Our team at Lemici consists of Mark Truskinovsky and Warish Zaman Orko. Mark completed his B.S. in Biomedical Engineering from the Rochester Institute of Technology. Warish graduated from the University of Rochester with his B.S. in Biomedical Engineering. We bring a shared passion for medical device innovation and a drive to improve quality of care in medicine.
The Lemici patient camera seeks to provide increased access to teleconsultations for patients with motor, cognitive, and/or communication difficulties, and full view control for clinicians. The system consists of a remotely operable camera able to move around the patient as required by their physician to take images of wound sites and remotely assess their healing. The camera maneuvers to preset positions around the patient of clinical interest to allow the clinician to look for signs of inflammation such as redness, fluid leakage, and scab formation.
We first embarked on this project after a discussion with Dr. Thomas Mattingly at the University of Rochester Medical Center during the team’s clinical rotations. We gained an appreciation of the new challenges facing clinicians around the world in adapting to a steep increase in the use of telemedicine, with practice far outstripping the pace of technological development.
In the status quo, clinicians experience difficulty in visualizing wounds located in areas away from the front of the patient. Current methods include the patient moving a smartphone around themselves, propping the phone up and moving in front of the camera, using a mirror to take images, and asking another person to take the picture. Smartphone methods are limited in image quality, ability to zoom in on the region of interest, muscle tremors, and ability to point the camera at the wound properly. Mirrors degrade image quality and suffer from glare as well. The last method is wholly inaccessible to the 37 million adults in the U.S. living alone . Relying upon the patient to move themselves introduces issues such as communication and patient mobility constraints.
Our latest prototype explores the articulated arm form factor. The device consists of a GoPro Hero camera mounted to the Tinkerkit Braccio arm, controlled by an Arduino microcontroller over the internet. The arm has preset positions to image the left, rear, and right sides of the patient without having to manipulate individual motors to achieve a desired image each time. Through a platform-agnostic web interface, the user can easily adjust the view as needed to make an assessment of the region of interest. We chose the Braccio platform as an easy-to-modify, well-documented robotics toolset which dovetailed well with the Arduino skills we brought to the table.
We created a test protocol with two main goals: inform device development by answering key questions, and to verify that the current prototype is able to reliably deliver an image from the desired position.
The rotational accuracy test was designed to evaluate the faithfulness of the motors to an expected angle of rotation, which is a key component that provides assurance of the repeatability of delivering a certain view of a wound site.
The rotational drift test was designed to identify if there is a gradual drift in the motors throughout sequential positioning events, which in tandem with the rotational accuracy test can help to confirm that the arm can achieve the same position repeatedly without a notable difference in the output view.
We made a test fixture and wrote test software to speed up the process.
The testing process generated evidence towards minimizing the number of motors required, for example, which helps to drive costs down and simplify the device to its minimal state. We hypothesize that the failure in motor 4 (furthest from the base) was a result of compounding slack in each link and its adjacent joints. However, the bulk of the positioning of the camera is done by the two joints closest to the base, motors 1 and 2 (both of which passed), which helps to lower the concern regarding the performance of motor 4.
This is great! This is particularly good for people who are not so mobile.Dr. Thomas Mattingly
We obtained feedback from five clinicians at Strong Memorial Hospital across a range of specialties, from neurosurgery to pediatrics to otology. Four out of five clinicians agreed that the setup was sufficient to make a clinical assessment when presented with the device. All of them responded positively when asked if they had patients who would benefit from the use of our device. We are encouraged by the feedback received at this stage of our project, as it strengthens the case for significant improvement in patient quality of care through adoption of our device.
We initially searched for predicate devices that would allow us to classify our device in the lowest risk-based classification, i.e. a Class I device exempt from filing premarket notification. After thorough predicate device comparison, we selected the SILHOUETTE Model 1000.01 as a predicate, given that it is also a camera for inspecting patient wounds. This would place our device in the FXN product code, reducing the time and costs of our commercialization pathway.
We then made use of the FDA Q-Submission program to obtain feedback from a reviewer about our regulatory determination, and are working to incorporate the feedback obtained into our regulatory strategy. This endeavor has given the team invaluable experience in preparing a regulatory submission, analyzing predicate devices, forming a regulatory strategy, and honed important skills required in communicating and working with regulatory bodies.
Our team focused on implementing a robust quality system from Day 1. Strict control over design changes, full traceability of aspects of the design, and documentation for every step from clinical observation to testing were integral to the team’s mission. Relevant standards were identified and adhered to, summarized in the list below.
- ISO 13485 Quality Management Systems
- ISO 14971 Application of Risk Management to Medical Devices
- ISO 10993 Biological Evaluation of Medical Devices
- ANSI/AAMI HE75 Human Factors Engineering of Medical Devices
Stemming from the conclusions of our tests as well as feedback from clinicians, there are several potential directions we foresee that the device could be taken in to improve it and increase its utility:
- Redesign of the robotic arm to remove unused joints and motors, driving down cost and simplifying operation.
- Transition to a more industrial grade construction to eliminate errors and/or slack from articulations, increasing the device’s accuracy in positioning.
- Implement a lower cost camera that is capable of a relatively large magnitude of digital zoom to help provide a closer-up image of a wound to clinicians.
- Improve the web interface through which the arm is operated by making it more user friendly and potentially implementing additional intuitive controls for both the arm and the camera.
- Implement cybersecurity measures using the FDA’s “Cybersecurity in Medical Devices: Quality System Considerations and Content of Premarket Submissions” draft guidance document to conform to 21 CFR Part 820’s Quality System Regulation (QSR) standards.
- Develop ancillary functions that serve to help measure other health metrics of the patient, such as vitals, to provide more clinically relevant information through a unified platform.
- Streamline the device’s compatibility with and assimilation into the workflow of commonly used video conferencing platforms.
- Develop software algorithms that help the camera track around a specified point of focus (i.e. the wound of interest).
- Dr. Greg Gdowski, CMTI Executive Director
- Martin Gira, CMTI Senior Research Engineer
- Dr. Amy Lerner, CMTI Academic Director
- Dr. Thomas Mattingly, URMC Neurological Surgery
- Dr. Jonathan Stone, URMC Neurological Surgery
- Dr. Benjamin Crane, URMC Otolaryngology
- Dr. Adam Kelly, URMC Neurology
- Dr. Adam Dziorny, URMC Pediatrics