Healthcare of the next decade -Discovering Small while Optimizing at Scale

Significant efforts are made to put new technologies into use in healthcare. Studies such as those showing the benefit of “AI” for breast cancer screening (1) or detection of lung nodules on CT scans (2) are indeed promising and could lead to fast adoption by clinicians and healthcare systems. The best example may be the automated screening tool for diabetic neuropathy called IDx-DR, (3) an FDA-approved system used on actual patients in this very moment. Because the technology looks so promising, I am sometimes under the impression that we often try to make giant leaps and talk about “replacing humans with machines” for example, instead of finding ways to incorporate these new tools into the existing system, for the benefit of every person in the chain — from patients to hospital systems.

However, we are witnessing extraordinary discoveries through the use of these powerful tools. For example, a team of researchers of Google showed that a neural network that looked at the image of retinal fundus could appropriately assign the gender of the person in 97% of cases. (4) The implication is that there is a gender-related difference in the anatomy of the retinal fundus — something that has never been considered before in medical literature. More importantly, it was never possible to evaluate such hypotheses, and we owe this to the technological revolution that is being played out in front of our eyes.

Then, we have deep learning models can predict in-hospital mortality based on electronic health records (EHRs) only with a staggering 93% accuracy. (5) In the ocean of data, this system is able to find common threads that lead it to a conclusion — a hidden discovery in the data that we are now capable of interpreting.

And finally, many “scientific dogmas” have been debunked just in the last year, and are kindly summarized by Eric Topol in one of his many inspiring tweets, (6) the most fascinating one for me is the rare, but possible paternal mitochondrial DNA transmission.7

All of these revelations have sparked a notion that the biggest breakthroughs and discoveries in healthcare will not come from the development of extraordinary tools that aim to completely disrupt the industry. Instead, I believe it will come from creative, but humble individuals trying to optimize current systems — collaborating with each other, but also with machines and these new tools that have been given to us.

I believe that progress will be based on the ability to collect, store, and interpret data while maximally relieving patients and clinicians of that painful interaction that is going on now through a failed EHR experiment. Sadly for many, there is nothing cool or fancy about this approach — which is why success can come only if we are guided by empathy and true desire to help patients.

By optimizing current practices (going all-digital, facilitating remote care, use of gadgets for tracking health parameters), we would substantially improve the lives of many patients. And in doing that, we would free clinicians and nurses to become creative and savvy thinkers.

We should be excited that we are now in possession of tools that can completely change the way we look at healthcare. But we must not rush to fix everything with one stroke — slow and steady wins the race.

References:

1. Wu N, Phang J, Park J, et al. Deep Neural Networks Improve Radiologists’ Performance in Breast Cancer Screening. IEEE Trans Med Imaging. October 2019. doi:10.1109/TMI.2019.2945514

2. Ali I, Hart GR, Gunabushanam G, et al. Lung Nodule Detection via Deep Reinforcement Learning. Front Oncol. 2018;8:108. doi:10.3389/fonc.2018.00108

3. van der Heijden AA, Abramoff MD, Verbraak F, van Hecke M V, Liem A, Nijpels G. Validation of automated screening for referable diabetic retinopathy with the IDx-DR device in the Hoorn Diabetes Care System. Acta Ophthalmol. 2018;96(1):63–68. doi:10.1111/aos.13613

4. Poplin R, Varadarajan A V, Blumer K, et al. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nat Biomed Eng. 2018;2(3):158–164. doi:10.1038/s41551–018–0195–0

5. Rajkomar A, Oren E, Chen K, et al. Scalable and accurate deep learning with electronic health records. NPJ Digit Med. 2018;1:18. doi:10.1038/s41746–018–0029–1

6. https://twitter.com/EricTopol/status/1205952384913903616.

7. Luo S, Valencia CA, Zhang J, et al. Biparental Inheritance of Mitochondrial DNA in Humans. Proc Natl Acad Sci U S A. 2018;115(51):13039–13044. doi:10.1073/pnas.1810946115

--

--

--

Clinician at Huma, PhD student of Public Health focusing on hospital-acquired infections in ICUs. Building medical devices.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

why MSG is ok

Is This Really a Book for Nurses?

Family | Medical | History

I am Addicted; Part 01

‘Long Covid’ may affect multiple parts of body and mind, doctors say

Tinnitus — Literally Suffering in Silence

The Beginner’s Guide to Dry Eyes

The Future of Healthcare: “We must change the fact that right now, the healthcare system is not set

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Alex Despotovic

Alex Despotovic

Clinician at Huma, PhD student of Public Health focusing on hospital-acquired infections in ICUs. Building medical devices.

More from Medium

Changing the World — One Mindset at a Time…

Dealing with LSAT Burnout

Elegran Insights: Weekly Manhattan & Brooklyn Market

The Solution to Wealth Inequality