Colorectal cancer (CRC) is a leading cause of cancer-related morbidity and mortality worldwide, with rates of CRC predicted to increase. Colonoscopy is currently the gold standard of screening for CRC. Artificial intelligence (AI) is seen as a solution to bridge this gap in adenoma detection, which is a quality indicator in colonoscopy. AI systems utilize deep neural networks to enable computer-aided detection (CADe) and computer-aided classification (CADx). CADe is concerned with the detection of polyps during colonoscopy, which in turn is postulated to help decrease the adenoma miss-rate. In contrast, CADx deals with the interpretation of polyp appearance during colonoscopy to determine the predicted histology. Prediction of polyp histology is crucial in helping Clinicians decide on a "resect and discard" or "diagnose and leave strategy". It is also useful for the Clinician to be aware of the predicted histology of a colorectal polyp in determining the appropriate method of resection in terms of safety and efficacy. While CADe has been studied extensively in randomized controlled trials, there is a lack of prospective data validating the use of CADx in a clinical setting to predict polyp histology. The investigators plan to conduct a prospective, multi-centre clinical trial to validate the accuracy of CADx support for prediction of polyp histology in real-time colonoscopy.
Colonoscopy is currently the gold standard of screening for CRC. A 1% increase in adenoma detection rate (ADR) estimated to be associated with a 3% decreased risk of interval CRC. AI systems can be broadly divided into CADe (for detection) and CADx (for diagnosis, or prediction of polyp histology in the context of colonoscopy). CADe has been extensively studied, with several randomized controlled trials and meta-analysis showing a higher ADR when CADe is used compared to the control groups without CADe. Besides the ADR, predicted polyp histology is a key component in the performance of colonoscopy as this enables the Clinician to make a decision regarding its management, as described above. In this regards, image-enhanced endoscopy (IEE) is often used to help Clinicians determine if colorectal polyps found on colonoscopy are neoplastic or hyperplastic. The most commonly used non-magnification classification is the NBI International Colorectal Endoscopic (NICE), while the Japan NBI Expert Team (JNET) classification is used where endoscopy systems with optical magnification and the proper training is available. However, these classification systems have varying diagnostic accuracy and interobserver agreement. Previous prospective studies looking at CADx have utilized endocytoscopy and autofluorescence imaging (CAD-AFI) with positive results. However, the major limitation in these CADx studies is that these imaging systems are costly and are not readily available in most centres worldwide. Furthermore, most Clinicians performing colonoscopies have not been trained in these modalities of imaging and will have to rely completely on the CADx function to detect polyps if these imaging modalities are used, without being able to fall back on their experience and training should there be doubts about the accuracy of a CADx diagnosis in a real-world setting. The Fujifilm 7000 System (Fujifilm Corp., Tokyo) has been in routine clinical use in all tertiary institutions in Singapore. The CAD EYE system was developed by Fujifilm Corp to aid Clinicians in colonoscopy with CADe and CADx functions. The basic functions and handling of the colonoscope, as well as the endoscopy processing unit, are similar to what is currently available in clinical practice, with the added CAD EYE software. The controller has been configured to allow the operator to activate and deactivate the CAD function depending on the need for it. These functions can be turned on and off using a button on the controller by the Clinician. The CADe and CADx functions operate when white light and blue laser imaging (BLI) are used, respectively. This provides a unique opportunity to externally validate the use of the CADx support tool by evaluating its diagnostic accuracy with final polyp histology as the gold standard, while also comparing its performance in a clinical setting against a Clinician using IEE (which is the conventional method of predicting polyp histology in colonoscopy). The investigators plan to conduct a prospective, multi-centre clinical trial to validate the accuracy of CADx support for prediction of polyp histology in real-time colonoscopy.
Study Type
OBSERVATIONAL
Enrollment
450
The CADx support tool operates when the Clinician switches the preconfigured CAD EYE function on using a button on the controller while the scope system is in BLI mode. This is performed after the Clinician first makes an optical prediction of polyp histology using IEE as described. The CADx support tool will make a prediction of polyp histology as "hyperplastic" or "neoplastic".
Changi General Hospital, National University Hospital, Singapore General Hospital and Tan Tock Seng Hospital
Singapore, Singapore
To evaluate the diagnostic performance of the CADx support tool compared to optical prediction of polyp histology by the Clinician in real-time colonoscopy in a clinical setting
Polyp histology used as gold standard
Time frame: 1 year
To determine the diagnostic performance of CADx versus optical prediction of polyp histology by endoscopist in the subgroup analysis
Subgroups include bowel preparation, size of polyp and location
Time frame: 1 year
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.