MISTac's 8mm vision-based tactile sensor gives surgeons a sense of touch
An 8mm sensor that detects forces as tiny as 24.3 mN could restore tactile feedback in robot-assisted surgery.
Minimally invasive surgery (MIS) offers patients faster recovery but forces surgeons to work without the sense of touch. To fix that, an international team led by Robin Koch, Annabella Mascot, and Roberto Calandra — including Stanford's Mark Cutkosky — developed MISTac, a vision-based tactile sensor specifically designed for palpation during MIS. At just 8mm in diameter, the sensor fits through the trocars used in minimally invasive procedures, while its modular 3D-printed case lets surgeons use bulky, off-the-shelf illumination and imaging hardware that can be easily swapped or upgraded.
MISTac packs serious specs: 176.68 µm optical resolution, 250 µm tactile resolution, and the ability to resolve forces as small as 24.3 mN. The team validated it in an in vivo study, collecting real tactile data and training a machine learning model for tissue classification that hit ~84% aggregate accuracy under leave-one-out cross-validation. That's a meaningful step toward giving surgeons intraoperative tumor localization and tissue typing. The researchers have open-sourced MISTac, making the full sensor design and code publicly available for other labs to build on.
- MISTac is an 8mm vision-based tactile sensor that fits through standard MIS trocars, with a modular 3D-printed case for easy upgrades.
- It achieves 176.68 µm optical resolution, 250 µm tactile resolution, and detects forces as low as 24.3 mN.
- In an in vivo trial, a machine learning model using MISTac data classified tissue with ~84% accuracy; the sensor is fully open-source.
Why It Matters
Giving surgeons tactile feedback in minimally invasive procedures could improve tumor detection and tissue classification, directly impacting surgical outcomes.