Design and Evaluation of ScanCap: A Low-Cost, Reusable Tethered Capsule Endoscope with Blue-Green Illumination Imaging for Unsedated Screening and Early Detection of Barrett's Esophagus.
Cheima HicheriAhad M AzimuddinAlex KortumJoseph BaileyYubo TangRichard A SchwarzDaniel RosenShilpa JainNabil M MansourShawn GrothShaleen VasavadaAshwin RaoAdrianna MaligaLeslie GallegoJennifer CarnsSharmila AnandasabapathyRebecca R Richards-KortumPublished in: Bioengineering (Basel, Switzerland) (2024)
Esophageal carcinoma is the sixth-leading cause of cancer death worldwide. A precursor to esophageal adenocarcinoma (EAC) is Barrett's Esophagus (BE). Early-stage diagnosis and treatment of esophageal neoplasia (Barrett's with high-grade dysplasia/intramucosal cancer) increase the five-year survival rate from 10% to 98%. BE is a global challenge; however, current endoscopes for early BE detection are costly and require extensive infrastructure for patient examination and sedation. We describe the design and evaluation of the first prototype of ScanCap, a high-resolution optical endoscopy system with a reusable, low-cost tethered capsule, designed to provide high-definition, blue-green illumination imaging for the early detection of BE in unsedated patients. The tethered capsule (12.8 mm diameter, 35.5 mm length) contains a color camera and rotating mirror and is designed to be swallowed; images are collected as the capsule is retracted manually via the tether. The tether provides electrical power and illumination at wavelengths of 415 nm and 565 nm and transmits data from the camera to a tablet. The ScanCap prototype capsule was used to image the oral mucosa in normal volunteers and ex vivo esophageal resections; images were compared to those obtained using an Olympus CV-180 endoscope. Images of superficial capillaries in intact oral mucosa were clearly visible in ScanCap images. Diagnostically relevant features of BE, including irregular Z-lines, distorted mucosa, and dilated vasculature, were clearly visible in ScanCap images of ex vivo esophageal specimens.
Keyphrases
- low cost
- convolutional neural network
- deep learning
- high resolution
- high grade
- optical coherence tomography
- early stage
- papillary thyroid
- high speed
- end stage renal disease
- artificial intelligence
- chronic kidney disease
- squamous cell
- mass spectrometry
- ejection fraction
- photodynamic therapy
- intensive care unit
- low grade
- newly diagnosed
- machine learning
- light emitting
- squamous cell carcinoma
- big data
- electronic health record
- case report
- neoadjuvant chemotherapy
- optic nerve
- fine needle aspiration
- sensitive detection