7 ms·

Catching It Early: New Ways of Detecting Coronavirus

82 points by sohkamyung 6y ago

usrusr 6y ago
If I get this right this sensor could be reused until it is saturated with positive detection? (one or more positive patients depending on intensity of the signal) That would be a complete game-changer compared to PCR where tests that turn out negative are exactly as expensive as positives which incentivizes against testing without suspicion.
- usrusr 6y ago
  ps: very unlikely to have reusability beyond one positive, I failed to consider the increasing concentration in the successive spikes. No change in my general assessment, could be the second best thing to a vaccine if it can be turned into a well performing product.
- karmakaze 6y ago
  This has a good impact on the economics of testing. If each test device paid for can find one positive case, there's motivation to expand testing.
isoprophlex 6y ago
What is really nice about this is that they show response to actual human samples, not just purified spike protein vs. control proteins/blanco buffer controls. This gives me hope that a finalized test could have good specificity... Finally, clinical samples were tested to analyze the performance of the biosensor. Human swab samples from a normal person and a COVID-19 patient were collected separately. The normal sample was loaded onto the device before the addition of the patient sample. Figure 3a clearly shows that patient samples produce a different signal from the normal sample within a couple of minutes. Even after diluting the patient sample 100000 times, clear responses were observed.
WilTimSon 6y ago
Impressive that it catches the virus even in a highly diluted sample. I'm trying to figure out if this also works faster than other tests? Can't tell from the excerpt.
- spuz 6y ago
  If you look at figure 2a, it looks like they are dropping successively more concentrated samples of SARS-CoV-2 onto the sensor and getting immediate spikes in the response. But it's not clear whether or not this technology could be developed into something that can be used at home or in the field. That is the real challenge you need to overcome in order to make a test with a quick response. Packaging a sample, sending it to a lab, getting a technician to put the sample through the test, recording the result and communicating it back to the patient - that is what takes time not the actual performance of the test itself.
  - usrusr 6y ago
    > successively more concentrated samples Good observation, I saw that but completely failed to consider the ascending nature in my assessment in my first post. So reusability would have an upper bound at the first positive. Could still be an incredibly valuable addition to the using Arsenal of they manage to shape a deployable product around the technology (ed if it turns out to be a car-sized behemoth, if the throughput is right)
- usrusr 6y ago
  Detecting spike protein omits the unwrapping steps that are part of the PCR test protocol. And this test appears to be formalized for saliva from the start, which is far more practical than throat swabbing. Some studies indicate that PCR would very similar results if it was switched to saliva, but that would require recertification. Switching would probably be a net negative because swabbing is far from the only bottleneck and because protocol change artifacts would taint statistics.
- Ensorceled 6y ago
  The graphs suggest results in minutes, especially graphs 3a and 3b. Unclear on amount of setup and cleanup between tests but 3a also suggests good results by adding a second sample to a test in progress.
OJFord 6y ago
'Biosensors' are a really cool blend of fields, and manufactured for more ordinary things like pH measurement too, but they (not SARS-CoV-2 ones specifically obviously) seem to me to be completely unavailable to the hobbyist/anyone that doesn't want to have a meeting to discuss a small sample of 100k units to get started. Is there just insufficient market for them to make them more available, or is fabrication really expensive, or something else? I asked about it recently, but no traction or comments: https://news.ycombinator.com/item?id=22996014 https://news.ycombinator.com/item?id=22996014
- usrusr 6y ago
  Lazy layman's guess, posted to provoke more knowledgeable answers: expensive combined with very low durability. I just can't imagine semiconductor surface doped with complex proteins to last very long when exposed to arbitrary samples. If I'd wanted to have this sensor in a near-future hard sci-fi I'd world-build it the sensor to last a day at most to keep it convincing. Comes in a cartridge of 16, each day the machine unseals a new sensor then you need to pop in a new disk.
  - OJFord 6y ago
    I don't think that's the case - when I was looking into it prior to posting the above Ask HN, I saw a video (a bit marketing-y, but interesting) for a company that was deploying pH sensing ones in their packaged product dotted around in the sea (harbour or fishery or something) in buoys for some reason or other. It didn't seem like it was a consumable that would need continually replacing, more like a 'deploy and forget' humidity or temperature sensor (as in thermistor, I don't know if you can also get temp FETs).
- amelius 6y ago
  You might be surprised: https://www.sciencedirect.com/science/article/pii/S1566119920301804 https://www.sciencedirect.com/science/article/pii/S156611992... > Organic-based field effect transistors for protein detection fabricated by inkjet-printing
  - 6y ago
toshk 6y ago
South-korea, Taiwan and a few other Asian democracies have shown great organisational skills in this crisis. The title "catching it early" suggested to me that the article would be about macro solutions for societies to "catch it early". For instance sewage detection: https://www.bloomberg.com/news/articles/2020-03-31/coronavirus-in-sewage-portended-covid-19-outbreak-in-dutch-city https://www.bloomberg.com/news/articles/2020-03-31/coronavir...
- Ensorceled 6y ago
  This appears to be a non seqitur. Why did the success of Asian countries make you think that this would be about something else? Especially since a lot of South Korea's success was tied to the their extensive testing efforts. Taiwan's success attributed almost entirely to their taking a far more aggressive approach to isolation than WHO was recommending (also notable, Taiwan is not part of WHO due to China's efforts to keep them out).
  - arcticbull 6y ago
    > Taiwan is not part of WHO due to China's efforts to keep them out. Taiwan isn't part of the WHO because the WHO is part of the UN and broadly speaking, member states of the UN -- the United States included -- don't officially recognize Taiwan. As an aside, the US relationship with Taiwan is entirely unofficial. The US political stance follows the One China policy and recognizes Beijing as the capital of a China inclusive of both what is colloquially the PRC and the ROC. For what it's worth, it's likely due to the close (if challenging) relationship the PRC and ROC enjoy that Taiwan was able to get a head start on containment.
    - adventured 6y ago
      > As an aside, the US relationship with Taiwan is entirely unofficial. The US political stance follows the One China policy and recognizes Beijing as the capital of a China inclusive of both what is colloquially the PRC and the ROC. That's false on multiple counts. The official US policy only acknowledges China's position and does not recognize China as having sovereignty over Taiwan. This has been clarified publicly on numerous occasions since the early 1980s. The Taiwan Travel Act further upgraded US-Taiwan relations and made them official. That's why China was so upset by it - publicly stating that it violated their One-China principle - and why the Taiwanese President was able to freely travel to the US afterward.
acqq 6y ago
Covered by ScienceDaily in April: https://www.sciencedaily.com/releases/2020/04/200420145029.htm https://www.sciencedaily.com/releases/2020/04/200420145029.h... What I didn't find covered is how hard is to maintain the "sensing" state of the sensor for more than a single measurement. If I understood correctly, actual antibodies have to be put on the graphene. The described process for that is far from trivial, in the paper: "The fabricated graphene-based device was soaked in 2 mM PBASE (Thermo Fisher Scientific, Waltham, MA) in methanol for 1 h at room temperature and then rinsed several times with PBS and DI water. Finally, the functionalized device was exposed to 250 μg/mL SARS-CoV-2 spike antibody for 4 h."
juskrey 6y ago
Tests are useful. Why no one is bothering about humane side of tests outcome? E.g. who compensates effectively lost airline ticket for someone positive right before boarding? For we stop him/her for public safety but also punish like criminal, what for?
- raziel2p 6y ago
  Travel insurance. "punish like criminal" - what are you talking about? Also, you're talking about the humane side but focusing on losing money because of void airplane tickets. I think hindering a pandemic is more humane than worrying about losing a few hundred dollars.
  - juskrey 6y ago
    You were never flying, right? No travel insurance covers this. I am glad you are talking about trillions of dollars in savings but against compensating tickets lost which is often a big personal issue.
- cozzyd 6y ago
  Every single airline is offering no change fees for a long time. Ok maybe there are are some airlines that don't (Spirit? Wizz Air? I don't actually know, just assuming.) So they'll be able to take their trip later.
  - juskrey 6y ago
    Dropping off at the airport (instead of concealing the infection by all means) is a public service - and should be compensated as such. Otherwise we'll get perverted incentives.
m0llusk 6y ago
It is important to be scientific about the context. There are already cheap, fast, accurate tests for COVID-19 related antibodies. This test is comparable to other antibody tests and not to the PCR tests which are fundamentally more complex and slow while also being better at detective active infection. Probably the current biggest competitor would be the Abbot Labs antibody test, so this is really going to come down to cost, availability, and just how much the speed matters. Little differences can really add up when doing population scale testing, but waiting a few minutes might not necessarily be different enough from nearly instant to be worth additional cost. It will be really interesting to see if production of transistor based sensors can keep up with chemical antibody test production.
- mark-r 6y ago
  I didn't think they were detecting antibodies, they were using antibodies built into the transistor to affect the transistor response.

82 points by sohkamyung 6y ago

If I get this right this sensor could be reused until it is saturated with positive detection? (one or more positive patients depending on intensity of the signal) That would be a complete game-changer compared to PCR where tests that turn out negative are exactly as expensive as positives which incentivizes against testing without suspicion.

usrusr 6y ago
ps: very unlikely to have reusability beyond one positive, I failed to consider the increasing concentration in the successive spikes. No change in my general assessment, could be the second best thing to a vaccine if it can be turned into a well performing product.
karmakaze 6y ago
This has a good impact on the economics of testing. If each test device paid for can find one positive case, there's motivation to expand testing.

isoprophlex 6y ago

What is really nice about this is that they show response to actual human samples, not just purified spike protein vs. control proteins/blanco buffer controls. This gives me hope that a finalized test could have good specificity... Finally, clinical samples were tested to analyze the performance of the biosensor. Human swab samples from a normal person and a COVID-19 patient were collected separately. The normal sample was loaded onto the device before the addition of the patient sample. Figure 3a clearly shows that patient samples produce a different signal from the normal sample within a couple of minutes. Even after diluting the patient sample 100000 times, clear responses were observed.

WilTimSon 6y ago

Impressive that it catches the virus even in a highly diluted sample. I'm trying to figure out if this also works faster than other tests? Can't tell from the excerpt.

spuz 6y ago
If you look at figure 2a, it looks like they are dropping successively more concentrated samples of SARS-CoV-2 onto the sensor and getting immediate spikes in the response. But it's not clear whether or not this technology could be developed into something that can be used at home or in the field. That is the real challenge you need to overcome in order to make a test with a quick response. Packaging a sample, sending it to a lab, getting a technician to put the sample through the test, recording the result and communicating it back to the patient - that is what takes time not the actual performance of the test itself.
- usrusr 6y ago
  > successively more concentrated samples Good observation, I saw that but completely failed to consider the ascending nature in my assessment in my first post. So reusability would have an upper bound at the first positive. Could still be an incredibly valuable addition to the using Arsenal of they manage to shape a deployable product around the technology (ed if it turns out to be a car-sized behemoth, if the throughput is right)
usrusr 6y ago
Detecting spike protein omits the unwrapping steps that are part of the PCR test protocol. And this test appears to be formalized for saliva from the start, which is far more practical than throat swabbing. Some studies indicate that PCR would very similar results if it was switched to saliva, but that would require recertification. Switching would probably be a net negative because swabbing is far from the only bottleneck and because protocol change artifacts would taint statistics.
Ensorceled 6y ago
The graphs suggest results in minutes, especially graphs 3a and 3b. Unclear on amount of setup and cleanup between tests but 3a also suggests good results by adding a second sample to a test in progress.

OJFord 6y ago

'Biosensors' are a really cool blend of fields, and manufactured for more ordinary things like pH measurement too, but they (not SARS-CoV-2 ones specifically obviously) seem to me to be completely unavailable to the hobbyist/anyone that doesn't want to have a meeting to discuss a small sample of 100k units to get started. Is there just insufficient market for them to make them more available, or is fabrication really expensive, or something else? I asked about it recently, but no traction or comments: https://news.ycombinator.com/item?id=22996014 https://news.ycombinator.com/item?id=22996014

usrusr 6y ago
Lazy layman's guess, posted to provoke more knowledgeable answers: expensive combined with very low durability. I just can't imagine semiconductor surface doped with complex proteins to last very long when exposed to arbitrary samples. If I'd wanted to have this sensor in a near-future hard sci-fi I'd world-build it the sensor to last a day at most to keep it convincing. Comes in a cartridge of 16, each day the machine unseals a new sensor then you need to pop in a new disk.
- OJFord 6y ago
  I don't think that's the case - when I was looking into it prior to posting the above Ask HN, I saw a video (a bit marketing-y, but interesting) for a company that was deploying pH sensing ones in their packaged product dotted around in the sea (harbour or fishery or something) in buoys for some reason or other. It didn't seem like it was a consumable that would need continually replacing, more like a 'deploy and forget' humidity or temperature sensor (as in thermistor, I don't know if you can also get temp FETs).
amelius 6y ago
You might be surprised: https://www.sciencedirect.com/science/article/pii/S1566119920301804 https://www.sciencedirect.com/science/article/pii/S156611992... > Organic-based field effect transistors for protein detection fabricated by inkjet-printing
- 6y ago

toshk 6y ago

South-korea, Taiwan and a few other Asian democracies have shown great organisational skills in this crisis. The title "catching it early" suggested to me that the article would be about macro solutions for societies to "catch it early". For instance sewage detection: https://www.bloomberg.com/news/articles/2020-03-31/coronavirus-in-sewage-portended-covid-19-outbreak-in-dutch-city https://www.bloomberg.com/news/articles/2020-03-31/coronavir...

Ensorceled 6y ago
This appears to be a non seqitur. Why did the success of Asian countries make you think that this would be about something else? Especially since a lot of South Korea's success was tied to the their extensive testing efforts. Taiwan's success attributed almost entirely to their taking a far more aggressive approach to isolation than WHO was recommending (also notable, Taiwan is not part of WHO due to China's efforts to keep them out).
- arcticbull 6y ago
  > Taiwan is not part of WHO due to China's efforts to keep them out. Taiwan isn't part of the WHO because the WHO is part of the UN and broadly speaking, member states of the UN -- the United States included -- don't officially recognize Taiwan. As an aside, the US relationship with Taiwan is entirely unofficial. The US political stance follows the One China policy and recognizes Beijing as the capital of a China inclusive of both what is colloquially the PRC and the ROC. For what it's worth, it's likely due to the close (if challenging) relationship the PRC and ROC enjoy that Taiwan was able to get a head start on containment.
  - adventured 6y ago
    > As an aside, the US relationship with Taiwan is entirely unofficial. The US political stance follows the One China policy and recognizes Beijing as the capital of a China inclusive of both what is colloquially the PRC and the ROC. That's false on multiple counts. The official US policy only acknowledges China's position and does not recognize China as having sovereignty over Taiwan. This has been clarified publicly on numerous occasions since the early 1980s. The Taiwan Travel Act further upgraded US-Taiwan relations and made them official. That's why China was so upset by it - publicly stating that it violated their One-China principle - and why the Taiwanese President was able to freely travel to the US afterward.

acqq 6y ago

Covered by ScienceDaily in April: https://www.sciencedaily.com/releases/2020/04/200420145029.htm https://www.sciencedaily.com/releases/2020/04/200420145029.h... What I didn't find covered is how hard is to maintain the "sensing" state of the sensor for more than a single measurement. If I understood correctly, actual antibodies have to be put on the graphene. The described process for that is far from trivial, in the paper: "The fabricated graphene-based device was soaked in 2 mM PBASE (Thermo Fisher Scientific, Waltham, MA) in methanol for 1 h at room temperature and then rinsed several times with PBS and DI water. Finally, the functionalized device was exposed to 250 μg/mL SARS-CoV-2 spike antibody for 4 h."

juskrey 6y ago

Tests are useful. Why no one is bothering about humane side of tests outcome? E.g. who compensates effectively lost airline ticket for someone positive right before boarding? For we stop him/her for public safety but also punish like criminal, what for?

raziel2p 6y ago
Travel insurance. "punish like criminal" - what are you talking about? Also, you're talking about the humane side but focusing on losing money because of void airplane tickets. I think hindering a pandemic is more humane than worrying about losing a few hundred dollars.
- juskrey 6y ago
  You were never flying, right? No travel insurance covers this. I am glad you are talking about trillions of dollars in savings but against compensating tickets lost which is often a big personal issue.
cozzyd 6y ago
Every single airline is offering no change fees for a long time. Ok maybe there are are some airlines that don't (Spirit? Wizz Air? I don't actually know, just assuming.) So they'll be able to take their trip later.
- juskrey 6y ago
  Dropping off at the airport (instead of concealing the infection by all means) is a public service - and should be compensated as such. Otherwise we'll get perverted incentives.

m0llusk 6y ago

It is important to be scientific about the context. There are already cheap, fast, accurate tests for COVID-19 related antibodies. This test is comparable to other antibody tests and not to the PCR tests which are fundamentally more complex and slow while also being better at detective active infection. Probably the current biggest competitor would be the Abbot Labs antibody test, so this is really going to come down to cost, availability, and just how much the speed matters. Little differences can really add up when doing population scale testing, but waiting a few minutes might not necessarily be different enough from nearly instant to be worth additional cost. It will be really interesting to see if production of transistor based sensors can keep up with chemical antibody test production.

mark-r 6y ago
I didn't think they were detecting antibodies, they were using antibodies built into the transistor to affect the transistor response.