Data Science

The design and fabrication of all 10 Newtons' technologies combine advanced
engineering and wearables that allow us to assess technical skills, operative skills,
real-time decision making, and overall competency in a variety of surgical and functional activities.
10 Newtons has created and adopted numerous types of data acquisition methods,
including video, audio, motion tracking sensors, pressure sensors, and tool-tip interaction
software to capture hands-on clinical performance at every point of contact.
Data is the driving passion behind 10 Newtons.
Our vast experience and large-scale data acquisition expertise have allowed
us to create validated models of mastery performance, objective measures of
efficacy, as well as high-leveldiscernment of hands-on clinical skills.

Description of Sensor-Based Technology:

With the advent of simulation technology, 10 Newtons has adopted several types of data acquisition methods, including motion sensors, pressure sensors and tool-tip interaction software, to capture hands-on clinical performance. 10 Newtons’ primary objective with the simulation technology component is to enhance clinical performance and thus, patient outcomes, by creating responsive and replicable clinical simulation environments that provide performance feedback linked to a rich database of expert performance and learning curve metrics.

Our sensor-based technology and vast experience enables 10 Newtons to:

  • Provide data for monitoring medical and surgical performance through the assessment of technical skill, operative skill, decision making and overall competency in a variety of diagnostic and therapeutic activities
  • Enhance clinical and functional skills by providing instantaneous, quantitative and personalized feedback and by applying touch metrics to the evaluation of hands-on skills
  • Create seamless surgical and clinical solutions with the integration of haptic algorithms to enhance clinical and operative decision-making

We Value Quality Assurance & Quality Control:

Junior data collection staff are trained by senior staff in simulation set up and tear down, stakeholder interactions and data back-up procedures. All survey entries and code book generation procedures are verified by an independent staff member. A random numbers generator is used to pick a 10% sample of the data set for re-review by a different staff member. This serves as the verification process for all of our data collection procedures.

Data Processing and Editing:

All of our data sets are reviewed for outliers and unreasonable values. For each data set, this information is reported upon publication of our results in peer reviewed journals. On average, we experience approximately 2-5% data loss either from incomplete surveys, or technology malfunction. We are now able to keep data losses at a minimum due to our strict, routine pilot exercises. Over the years, we have learned that most of our data losses are due to human error. Hence, we have significantly improved our staff training exercises over the years. All of our data sets have associated code books with legacy coding schema which allows us to pool data sets when necessary.

Analytic Notes:

Our data is all backed up into one system. Both SAS and SPSS have been used seamlessly to run analysis on our data. We have participated in numerous data sharing contracts and relationships with various academic groups who have an interest in collaborative data analysis.