduce the risk of developing lung cancer, myocardial infarction and other smoking-related diseases
OTHERCarbon monoxide measurment
Measurment of Carbon monoxide (CO)
OTHERCardiovascular primary prevention
Intervention done in order to find the presence of coronary calcifications
Sponsors
Scientific Institute San Raffaele
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
SCREENING
Masking
NONE
Masking description
No blinding in this study
Intervention model description
In the study are considered 4 different cohorts. Lung cancer patients diagnosed outside screening and treated at San Raffaele Hospital; smokers and former smokers aged over 50 years old at high risk for lung cancer, cardiovascular diseases and chronic obstructive pulmonary disease (COPD); subjects enrolled in previous screening cohorts in this Institute, which presented a computed tomography (CT) with the presence of lung nodules \>4 mm and to conclude, subjects enrolled in previous screening cohorts in this Institute, which presented a negative computed tomography (CT).
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
High-risk screening subjects Inclusion Criteria: * Age \>= 50 years old * Active smokers * Former smokers (from no more than 15 years) * Pack/year \>20 * Risk-prediction model from Prostate, Lung, Colorectal, and Ovarian study (PLCOm2012) \>1.2% * Provision and signature of informed consent
Exclusion criteria
* Previous or concurrent neoplastic disease, excluding skin cancers * Cognitive or other problems that could hinder the collection of informed consent * Severe pulmonary or extra pulmonary disease * Previous low-dose computed tomography (CT) scan in the past 12 months Previous high-risk positive screening subjects Inclusion Criteria: * Subjects enrolled in previous lung cancer screening with the presence of lung nodules \>4 mm and candidate to additional computed tomography (CT) * Signed informed consent
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Creation of Artificial Intelligence (AI) algorithm | from enrollment to 48 months | To train and validate a reliable and unbiased Artificial Intelligence (AI) algorithm that detects the presence of nodules and differentiates between malignant or benign tumor types. AUC (Area Under the Curve) values, expressed as mean and standard deviation (SD), comparing the ability in detecting the presence of nodules and differentiating the malignancy or benignity of a radiologist versus an AI algorithm, both trained on the same patient group. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Multimodal program | from enrollment to 48 months | Develop a multimodal program to enhance the prevention and the early detection of multiple smoking-related diseases Presence and absence of lung nodules \> 4 mm with computed tomography (CT) scan |
Countries
Italy
Contacts
Primary ContactPiergiorgio Muriana, MD
Outcome results
None listed