AI for accessible, affordable and accurate healthcare

13

September

2020

4/5 (1)

AI for accessible, affordable and accurate healthcare

One of the most important inputs to health is healthcare. Unfortunately, a big problem in today’s society is unequal access to this basic human right. According to a report by the World Bank and the WHO at least half of the world’s population is too poor to access essential health care service [1]. Artificial intelligence (AI) could be a piece of the puzzle in solving this problem. AI makes use of complex algorithms and software to emulate human cognition in the analysis, interpretation and comprehension of complicated medical and healthcare data [2].  The primary aim of health-related AI applications is to analyse relationships between prevention or treatment techniques and patient outcomes [3]. But how can all this make an impact?

 

Accessibility

Access to basic and specialist healthcare varies vastly across countries and even cities. In order to bring healthcare closer to the patient AI can help by providing doctors with high-quality diagnostic images regardless of geographical location. This means that doctors are able to analyse and diagnose from anywhere in the world! This application is already making traction in the U.S. Start-up InfiniteMD provides second-opinion video consultations to patients around the world [4]. Building on that they are even developing an algorithm that aids in treatment decision-making and connects them with a global treatment operation or clinical trials [5].

 

Affordability

AI can significantly cut down the cost of healthcare by analysing medical and visitation data. Machine learning can find patterns in patient admission and discharges to determine which patient categories tend to overstay in hospitals and reduce patients stay accordingly. Additionally by powering virtual chat bots, AI can eliminate unnecessary in-person doctor visits and readmission, which can potentially save billions of dollars annually.

 

Accuracy

AI can be used to predict diseases in early stages by analysing data or imaging. This could aid doctors in diagnosing diseases and rule out false negatives, which is very valuable. The UK government is already working on this. They are building a new medical technology centre that will use AI to aid in disease diagnosis. The London Medical Imaging and Artificial Intelligence Centre for Value-Based Healthcare, will apply AI to detect anomalies in scans, helping with earlier detection. Intelligent information will set new precedents when it comes to diagnosis accuracy [6].

 

As with many new technologies, AI still has a lot to answer for. There are questions about privacy, safety, accuracy and ethics that we still need to examine. Yet, I do believe that there is a lot of potential and ground to be gained by combining machine systems with care delivered by humans.

 

References:

  1. http://pubdocs.worldbank.org/en/193371513169798347/2017-global-monitoring-report.pdf
  1. https://en.wikipedia.org/wiki/Artificial_intelligence_in_healthcare
  1. Coiera E (1997). Guide to medical informatics, the Internet and telemedicine. Chapman & Hall, Ltd.
  1. www.infinitemd.com
  1. https://www.bizjournals.com/boston/inno/stories/profiles/2018/01/29/infinitemd-raises-15m-for-ai-that-gives-patients-a.html
  1. https://www.kcl.ac.uk/bmeis/research-impact/london-medical-imaging-and-ai-centre-for-value-based-healthcare

Please rate this

Connecting Computers to Brains

9

September

2020

5/5 (1)

Last week Elon Musk surprised the world again with a live demo from one of his futuristic startups. This time it’s not about bringing people to Mars, building electric cars or creating an efficient transportation infrastructure. No, Neuralink is trying to connect computers to human brains by designing the first neural implant that allows humans to control a computer or mobile device with their brain.

At the event Elon demonstrated several pigs that had a prototype version of the neural links implanted in their head. On a screen the audience was able to see how the implant reads brain activity allowing for prediction of the movement and positioning of the joints with a high accuracy. Pretty cool right?

 

How does it work?

A neuron consist of three parts. There is the dendrite that receives a signal, a cell body called a soma which computes the signal and an axon which sends the signal out. In a human brain there are Human over 86 billion neurons sending and receiving information by communicating through electric signals. All these neurons together form a web that is continuously firing away resulting in about 15 million billion bits of information moving around per second. All these bits together carry information about everything we see, feel, touch or think.

Neuralink’s brain machine interface (BMI) system places micron-scale threads near these neurons allowing them to decode information that is communicated by these cells. These threads are so fine that they need to be inserted by a robot that ‘sews’ these threads exactly in the right areas of the brain with micrometre spatial precision. As the implant contains arrays of flexible electrode threads with up to 3072 electrodes per array, distributed across 96 it can gather a lot of detailed information. This direct connection allows for reading brain activity but information but also for the implant to write information back into the brain.

 

Why is this interesting?

Using this BMI technology Elon aims to improve the lives of a broad range of patients. The short-term goal of Neuralink is to help people with a paralysis interact with the world around them by allowing them to communicate using computers and mobile devices. In the long-term, as the technology improves, it could potentially treat a wide variety of neurological disorders in the fields of sensory and motor disabilities, neurocommunication, BMI control of exoskeletons and cognitive state evaluation.

Personally, I think that potential application of this technology is super interesting, especially for people that are currently paralysed. But, we also have to be careful on how far we want to take this in a world were advanced technology is mostly available to the wealthy and the implications of inserting computer hardware into human brains is still unknown. As George Orwell’s said it best:

“Nothing is your own except the few cubic centimetres inside your skull.”

Maybe we should keep it that way in a world with every increasing connectivity.

 

 

References:

http://thephenomenalexperience.com/content/how-fast-is-your-brain#:~:text=Each%20neurons%20fires%20(on%20average,other%20neurons%20get%20that%20information.

Neuralink Progress Update, Summer 2020: https://www.youtube.com/watch?v=DVvmgjBL74w

https://neuralink.com/

https://www.vox.com/recode/2020/8/28/21404802/elon-musk-neuralink-brain-machine-interface-research

Daly JJ, Wolpaw JR. Brain-computer interfaces in neurological rehabilitation. Lancet Neurol 2008 Nov;7(11):1032-1043.

Maksimenko VA, Hramov AE, Frolov NS, Lüttjohann A, Nedaivozov VO, Grubov VV, et al. Increasing Human Performance by Sharing Cognitive Load Using Brain-to-Brain Interface. Front Neurosci 2018 Dec 13;12:949

Kawase T, Sakurada T, Koike Y, Kansaku K. A hybrid BMI-based exoskeleton for paresis: EMG control for assisting arm movements. J Neural Eng 2017 Feb 09;14(1):016015

Maksimenko VA, Runnova AE, Zhuravlev MO, Makarov VV, Nedayvozov V, Grubov VV, et al. Visual perception affected by motivation and alertness controlled by a noninvasive brain-computer interface. PLoS ONE 2017 Dec 21;12(12):e0188700

 

Please rate this