Call to Arms with Big Data

0
198

The tools for the next major disruption in neurosurgery are already here, we just have to leverage them. Large data sets from our daily lives will allow insights into health that, when merged on national and international levels, will dramatically change our treatment of neurological disease.

Big Data

Most people already have the tools for gathering precise, in-depth information about their life and health: today’s cell phones, smart watches and home controllers contain powerful computers with a broad array of sensors. These devices can sense heartbeats, monitor respiration, quantify mobility, analyze gait, fatigue and interpret speech.1-3 For the surgeon, this data has the potential to dramatically change patient selection and treatment planning. Two poignant examples are:

o
  • GPS analysis over weeks or months and accelerometer data from a cell phone can yield in-depth information on gait, flexibility, walking speed, ambulatory range and social independence, all of which influence considerations in spine surgery.
  • Monitoring patients with a history of low-grade glioma via a home digital assistant that can continuously record speech, including the examination of language patterns, decision-making and memory, over several months could alert providers to subtle cognitive changes from tumor recurrence.

Ultra Big Data

However, the greater power from harnessing personal data will be seen when huge personal data sets are combined with hospital data, surgical treatments and outcomes. Artificial intelligence algorithms can find subtle connections between habits, diseases, treatments and outcomes. In contrast to clinical trials testing one or two variables in a highly select population, these analyses could concisely elucidate the contributing effects of hundreds of variables in a truly representative population. Gone would be bias based on:

  • Self selection
  • Socioeconomic factors
  • Diversity
  • Practitioner preconception
  • Study design

This approach would be a boon for neurosurgery, where the infrequency of diseases and variability of treatments have made randomized clinical trials and treatment consensus difficult. Imagine analyzing in-depth lifestyle, treatment and outcome data on hundreds of thousands of patients with spontaneous intracranial hemorrhage. Such a data set would reveal which patients are most likely to benefit from surgery, which surgery does best for which type of lesion and what co-morbid conditions will affect recovery. Algorithms will give accurate expected functional recovery, and discussion on surgical risks and benefits will involve individually precise statistics. Therefore, surgical decision-making will dramatically change, as surgeons will simply input salient patient characteristics and a computer program will provide advice on best treatment, best approach and expected outcomes.

Hurdles to Utility

To reach this point, we still have a great number of hurdles to overcome.

  • Data cohesion across the country will require standardization, and compiling and sharing the data will require a dedicated effort on all levels.
  • Public acceptance will be a barrier. With recent breaches in consumer and patient privacy, we will have the difficult task of convincing the public that we can use this wealth of data in a safe and unbiased manner.
  • Surgeons will also be highly resistant to these changes, as it will mean losing much of our autonomy in decision-making as well as publicly opening ourselves up to what these deep analyses may reveal about our practices, biases and outcomes.

Regardless of whether we want this disruption, changes are coming. Insurance companies are already harvesting highly personal data to assess risk and grade premiums. Post-hoc meta-analyses leverage many small data sets to detect subtle findings not seen in larger trials and surgeon statistics are being factored into reimbursement. Like all disruptive technologies, the benefits will be balanced by the potential for misuse. Right now, we must be proactive and involved. Otherwise, we will lose the opportunity to guide how this technology is used and direct how these changes will affect the future of neurosurgery.

 

References

1. Haescher, M., Matthies, D. J., Trimpop, J., & Urban, B. (2015). A study on measuring heart- and respiration-rate via wrist-worn accelerometer-based seismocardiography (SCG) in comparison to commonly applied technologies. Proceedings of the 2nd International Workshop on Sensor-based Activity Recognition and Interaction – WOAR 15.

2. Iso, T., & Yamazaki, K. (2006). Gait analyzer based on a cell phone with a single three-axis accelerometer. Proceedings of the 8th Conference on Human-computer Interaction with Mobile Devices and Services – MobileHCI 06.

3. Phan, D., Siong, L. Y., Pathirana, P. N., & Seneviratne, A. (2015). Smartwatch: Performance evaluation for long-term heart rate monitoring. 2015 International Symposium on Bioelectronics and Bioinformatics (ISBB).

[aans_authors]

Print Friendly, PDF & Email
o