Endoscopy | Shaping the limitless future of endoscopy for the next decade

Anticipating a decade of growth, improvement, and transformation, endoscopes are shaping advancements in technologies and practices.

Endoscopy has a history of over 200 years. It began with the invention by a German physician named Phillip Bozzini in 1806, with a device called the Lichtleiter, and further pursued by Antonin Jean Desormeaux, a French physician, called the father of endoscopy. Desormeaux was the first to use the term endoscopy.

From the first simple tube endoscope to the modern world’s latest technology, the endoscope has helped a lot in the medical profession in diagnosing and treating a wide range of conditions.

Endoscopic retrograde cholangio pancreaticography (ERCP) is now used for diagnoses and treatment of diseases of the bile and pancreatic duct. Chromoendoscopy, in which a special dye is used, helps in better visualization. Endoscopic mucosal resection is used to remove cancerous tissue in the digestive tract. Endoscopic ultrasound, narrow-band imaging, uses a filter that differentiates mucosa from the vessel in the digestive tract.

Today, the increasing cases of chronic constipation, pancreatitis, inflammatory bowel diseases, stomach ulcers, and other serious health conditions have enhanced the use of endoscopy devices. The latest developments in endoscopy devices and techniques are shaping the global endoscopy devices market and the contribution of leading players in the industry.

Advancements in the technical trends brings forth significant and tangible benefits, often resulting in life-changing or life-saving events. Endoscopy, in particular, finds itself amidst a transformative phase, undergoing a tectonic transition, leading to substantial improvements in existing instrumentations and rendering traditional methods obsolete.

Colon capsule endoscopy (CCE) is increasingly considered as a viable option for colon imaging, supported by recent advancements and evidence. Despite initial reservations, meta-analyses have shown CCE to be on par or superior to colonoscopy in terms of precision, complication rates, and patient preference. However, a notable challenge lies in the high reinvestigation rate post-CCE. Efforts to reduce this rate include automated image analysis and the integration of artificial intelligence (AI), which could enhance quality and decrease costs, making CCE more competitive.

Third-space endoscopy (TSE) or submucosal endoscopy, a group of advanced endoscopic techniques and procedures involve accessing the peritoneal cavity and mediastinum safely by securing a mucosal flap to explore the submucosa and beyond. This concept gained traction with the success of peroral endoscopic myotomy (POEM) in treating achalasia. The field has witnessed exponential growth, driven by the development of new devices enhancing efficacy and safety.

For endoscopic closure of mucosal defects in TSE, novel devices include a through-the-scope suturing system (X-Tack Endoscopic HeliX Tacking System) and clips like the dual-action tissue clip (DAT) and the anchor-pronged TTS clip (Mantis). The X-Tack system uses preloaded helical coil tacks, providing an alternative to conventional clips. The DAT clip’s unique design allows independent arm action for secure tissue grasping, facilitating closure of large defects. The Mantis clip features angulated prongs for anchoring, reducing the risk of slippage during closure.

While these advancements in TSE, devices have shown promise in various clinical scenarios, further well-designed trials are necessary to establish their role in routine clinical practice.

Currently, TSE is widely used to resect subepithelial tumors and to manage refractory gastroparesis and Zenker’s diverticulum. The potential of TSE to provide easy and safe access to the mediastinum and peritoneal spaces may open doors to novel indications and rejuvenate the interest of endoscopists in natural orifice transluminal endoscopic surgery in the future.

A computer program software can now automatically calculate the adenoma detection rate (ADR) during colonoscopies. This metric evaluates how effective a colonoscopy is at detecting potentially cancerous growths called adenomas. Traditionally, calculating the ADR was time-consuming as it involved combining information from both the colonoscopy procedure and the pathology reports of tissue samples taken during the procedure. The new software can do this automatically, saving time and making it easier to monitor the quality of colonoscopies. This software highlights the potential for further development to incorporate additional metrics, ultimately enhancing performance benchmarks in colonoscopy practice.

AI in endoscopy
The latest progression in endoscopic imaging with rapidly evolving AI with recent studies provide real time diagnoses. It offers a ground-breaking approach for evaluating disease activity in ulcerative colitis (UC). Traditional scoring systems, such as the mayo endoscopic score (MES) and ulcerative colitis endoscopic index of severity (UCEIS), exhibit limitations due to inter- and intra-observer variability. AI, particularly machine learning (ML) and deep learning (DL), address these challenges by automating the assessment process, offering increased consistency and objectivity.

AI, applied to still images from endoscopic examinations, can achieve success rates comparable to human experts. Various models, including convolutional neural networks (CNN), have shown high discrimination capabilities for distinguishing between active disease and remission. Real-time endoscopy presents new challenges, such as patient movements and bowel-related factors, but AI models trained on high-definition videos have demonstrated significant agreement with expert assessments.

In addition to assessing endoscopic severity, AI extends its impact to histological prediction in UC. AI algorithms trained on colonoscopy images successfully predicted mucosal healing and histological remission with high accuracy, correlating with clinical outcomes. These advancements in AI technology, from real-time analysis to histological assessment, signify a transformative era in endoscopy, offering a more objective, consistent, and efficient approach to evaluating UC activity.

AI is a breakthrough in digestive endoscopy too. Screening gastric and colonic cancer detection should be improved especially outside of expert’s centers. Many software programs based on AI are now being used in endoscopy to aid doctors in their work. Some of these systems can even help with treatment by assisting in removing lesions completely during endoscopic procedures or predicting potential complications after treatment.

The next big step in this field is ensuring high-quality endoscopic procedures through quality assurance measures. This involves closely monitoring every aspect of a colonoscopy to make sure it meets the highest standards.

Overall, while AI plays a significant role by providing assistance to doctors, the final decisions and outcomes still rely on the expertise of the physician. However, there’s one unique example, video capsule endoscopy where the computer takes the lead, capturing multiple images along the way and even helping in making accurate diagnoses based on these images.

AI for cystoscopy, is likely to become an important tool and is expected to have real-world clinical applications comprehensively linking AI and imaging, data management systems, and clinicians.

It has shown promising potential in improving outcomes for bladder cancer (BCa) patients by assisting in various aspects of diagnosis and treatment. AI applications, particularly in computer-assisted detection (CADe) and computer-assisted diagnosis (CADx), have been evaluated extensively in BCa diagnostic work-up, aiming to enhance oncologic outcomes, quality of life (QoL), and reduce financial burdens associated with BCa treatment.

In the context of cystoscopy, which is a key procedure for BCa diagnosis, AI methods have been explored to improve the detection of bladder tumors and suspicious areas. By analyzing cystoscopy images, AI algorithms can help mitigate human errors in image interpretation, potentially leading to more accurate diagnoses. However, most studies have been retrospective and focused on static images, limiting their clinical applicability. Real-time AI-based tools applied to endoscopic videos could address this limitation, although prospective studies on their real-time performance are lacking.

Furthermore, AI has been utilized in urine sample analysis, particularly in urine cytology, to enhance diagnostic accuracy, especially for low-grade BCa cases. AI-based tools can automatically identify atypical and BCa cells, improving sensitivity for BCa detection.

However, research in this area is still evolving, and further validation is needed before widespread clinical implementation.

Outlook
The future of endoscopy devices with AI is promising and transformative. With ongoing advancements in AI algorithms and endoscopic techniques, we can expect to see more innovative applications that further optimize endoscopic procedures, leading to better patient care and outcomes in the future.