In recent years, steerable needles have attracted significant interest in relation to minimally invasive surgery (MIS). Specifically, the flexible, programmable bevel-tip needle (PBN) concept was successfully demonstrated in vivo in an evaluation of the feasibility of convection-enhanced delivery (CED) for chemotherapeutics within the ovine model with a 2.5 mm PBN prototype. However, further size reductions are necessary for other diagnostic and therapeutic procedures and drug delivery operations involving deep-seated tissue structures. Since PBNs have a complex cross-section geometry, standard production methods, such as extrusion, fail, as the outer diameter is reduced further. This paper presents our first attempt to demonstrate a new manufacturing method for PBNs that employs thermal drawing technology. Experimental characterisation tests were performed for the 2.5 mm PBN and the new 1.3 mm thermally drawn (TD) PBN prototype described here. The results show that thermal drawing presents a significant advantage in miniaturising complex needle structures. However, the steering behaviour was affected due to the choice of material in this first attempt, a limitation which will be addressed in future work.

Needle bending is a significant cause of error in biopsies, leading to lesion missampling and consequent cancer misdiagnosis. This paper presents the design of a new mechanism that detects the needle bending as soon as it occurs and immediately reduces the insertion force. Importantly, this is achieved without employing external sensors or electromechanical actuators. Experiments on a silicone-rubber phantom indicate that the proposed device can help to avoid deep insertions with bent needles, thus potentially reducing the associated risks and improving patient safety in biopsies and percutaneous interventions.

High‐aspect‐ratio polymer materials are widely utilized in applications ranging from everyday materials to specialized equipment in industrial and medical fields. Traditional fabrication methods, such as extrusion and molding, face challenges in integrating diverse materials and achieving complex geometries. Thermal drawing transforms macroscale preforms into long, thin fibers while preserving cross‐sectional geometry, but preform fabrication remains a key limitation for producing complex and functional fiber architectures. Here, commercially available 3D printing is used as a fast, low‐cost route for fabricating polymer preforms for thermal drawing. By optimizing printing parameters, fibers with diameters down to 100 µm are produced while preserving complex cross‐sectional features at the micron scale. This capability enables miniaturization of interlocking fiber geometries designed for steerable catheter concepts. Beyond amorphous thermoplastics, semi‐crystalline polypropylene preforms are drawn to produce fibers that exhibit a fivefold greater tendon‐driven bending response than comparable polycarbonate fibers. Functional material integration is further demonstrated using magnetically loaded filament to produce magnetically responsive fibers for robotic applications. Longitudinal periodicity engineered into the preform produces tapered fibers with programmable stiffness gradients in a single fabrication run. This approach supports rapid translation from preform design to thermally drawn polymer fibers with complex geometries, broader material selection, and application‐specific functionality.

Soft optical sensors hold potential for enhancing minimally invasive procedures like colonoscopy, yet their complex, multi-modal responses pose significant challenges. This work introduces a machine learning (ML) framework for real-time estimation of 3D shape and contact force in a soft robotic sleeve for colonoscopy. To overcome limitations of manual calibration and collect large datasets for ML, we developed an automated platform for collecting data across a range of orientations, curvatures, and contact forces. A cascaded ML architecture was implemented for sequential estimation of contact force and 3D shape, enabling an accuracy with errors of 4.7% for curvature, 2.37% for orientation, and 5.5% for force tracking. We also explored the potential of ML for contact localization by training a model to estimate contact intensity and location across 16 indenters distributed along the sleeve. The force intensity was estimated with an error ranging from 0.06 N to 0.31 N throughout the indenters. Despite the proximity of the contact points, the system achieved high localization performances, with 8 indenters reaching over 80% accuracy, demonstrating promising spatial resolution.

To navigate medical instruments safely and accurately inside a patient’s vascular tree, combining X‐ray fluoroscopy with intermittent contrast injections is the gold standard. However, prolonged exposure to ionizing radiation poses health risks and necessitates the use of cumbersome lead vests for the clinicians, and contrast injections can lead to acute kidney injury in patients. Bioelectric Navigation, a non‐fluoroscopic tracking modality, aims to provide an alternative. It uses weak electric currents to detect local anatomical features in the vasculature and localize instruments without X‐ray imaging. In this work, we advance Bioelectric Navigation on two frontiers. Firstly, we introduce a new class of bespokely designed electrode catheters. They are fabricated using 3D printing, thermal drawing, and laser micromachining. Specifically, we manufacture a 6 Fr catheter incorporating 16 electrodes, a guidewire channel and an additional sensor compartment. We thoroughly assess the catheter’s mechanical and electrical properties. Secondly, we introduce an algorithm to localize the catheter along the centerline of a vascular phantom, for the first time fusing electric detection of vascular geometry with electric distance estimation. We report both tracking accuracy and usability evaluated by an expert endovascular surgeon, demonstrating the strong potential of this technology for integration into the existing clinical workflow.