Hacker News new | past | comments | ask | show | jobs | submit login
Indian researchers create a Raspberry-Pi-based device to monitor health (ieee.org)
207 points by sebastianvoelkl 7 days ago | hide | past | favorite | 54 comments

This is actually how existing diagnostic machines work generally - mix blood or plasma with a reagent, incubate at body temp for some time, measure how the light properties of the sample (color, turbidity, depends on the test) change over time. Some tests measure 1 point at the end of the test, some measure several and construct a non-linear function.

The interesting part is that they're building a machine that can get accurate results with low-cost off-the-shelf components. State-of-the-art systems typically cost somewhere between 50k and 500k, which may or may not include service agreements (they break down all the time - wet chemistry) and ongoing costs for supplies.

I imagine the main wins in this area would be:

- built with low-cost/commodity components (done)

- small physical size for transport (done)

   -- systems come in a variety of sizes from bedside to the size of a small car, depending on where you want to run them (operating suite/doctor's office vs central lab) and the volume you need.
- expanded operating temperature and humidity

   -- existing systems have fairly tight tolerances on humidity especially, all the systems I worked with had environmental sensors installed nearby.
- low-power enough to run on batteries or small generators easily in remote environments

That these women are working their way down the list is impressive and much needed. As mentioned in the article, reagents are available commercially so that part is relatively "solved," although depending on the test they can be quite expensive and have their own cold-storage and transport problems for remote areas. The other big problem is affordable control material to ensure the systems are still accurate, and calibration material to adjust system constants when they inevitably drift (same cost/storage/transport problems as reagents). Still though, I'm glad they're making progress.

I look forward to more!

Source: worked in lab diagnostics for a while

It's a neat system, although I fear the major hurdles (in the early days at least) will be getting a license for a diagnostics lab with these equipments and doctors approving the test results.

Also worked in lab diagnostics for quite a while. You're correct about the process. I'd add that the only particularly expensive piece of hardware along the chain is the light detection at the end. Sure, you need high precision pumps to dispense reagents & sample accurately, but if you're not running at industrial scale, those aren't that hard to DIY. And, of course, some reagents will be expensive as you said.

On the machines we built, the reaction products bound to a light-emitting material and we used photomultiplier tubes to measure the light output at the end. I believe this is more sensitive than the occlusion/scattering methods in the article. The problem is that PMTs are pretty expensive to buy new and their calibration needs to be checked periodically.

I have wondered if it would be possible to make a cheaper version using large-area PIN diodes, or with lenses to collect as much light as possible. Never saw any research being done in this area though.

SpaceX did the same thing.

They took a very complex system (rockets) that was custom-made and had to be extremely reliable. And then they looked at every component and tried to re-engineer it to optimize for costs. And they have been hugely successful doing it.

I have a common medical condition which is easily monitored via a common blood test.

It seems to fluctuate quite a bit, and if I could get frequent readings I could figure out what is affecting it, and I could better understand it's effect on me.

But the medical system in my country does not allow for frequent readings, probably for good reason - it would be very expensive. And a hassle going to a lab frequently.

So I'm excited to read about this device and hope the tech matures quickly.

How would you go about drawing blood? Is a finger prick good enough for these tests or would you need to know how to draw venous blood?

With titles like this, I always think it must either be trivial nonsense, or (more likely!) the journalist just really missed the point.

It seems that a couple of researchers have used RPis to analyze blood samples in an innovative way.

They are published: https://ieeexplore.ieee.org/document/9524612

So, you (OJFord) could not be bothered with either reading the simplified synopsis or if you did, understanding the original article.

Why even bother to comment, when you have nothing to give?

I have nothing bad to say against the researchers, as I said was likely, and as I ascertained to be the case after that and you commented on, it's just shoddy journalism. RPi is not the interesting part (at a stretch, a proxy for low powered compute?), 'Indian' is not the interesting part (but I can excuse that more readily as similar to 'XYZ university ...') and 'monitor health' is uselessly vague.

(The title could apply to a calorie counting app, running (arbitrarily) on a raspberry pi, for example.)

This valid contribution is dealt disservice with this title, and 'I (OJFord)' commented on that, 'giving', I hope, the suggestion that it might be better.

It’s a relatively descriptive title and accurate to the article. I don’t think it is terrible.

Here’s a link to the actual paper: https://ieeexplore.ieee.org/document/9524612

Abstract: The development of a cost-efficient and sensitive platform for biochemical analysis of blood serum and its realization in the low resource areas is one of the imperative challenges to establish a robust healthcare ecosystem. The present work demonstrates the design of a universal platform, capable of performing all biochemical analyses of blood serum by measuring the absorbance of light through the test sample. To verify the working of the developed platform, the concentration of glucose was estimated in blood serum. The detection of glucose has been accomplished in a linear range of 1 mg/dL to 400 mg/dL with detection of limit 1 mg/dL (R2 = 0.9875, n=3). The stability analysis demonstrates improved stability in the output as compared to the conventional analyzer with an average standard deviation of 0.32 calculated for n=5. Human blood samples were tested with the developed platform and the results were in line with the pathology laboratory. The developed platform offers the advantages of automation, low cost, portability, simple instrumentation, flexibility, and an easily accessible interface. Due to the use of a huge processing capability processor, the analysis time reduces to half a minute which yields fast analysis and high throughput. The stability and accuracy also improve owing to the employment of high-resolution electronics components. Overall, the proposed framework is an attractive solution to be incorporated in the low resource area as a universal platform for all biochemistry analysis simply by varying the wavelength of light and reagent.

What would be really cool and useful would be an open source bedside (ICU/OR) monitor. RPi would probably be a good base, I guess...

No way. Medical devices need very strong guarantees. RaspPi does not offer those. You want a cheap RTSoC.

Monitors and tests don't need to be perfect to be useful. If you were a doctor running a poor rural hospital, would you rather have:

1. No monitoring, or

2. Monitoring that fails to alarm for 50% of events.

If you have plenty of staff, maybe you'd prefer no monitoring, as it's worth than having someone sitting next to each bed. But that's probably not the reality of the situation.

I could be wrong, but I'm not convinced RaspPi is better suited for the task, even cost-wise. SoC are cheap these days - I have zero evidence to back this up, but I would wager most of the "cost" associated with e.g. ICU monitoring devices is strictly to non-technical aspects (brand name, margin, etc.)

In theory, an open source and safe design is definitely feasible. I just wouldn't marry "safe" and "raspi" on most days.

Sure, at scale there might be better options. But it's probably easier to develop a prototype on something like a Pi, than on an ESP32 or other development board. Because you can plug a keyboard, mouse and monitor into the Pi.

That's not how you develop embedded stuff normally. You usually do that over some serial line to control the remote debugger-related interrupts and whatnot and drive it via GDB. You can also emulate many SoC's via Qemu, Bochs, etc. (to varying degrees) during the development process.

Further, you don't run hard realtime applications on a normal OS. Preemptive multitasking can cause a userspace program to drift out of sync, or have other side effects - for example.

There is a pretty well known RTOS but the name is escaping me.

You wouldn't really benefit from prototyping on the pi as a lot of that software has to be purpose built for the hardware.

RPi+Linux+sensors has an easier learning curve, and is accessible to people with a wider range of backgrounds/skills.

I don't disagree with you, it's just that medical equipment that is critical (e.g. ICU monitors) would benefit from real, robust hardware. It's not about learning curves.

You'd be surprised. When I started my career in medical devices, we specifically based our product on Windows & C++ in part because it was a well-known platform and we would have no trouble finding developers. That system, and its descendents ran on off the shelf single-board Windows PCs.

There is a lot of medical equipment out there that's running Windows. FWIW, I have one next to me right now that uses Linux for its UI.

> It's not about learning curves.

When it comes to experimenting and prototyping I think it is. The lower the learning curve the more cross polinations take place, projects and ideas get shared.

Clearly RPI is not optimal for clinical machinery that is supposed to run unattended by engineers in the real world but for prototyping why not?

Because there isn't a lot in the way of "prototyping" necessary. We already know how to effectively monitor patients. A lot of the shitty parts are actually in software and licensing and process - not hardware.

If you "prototype" for raspi, you will have to completely start from scratch if you want to manufacture reliable devices later on. It's not like porting a program from one system to another.

Go build it!

> Monitoring that fails to alarm for 50% of events.

As Waymo found out early on - unreliable automation may lead to counter-intuitive results when people put too much trust in such systems, as humans do. It is in the realms of possibility that patients will get worse outcomes because doctors/medical staff are paying them far less attention than they were without any monitoring.

Although that would be the end game, trying to get off the ground would be much easier with a RPi

I agree. Nothing running a GPOS is suitable here.

But do those include graphical capabilities, USB connectivity, jack for alim, etc. all-in-one and for a few bucks? I am aware medical devices should provide guarantees, but a smaller guarantee is still better than none at all, innit? For context, in many countries people are simply left to die in the most atrocious ways if they cannot afford care. You should see how african people react when they're admitted to the hospital, even in developed countries.

> but a smaller guarantee is still better than none at all, innit?

Raspberry Pi comes with no guarantees at all and has well documented problems with flaky power and burning out SD cards. It works great for the hobbyist/learning system that it was designed for and might even be good enough for the blood testing system in the article, but active monitoring for intensive care? Just...no.

I'm no expert in this field and was genuinely asking, but if someone knows of better things then please go for it!

The raspberry pi will be a very capable CPU for a medical monitoring. The issue is the sensor and the power unit. Isolation is very important and the sensor reading needs to be correct.

It's certainly a fair question. I don't work in medical devices, but know people that have, and the problem for medical hardware has two steps: One, build something that won't kill the patient, and then step two is proving that you're correct about step one. Step two is the expensive bit.

Yes. For example, most of the STM32 lineup has builtin USB support, and can connect to displays through a range of protocols.

Stuff like this is always cool and gee whiz, but it will never muster in the US that I know of, and I suspect most developed countries.

Anything on the bedside, especially in ICU, has to be examined by the FDA to the point where they want to know exactly what the chemical composition of parts are. If there's a risk of things releasing fumes that might be dangerous for a sick patient if the device gets too warm, they'll ding it. Diagnostic software powering these devices are scrutinized by the FDA, too.

Ah yes, in the US they can't occur the risk of "fumes" from an affordable electronic device, that's why they need to charge people a year's income for a minor emergency room visit.

Never mind the fact that for millions in the US the air is toxic, the water is full of lead, and the food full of agro-toxins. But they are lucky, because the FDA makes sure medical devices don't release any accidental fumes and they only cost 2 or 3 magnitudes more than they need to.

Of course it would never do in developed countries! For context, I work with such monitors daily in a (very) rich country. I was thinking of the most backwater context you could still find use for it. Some parts of Africa, for instance.

I suppose a parallel would be a Mars Helicopter based on off the shelf components vs making one of radiation hardened components.

It may work, and it may work very well in fact. However, if you can afford it - you'd opt for something with more reliability and consistency.

In economics, the most important question is "as compared to what?" You might be able to afford a gold plated system designed with 15-9's of reliability, but do you want to work 10 hours week to afford that system? On the margin, do you want to have a 10% smaller house to pay for that level of medical awesomeness?

Not Open Source but relatively cheap : https://shberrymed.com/products/patient-monitor-pm6100

A company called "Anidra" (Australia + India) is offering the full package (HW based on the PM6100 and cloud based Monitoring SW) both for Hospital and Home use: http://www.anidra.com.au/ and https://www.anidra.in/

Weird choice of name Anidra in Hindi means sleeplessness. Nidra is sleep, anidra is its opposite.

On second thought, I think they are probably implying their devices (or they themselves) do not sleep.

>implying their devices (or they themselves) do not sleep.

Yep, 24/7 monitoring. Their website mentions that they are already in trials which is good.

Let’s do it!

Why aren't more low cost Medical devices available easily and cheaply? Given the power/price ratio we have achieved with cellphones, why can't the same be done in the field of Medical devices?

For example, i was looking for a all-in-one monitoring device for an elderly patient and came across "PM6100" made by "Shanghai Berry Electronics" (https://shberrymed.com/products/patient-monitor-pm6100) Low-cost but still more expensive than many cellphones. These things should be commodity priced. It almost feels like there is some "organized cartel" preventing the invention and marketing of low-cost medical devices.

Red tape. Medical devices need to (depending on type and country) conform to medical device regulations, be approved by a medical device regulator, be sold only to medical professionals or people with a script from a medical professional. Liability in case something goes wrong because of your device is expensive, adversarial and dependent on crossing all the 'i's and dotting all the 't's. There are "lower levels" of this due to regulations being less for more harmless devices. E.g. you can nowadays get thermometer, pulse-oximeter or a blood pressure meter quite cheaply. But anything just mimimally more complicated or critical gets expensive very fast.

Reasons for this beside the red tape are imho the low number of customers (most slightly specialized medical devices are needed once per patient with $rare_disease, once per lab or once per doctors office), the high need for customization (one-size-fits-all doesn't even work for blood pressure cuffs, let alone prosthetics), localization (broken i18n can kill, most customers are elderly and therefore not as versed in engrish), higher component cost (sterilizable plastics are more expensive, bigger displays for vision-impaired elderly clients are more expensive) and acceptance of foreign/small/unknown manufacturers (won't trust my elderly mother's health to a device from "Corty's Refurbished Asbestos Plates, Health Equipment and Luxuries Ltd., Templestreet, HongKong (CRAPHEALLTH)").

There is also a cartel of each medical professionals, manufacturers, insurance companies/public insurance pools and politicians, complete with revolving doors, kickbacks, fake or real -but always suspiciously convenient- scientific data, and exclusionary legal situations. All cementing the status quo and the wealth and standing of all participants (except the patients' of course).

>But anything just mimimally more complicated or critical gets expensive very fast.

"Critical" i understand, "Complicated" i dispute. Our existing technology has significantly brought down this threshold and it should no longer be a limiting factor.

>There is also a cartel of each medical professionals, manufacturers, insurance companies/public insurance pools and politicians, complete with revolving doors, kickbacks, fake or real -but always suspiciously convenient- scientific data, and exclusionary legal situations. All cementing the status quo and the wealth and standing of all participants (except the patients' of course).

This is what i believe is the real reason. In fact sometimes i think i should spend the rest of my career/life to overturn the status quo with the help of professionals from the Open Source community many of whom would gladly spend their time and money in helping their fellow human beings get affordable healthcare.

>These things should be commodity priced.

These things couldn't be commodity priced (comparable to a smartphone) until they are manufactured at similar scale. You just don't need as much medical devices.

I don't buy this commonly given excuse at all. Companies like Omron/Rossmax etc. have made BP monitors dirt cheap (and now pulse oxymeters). Add a few more functionalities needed for patient monitoring, price accordingly and you have a market.

It is well within the capabilities of existing technologies (both HW and SW) and though there are stringent regulations i don't think they are the limiting factor. I am convinced the Companies involved in the manufacture of these devices (eg. GE Healthcare) are artificially keeping the prices of these high enough to deter widespread adoption.

> though there are stringent regulations i don't think they are the limiting factor

This is exactly the limiting factor. I could build you a Pulse Ox device with a nice graph on a Raspberry pi in a few days/weeks quite easily and put it in a pretty case with battery power and connectivity to the internet and your phone if needed. A cool hobby project.

However, selling it as a Medical Device (those words have a specific meaning) will raise the price exponentially. Let's look at some of the things we need to take care of in order to keep the FDA happy.

Development process: the software & hardware needs to have been developed with a predefined process, with specific minimum steps taken and certain documentation needs along the way. This doesn't sound like a big deal until you realize just how much it can slow down the development process and add more cost. If I can't prove that those steps were all taken, then I have to create a validation process to prove that the part I bought off the shelf (like the Pi) is suitable for the task -- not a difficult thing to do, but all these little things add to the cost of producing something.

Supply chain: you need to be able prove the provenance of every item in your device back to the originating manufacturer. If the manufacturer changes something, there needs to be a mechanism so you can be alerted to evaluate if it has an impact on your product. No buying cheap parts from pulse0x7337 on Amazon.

Manufacturing: basically set out a process for building it and make sure there are no deviations from that process and be able to prove that you are always following that process.

Complaints: you need to have in place a method for users of your device to report issues with it and to be able to respond quickly and correctly, in some cases within a federally mandated time (if in the US) depending on how serious the complaint is.

Compliance: ensuring compliance with 21CFR is such a huge task that there are entire companies that are dedicated to doing this for you if you can't afford your own team.

Support: can't remember details off the top of my head but you are required to provide support for your device for xx years after you remove it from the market. This means the capability to answer user questions and perform repairs if needed.

So yeah, it might seem like an excuse, but it really isn't!

Nice writeup, appreciate it !

However the points you raise need not be that stringent towards "non-critical" and "mass use" medical devices. Add proper disclaimers to the sale of these devices and everybody benefits. In particular, the people in the developing world many of whom lack even the most basic medical support. Use these in the field (eg. underdeveloped villages) and at homes to get an approximate reading after which you can be referred to a proper diagnostic lab/hospital. Incidentally, some of these diagnostic labs in the developing world are a absolute scam with completely broken, unmaintained and error-prone devices. If the cost is brought down they can actually afford to buy decent devices.

For example, at one point you couldn't even afford home BP monitors and Pulseoxymeters. But now they are common place in particular the latter which before the Covid-19 pandemic were expensive. This in turn has resulted in the public getting better educated in learning to measure and monitor basic health signs which is very good.

In the end it's always a Regulatory issue depending on the regulations of the country where the device is sold. I'm sure there are countries where regulations are very lax, but those will tend to be countries where the market is tiny, so most reputable manufacturers will avoid them.

I guess what I'm trying to get at is that I think these Open Source Hardware/Software medical device projects are great. I really do. However people need to understand why they can't be considered the equivalent of a device that's built following the necessary regulations so they can make informed decisions about how they can be used.

> the Companies involved in the manufacture of these devices (eg. GE Healthcare) are artificially keeping the prices of these high enough to deter widespread adoption

One of the ways this will happen is the regulatory burden required to bring a product to market, which they will be keen to keep in place as a competitive moat.

>which they will be keen to keep in place as a competitive moat.

This is exactly it. Their size/money/monopoly gives them the advantage if they can keep "disruptive" competitors at bay by mere red tape/regulations/paperwork. Common business strategy but deplorable.


why do you think so?

Isn't this Theranos 2.0?

why do you think so?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact