In the robotic mode under Scenario-A, the subjects preferred providing unidirectional step inputs (for example ‘left’ and ‘up’) to traverse diagonally in track #1 or along the curvature in track #3. This resulted in similar timing to that of manual mode. However, the percentage of time for which the scope’s focus shifts outside the track and the number of missed segments were significantly lower for robotic mode for all the three tracks. This validated the use of robotic mode to precisely navigate through tracks of 2 mm in width.

In the manual mode under Scenario-B, the percentage of time for which the scope’s focus shifts outside the track and the number of missed segments were higher due to (a) misalignment of the scope focus caused by abrupt manual insertion/retraction motion, (b) difficulty to hold the deflection knob stable at the intended position, (c) disorientation of the endoscope distal tip inside the stomach cavity. These factors collectively led to a prolonged effort to reorient the scope’s focus back onto the track after deflection, consequently resulting in an increase in the task duration. These challenges were addressed by the robotic mode. In this mode: (a) The operator was relieved from the burden of supporting the endoscope’s weight, (b) the adaptable speed control facilitated quicker movements on straight sections of the track while offering better control and slower motion on curved sections, and (c) the scope adapter maintained endoscope stability, preventing sudden unintended movements and focus misalignments. The NASA-TLX analysis also indicated that the subjects favored the robotic mode as compared to the manual mode providing a user-friendly control mechanism. Significant differences were observed for mental demand, physical demand, effort, performance, and frustration. For temporal demand, the subjects did not feel the time pressure for completing the tasks under two modes, and this resulted in similar scores.

This work demonstrates the advantage of extending the usage of scope adapter (originally built for a rigid zero, angulated [9, 11], and articulated scopes [10, 14]) for flexible endoscopes. The generic design enables utilizing the proposed system across different endoscopic procedures (such as bronchoscopy, colonoscopy, or duodenoscopy). For example, a new support plate could be designed to host a flexible bronchoscope (Fig. 6a). Though a new design of support plate (to actuate deflection knob) is required, the same design of the scope adapter (Fig. 2c) can be used to achieve insertion/retraction and the rotation of the bronchoscope along its shaft. In several endoscopes, such as colonoscope (Fig. 6b) and duodenoscope (Fig. 6c), the support plate designed for gastroscope (Fig. 2b) can be reutilized. This further underscores the value of using a modular system where the components can be configured to facilitate multiple types of usage. The proposed support plate and scope adapter design do not impede access to the buttons on the endoscope’s control section (Fig. 2d). The overall system can be connected to conventionally used air/CO2 insufflators, such as UCR (Olympus Medical Systems, Tokyo, Japan) or CO2MPACT Endoscopic Insufflator System (Steris, Mentor, U.S.A), and irrigation pumps. Alternatively, smart pressure control surgical insufflators such as EVA-15 (Palliare, Galway, Ireland) [15] and UHI-3 (Olympus Medical Systems, Tokyo, Japan) [16] can be integrated for a seamless clinical experience.

Support plates hosting a bronchoscope, b colonoscope, and c duodenoscope. The former support plate host mechanism to actuate up–down movement of the bronchoscope’s distal end. The latter support plate hosts a mechanism to enable both up–down and left–right movements of the distal ends of the bronchoscope and duodenoscope. Zoomed panels display the distal ends of each endoscope. The compartment inserted into the scope adapter (marked by a rectangular dotted line) is the same for all the support plates

Several enhancements can be made to the current design of the proposed system to improve its applicability. First, the current version uses a rail to prevent buckling and guide the flexible endoscope during insertion/retraction. This increases the workspace of the setup in the operating room. To reduce the footprint of the proposed system, a highly compressible origami-based anti-buckling support sheath can be deployed [17]. Second, a mechanism comprising of push button switches could be used to enable pressing of air/water irrigation and suction buttons. Third, while a game controller was used as an input device, other human computer interfaces [18, 19], such as stylus/handle [6, 20], interfaces based on eye-gaze and/or head motion [7], can be integrated with the proposed system to provide actuation commands. While game controller and stylus/handle engages operator’s hands, head/eye tracking devices offer a hands-free solution for endoscope movements [11]. In addition, instead of displaying the view acquired from the endoscope on a 2D display screen, head mounted display devices can be used to render the view in virtual reality or mixed reality environment [21]. These devices also have inbuilt sensors to detect head pose and can provide actuation commands for the endoscope movement. Lastly, visual cues can also be rendered to compensate for the loss of tactile feedback. These visual cues could assist in depicting the endoscope shape (e.g. inserted length and deflection angles) based on the motor state, and the forces exerted on the lumen using fiber Bragg grating sensors [22].

The proposed system was trialed in an environment using basic endoscope manipulation techniques. To further assess the system, we plan to evaluate it in high-fidelity phantoms, such as a colon phantom during cecal intubation, where advanced techniques (e.g., hooking, left turn shortening, and right turn shortening [12]) can be employed to navigate in the presence of luminal tissue deformation. The user study was primarily conducted to demonstrate the ease of using the actuated system for maneuvering. Experienced endoscopists were not included as subjects, and they may perform as well as the proposed system [23]. However, robotic actuation offers several advantages over manual manipulation, including (a) superior endoscope stability, (b) effective alleviation of hand tremors and wrist discomfort associated with prolonged endoscope manipulation, and (c) reduced operator fatigue by enabling procedures to be performed comfortably while seated. These ergonomic benefits may prevent musculoskeletal pain issues reported by expert endoscopists [24]. Evaluating the system in clinical use case scenarios will be necessary to confirm this. Additionally, advanced guidance solutions that combine sensing, intelligent algorithms, and motor control for actuation can be used to automate the navigation.

In conclusion, the proposed system enhances the maneuverability of flexible endoscopes. Moreover, the simplified design eases the installation and disassembly of the system for usage. The user study validated the effectiveness of the proposed scope adapter and laid a strong foundation for future development of modular and generic scope assistant systems.