We have all been impacted by Covid-19 in some way. The media is full of stories and speculation that are not always accurate. Today I will highlight a few of the technology developments that are being quickly progressed to help the globe to get on top of this virus. The speed of development in this area is truly amazing.
Hopefully this is slightly different look at the issue than the flood of hype and insanity currently dominating social media.
Identifying people with the Virus
Early identification of those with the virus is critical to reducing the spread of the virus. Here are two (of the many) ways that technology is being used to find people with the virus.
AI powered smart glasses
These smart glasses use thermal imaging to measure a person’s temperature from up to 1 meter away. Developed by AI startup Rokid in China, it is claimed that each set can check the temperature of several hundred people in 2 minutes. Security Guards in Hongyuan Park in Hangzhou (150 kilometers south west of Shanghai) wear the glasses to check on visitors.
Once a person with a fever is identified an automatic alert is sent to staff and action is taken. The devices measure skin temperature which is not the same as core body temperature and people not yet displaying symptoms will not have a temperature however the system does have some use. It won’t catch all people with Covid-19 however it will help to reduce the number of people with a fever (flu or Covid-19) out and about.
An App to analyze your voice
A team at Carnegie Mellon University and other institutions have released an early version of an app that they claim can determine whether you might have Covid-19 just by analyzing your voice. The Algorithm seeks out signs of infection in your voice. It is not a diagnostic tool however it may be a useful early screening tool or a tool to be able to track the spread of the virus. The team has used many hours of audio from Covid-19 patients and from patients with other viruses. Covid-19 has an affect on the lungs that impacts breathing patterns and other parameters which allow differentiation. The major concern for the team is false negatives. They don’t want to be responsible for people with Covid-19 moving about in the community thinking that they are fine.
Cordio Medical is an Israeli startup that is developing a very similar system of using voice to detect illnesses. They began clinical trials on 2 April at Ramona Hospital. Their system will be used to remotely monitor the progress of Covid-19 patients that are self quarantined. The solution is built upon Cordio’s existing HearO product which is used to remotely monitor patients suffering from heart failure.
Corido engineers were able to adapt the HearO algorithm to identify bilateral pneumonia with edema in the lungs, a common characteristic of Covid-19. Cordio claims they can detect changes in lung fluids and the onset of inflammation prior to patients noticing. The system uses a baseline of the patients voice (any recording prior to the onset of illness) and the algorithm measures changes away from that baseline. They hope to eventually be able to detect Covid-19 deterioration within 10 to 48 hours of onset.
Cordio’s HearO product does not yet have FDA approval however it is in use in 10 hospitals in Israel. On a side note, there are a range applications using voice detection that are being developed. The Mayo Clinic is working on using voice as a predictor of stroke, heart attack, concussion and Parkinson’s disease (very much early research). Preliminary research at Carnegie Mellon has indicated voice analysis can fairly accurately diagnose post traumatic stress disorder and high blood pressure. Canary Speech has completed clinical trials for vocal diagnostic models for Alzheimer’s, Parkinson’s and depression, WinterLight Labs uses its’ platform of vocal biomarkers to determine cognitive impairment and depression. Corti, a Danish startup, sells a product that helps emergency call dispatchers identify patients experiencing cardiac arrest. Corti has also recently started researching Covid-19. Voice Analysis is a relatively new field but one that shows a lot of promise.
Data, its all about the data
In any pandemic the more data available that can tell us what is happening and how it is happening the better. A couple of significant efforts to collect relevant data to help in combating the virus are:
This is an Australian initiative lead by Professor Richard Osborne Head of Global Health and Equity in Melbourne and Professor at Swinburne University of Technology. He is being assisted by several astrophysicists from ANU (they have put together the website and are conducting the data analysis under direction from the Epidemiologists).
The goal is to collect data every day from healthy and sick people (anywhere in the world) in order to allow them to track the development and onset of Covid-19 symptoms. Please help them out by completing the survey (daily if possible, once is also fine) here.
European Commission initiative for collecting ideas about deployable AI and Robotics solutions
This initiative aims to create a repository of all the ideas that everybody has for defeating Covid-19. They are particularly focused on Robotic and Artificial Intelligence applications and solutions. The Repository is open for all to read and to contribute here.
When I checked the Repository there were already 70 well developed solutions that will be able to be deployed within the coming 2 to 3 months. The solutions ranged from apps to monitor people and symptoms, robots to undertake work in hospitals, disinfection robots, home Covid-19 test kits using online computer vision to analyze saliva crystals and automated detection of early breathing difficulties. The range and scope of the work in such a short time is quite staggering.
Metaculus
You may have heard about the wisdom of crowds. The thinking is that the crowd will be more often correct about a predicting an outcome than any one individual. Not always the case however this is often true. Metaculus has been testing this thesis since 2015. They now have a range of predictions that they are seeking input on for Covid-19. To predict you will have to register however you can see their site and the propositions that they are examining here.
Covid-19 Modelling
Once we have the data, a model needs to built that will allow us to try and predict the future behaviors of the virus and how it will impact society. This Epidemic Calculator lets you play with the assumptions and see the impact of different variables.
Minor changes in some of the variables can make a huge impact in the eventual outcome. One such variable is the number of people exposed to the virus. R0 or the Basic Reproduction Number is the average number of people an infected person will infect. Covid-19 has a high R0 of approximately 3. The flu has an R0 of 1.4, to give an example, after 10 days on average 14 people would have been infected with the flu however with an R0 of 3, after 10 days an average of 59,000 people would have been infected with Covid-19. We can’t change R0 of the virus however we can change the number of people exposed via Social Distancing and isolation. The model below shows a city of 7 million and the impact (e.g. Hong Kong). If you wish, you can play with this model and its’ assumptions by using the sliding inputs at the bottom. The model is on GitHub here.
The NZ Government released its’ Covid-19 model this week and the Australian Government is expected to do so shortly (the NZ model was based on the unreleased Australian model adjusted for the current NZ experience with Covid-19).
How can I avoid touching things
Easpeed from China, has developed a holographic control panel for elevators. At least one elevator in Hefei now lets people use holographic images of buttons to select the floor they want to go to. The company claims another 100 orders.
Google filed a patent for a similar product in 2003. They might now have a reason to develop and commercialize it. Easpeed also claims that they have a holographic registration and payment system for a hospital in Hefei. Covid-19 might speed up adoption of this sort of holographic interactive display.
Can we predict a future pandemic?
We have previously mentioned BlueDot from Toronto who alerted their clients in late December of the pandemic to come. The CSIRO in Australia has also developed models to help predict pandemics. In 2014 the CSIRO model predicted the Ebola epidemic 3 months before it was declared. The researchers are now applying the model to assist in predicting the spread of the Coronavirus in Australia.
The tool combines natural language processing, data science and statistical time series modeling to identify specific syndrome keywords and their context mentioned in Twitter posts. Thus facilitating the early detection of outbreaks whilst taking into account expected daily, weekly and seasonal influences.
The tool monitors flu like symptoms, coughs and head colds etc., which are currently most prevalent in NSW, Queensland and Victoria. Changes to the frequency and location of mentions of symptoms can provide predictive data for modeling. This allows authorities to proactively prepare for changes in demand for health interventions. This ultimately leads to better patient outcomes and less deaths. The data may also be an early indicator of any “flattening of the curve” in relation to Covid-19. Over time, analysis of the data can show better linkages of symptoms to catching or the severity of Covid-19 and thus even earlier intervention.
Some things to do during lockdown
Once you have finished Netflix and completed all your jigsaws, what will you do with all your time (working from home seems to take so much less time than in an office). Here are a few different suggestions that relate to Space and Technology:
Visit CERN and the Large Hadron Collider
CERN is the home of the Large Hadron Collider where scientists are trying to discover the fundamental building blocks of nature. You may have heard of the Higgs Boson. It was predicted in the 1964 by Peter Higgs but it was not experimentally proven to exist until the Hadron Collider confirmed its’ existence in 2012. Thanks to the magic of Google Streetview you can virtually walk around the Large Hadron Collider. Try it out here. It is an amazing feat of scientific engineering.
Discover a new Galaxy with GalaxyZoo
The recent improvements in the surveying of the night sky has given astronomers an incredible amount of data that needs to be analyzed. GalaxyZoo is an initiative that asks the public to help them classify galaxies based upon their shape. They currently have over 41,000 volunteers who have made 1,250,000 classifications. There are over 2 trillion galaxies in the universe so there are a lot of great galaxies still be identified. Coronavirus55 will have been discovered before you are finished.
Some Resources that may assist
Leading an Organization during Covid-19
Professor Harry Kramer from Kellogg School of Management at Northwestern University (one of my alma mater’s) has written an article “Two Principles for Leading your Organization through the Covid-19 crisis”. The article is available here. Prior to teaching at Kellogg, Professor Kramer was CEO of Baxter Healthcare in the US.
Australian Government app
Australian technology giant Atlassian and Canberra based Delv have developed an app for the Australian Government to disseminate information re Covid-19. Given the amount of bogus and outright false information circulating this is a trusted source. There is also a WhatsApp channel. The app can be downloaded from the Apple App Store or from Google Play.
Updated Data on Infection rates
If you want to see the underlying data on a global basis one of the best sites is Worldometers.info It is regularly updated from Official sources. It includes data on underling conditions, age of patients and infection and death rates per 1 million population.
Finally
If you have any questions or comments please email me via my website craigcarlyon.com or comment below.
I would also appreciate it if you could forward this newsletter to anyone that you think might be interested.
Till next week’s regular newsletter. Please stay safe from Covid-19.