DNA-LAMP Diagnostic Device

Detect HIV, Ebola, Tuberculosis and other diseases, cheaply and reliably.

Similar projects worth following
Billions of people have no way to access modern medicine, and millions die there each year from communicable diseases. A baby girl born in Sub-Saharan Africa faces a 22 per cent risk of death before age 15. Most of these people will never have access to modern medicine, and by the time they could reach a doctor from their remote homes, it's often too late to help.

This project will help these people. We can bring the very latest in advanced molecular biology techniques to their villages, in a simple, cheap and rugged device. With the push of a button, it will diagnose their diseases by analyzing pathogens at the DNA level, and give a simple LED read-out.

It might sound like science-fiction, but the research is real. And it's ready to make a big difference to billions of people.

There are 3 billion people on our planet who live in extreme poverty.

Living on less than $2.50 a day, medicine is a luxury.

This project is for them.

Combating diseases in developing countries is vitally important. Each year over 1,5 million people in these countries die of AIDS, and another 1,5 million people dies of Tuberculosis. Malaria, Ebola, SARS and many more infectious killers add to the misery. Treating these diseases correctly and preventing their spread both alleviates suffering, and also paves the way to economic growth and a way out of poverty.

The very first step in tackling disease is diagnosis, and its a hard problem. For example, during a Tuberculosis outbreak some people might be asymptomatic and spread the disease whilst appearing healthy, and until they are treated, they could infect hundreds of people. Or during an Ebola outbreak, placing a patient with a harmless fever in quarantine with Ebola-infected patients is a huge risk for them. Unfortunately, diseases do not always give reliable symptoms during their onset, and for the poorer half of the planet, waiting to long for the right treatment can kill.

But we know the causes of these diseases. We have given the virus, bacteria and parasites scientific names, analysed their life-cycles and molecular structures, and have sequenced their genomes. But bringing the latest technology to the poorest countries in the world is not easy. However, in just the past few year, a new diagnostic procedure has been developed, and it could change everything.

LAMP is a special type of DNA amplification process, a bit like PCR. It is highly specific to a defined DNA sequence, and can correctly identify specific disease-causing agents from small samples of blood or saliva. It has isothermal reaction conditions, which mean the reaction is carried out at a single temperature, simplifying the diagnostic procedure. It uses cheap reagents that can be desiccated and stored at room-temperature for months. And it’s results can be processed easily, without special equipment.

This project aims at taking current scientific knowledge, generated from hundreds of researchers from around the world, and creating a simple and reliable diagnostic device. A machine that is so simple and cheap that it can be posted in an envelope, and can be run off a motorcycle battery in the most remote villages. And its protocol is so easy to use, that it can be learnt with a simple demonstration, even for people who can’t read or write. You just mix the samples, push a button, and 30 minutes later a LED lights up to give the diagnosis.

Molecular diagnostics, with the push of a button.

It seems like magic, the ability to detect disease from just 10 or so single bacteria, with a device that costs less than $10, but I think it can be done. Over the next few week, and by the end of the competition, I intend on:

  • having a working prototype
  • have validation studies completed
  • having full documentation on the hardware and software published
  • and having the molecular biology protocol ready

I will also go through the literature, and create an ever-expanding list of diseases and the required DNA primers to detect them, which I will add here. I will start with disease don't need a reverse transcription step, as that requires a longer protocol.

Disease Detection Primers

Each set of primers should cost about $50-100 and be enough for about 2,500 reactions, or about 2-4 cents per test. Buying in larger quantities brings the price down dramatically.

Mycobacterium tuberculosisPrimer

Staphylococcus aureusPrimer

Shigella speciesPrimer
Read more »

  • 1 × ATTINY84-20SSU Microprocessors, Microcontrollers, DSPs / ARM, RISC-Based Microcontrollers
  • 1 × PMV40UN2R N-Channel Mosfet - heater control
  • 1 × TPS560200DBVR Power Management ICs / Switching Regulators and Controllers
  • 1 × VLMU3100-GS08 SMD UV Clear Non-Diff
  • 1 × VEML6040A3OG Photo IC Sensors RGBW Clr Snsr w/I2C 16-Bit 2.5-3.6V

View all 14 components

  • Parts have arrived!

    David09/13/2015 at 18:53 2 comments

    With just a week to go, and two big boxes full of parts from Mouser on my bench, this week is assembly time.

    On the last update, I showed off my new laser-made PCB boards, and now I have the parts to populate it. It's my first SMD design and I have to hand-solder everything as I don't have a reflow oven. So, just is case I mess up with such little time left, I have also ordered through-hole parts to make a perfboard functional prototype. The perfboard version will also be much easier to iterate on, as I use an Arduino Uno in place of a ATTINY841, which means I have enough programming space to include debugging and logging code over serial. I'm left with so little space on the 841 that I can't afford that luxury.

    In parallel, I have Bst Polymerase for the wetlab experiments, to generate fluorescent sample to test my spectrofluorometer for the activated Calcien dye.

    Wish me luck!

  • Laser-assisted PCB Fabrication

    David09/04/2015 at 20:33 6 comments

    One of the 2015 Hackaday Prize criteria is "“Wow” factor: is the entry innovative, is the build impressive?" Well, aside from the "Wow" of a DNA-based molecular diagnostic device, I hope this entry log is a little bit 'wow' on the build side too :)

    PCB boards, stat!

    Having got my (hopefully!) final schematic and PCB layout finalized, today it was time to fabricate my first PCB. Luckily, if I have messed it up, I can make new versions quickly using this technique :)

    Board Preparation

    I used off-the-shelf single sided copper board. I gave them a light coat of black spray paint, waited 5 minutes, and a second light coat covered the copper I missed on the first pass. After 10 minutes, the boards were touch dry, and ready to use. I used a hacksaw to roughly cut out a section of board the right size for my design.

    File Preparation

    During the drying time, I prepared my PCB files. My board layout has the interface at the top, consisting of a reset and run button, and a RGB status LED. In the middle is the power input on the left, microcontroller in the middle and ISP programming port on the right. At the bottom is the heater, and the sample chamber location, marked by the grey square. Under it is the RGB sensor and UV excitation source for the fluorescent dye.

    Making a PCB quickly requires just a few steps. The first is to hide all layer that should not be etched. I hide everything except for the top layer, pads and the dimension layer.

    Then, I exported the image, at the highest resolution of 2400 DPI, and set the export to Monochrome. After that, use any graphics program to invert the black and white image, and you are ready to laser your boards.

    Laser Engraving

    Like most CO2 laser cutters, the one I used has an "engraving mode", that can mark acrylic, anodized aluminium and wood etc. with a black and white image. A quick test found that the engraving settings for acrylic were perfect for ablating the spray paint from the copper surface, allowing the exposed copper to be etched in the next step. It's import to mention that you can't cut the copper laser directly with a regular CO2 laser system. You can do it with a expensive pulsed laser system, but you won't find one of those at your local Hacker Space. Anyway, you just place your painted PCB in the machine, load up the files with the settings for acrylic engraving, and you have your mask ready in about 3 minutes!

    PCB Etching

    After that, you do a regular PCB etching. I used a Ferric Chloride bath using the regular method you can find on google, and then removed the black spray paint mask with some solvent. Acetone removed spray paint very easily, so you won't need more than a few millilitres on some tissue paper. A quick soak and wipe, and you are finished!

    If you have a few painted copper boards ready, you can go from the designs on your computer to a physical PCB board in about 5 minutes more than your etching time!

  • Initial Schematic done!

    David08/28/2015 at 22:34 0 comments

  • $2 Fluorescence Spectroscopy

    David08/21/2015 at 14:37 0 comments

    I just got my sensor working! This is the key element of the LAMP diagnostic device, and something I don't think has ever been done before. Digital multi-channel fluorescence spectroscopy with a sub $2 chip, The other half of the device, the precision temperature regulation system, is something I already have working in my PCR Machine. So, the main components of the diagnostic device are past the proof-of-principal phase!

    A big thanks goes to Christoph on the Hacker Channel, who helped get my I2C device working!

    I'm using the VEML6040 RGBW color sensor, and a 405NM UV led to excite fluorescein, which had the same excitation and emission profile as Calcein, the dye I plan to use for the detector device.

    The SMD part is a tiny 4-pin device, which I have soldered onto copper wire, and then onto header pins.

    The illumination is set at a 90 degree offset to the sample.

    I use the UV to excite sample of water and 50 micromolar Fluorescein.

    The fluorescein solution generates a bright green fluorescence, and the water none. Then if was time to record data, over the serial port from the Arduino:
    UV led H2O:
    red channel8291
    green channel11501
    blue channel46281
    white channel65535
    UV led 50um Fluorescein:
    red channel11543
    green channel17613
    blue channel44271
    white channel65535

    Theres a good increase in the Green channel, over 50%. Whats great, is that by using a RGB led, we can do positive and negative controls!

    UV led 50um Fluorescein, lid open
    red channel 43481
    green channel 41390
    blue channel 35962
    white channel 65535
    UV led 50um Fluorescein, LED off
    red channel 311
    green channel 271
    blue channel 111
    white channel 1208

    If the UV led breaks, the blue channel drop by 99%, and if the lid is opened, the red channel goes up 500%. What that means is we can detect if there is a problem with the UV led, or if someone opens the device while the reaction is going on, both of which are vital for a diagnostic device!

    Thing are going really well! I think with this key step done, I'll have the device completed in time for the 2015 competition!

  • Cheap Spectroscopy

    David08/18/2015 at 15:56 0 comments

    I found a nice research paper on using LEDs as wavelength specific light sensors. You can find it on google:

    'Spectral and emission characteristics of LED and its application to LED-based sun-photometry'

    As one might suspect, the maximum absorption wavelength is blue-shifted compared to the emission spectrum. So whats nice is that LEDs are dirt cheap, and the yellow ones are sensitive to Green light, but reject blue and UV. They might make an inexpensive sensor for my diagnostic device. I have ordered a bunch, from multiple manufacturers, and well as regular RGB detector devices. Let hope I find a cheap solution!

  • Order complete, onto testing!

    David08/15/2015 at 21:19 0 comments

    OK, I went a bit overboard, and ordered about 100 euros worth of LEDs, optical sensors, temperature sensors, regulators and mosfets... Lucky I won a minor prize in last years competition!

    My goals for this week are to test the best method for DNA detection, either Absorption spectroscopy are 650nm with HNB, or Calcein fluorescence. Both are tricky, when your budget is so small, and trivial with a $30,000 spectrofluorometer...

    Product Detail

    Order Qty.

    Order Qty.

    50,851 €
    50,492 €
    50,463 €
    200,258 €
    50,624 €
    50,34 €
    50,624 €
    LH R974-LP-1
    50,101 €
    LS R976-NR-1
    50,101 €
    50,467 €
    50,606 €
    101,31 €
    50,523 €
    101,22 €
    51,96 €
    50,151 €
    50,119 €
    50,131 €
    50,11 €
    LY R976-PS-36
    50,101 €
    50,176 €
    50,16 €
    LA L296-P1R2-Z
    50,171 €
    50,516 €
    51,72 €

  • Parts list

    David08/14/2015 at 10:00 0 comments

    OK, I have a new design in mind. I will use Calcein as a DNA detection dye, and use a UV led and green photosensor to detect is the reaction has occured.

    The parts I plan on ordering from Mouser are:

    Intersil ISL29035IROZ-T7

    Light to Digital Converters Integrated Digital Light Sensor


    Vishay Semiconductors VLMU3100-GS08

    405nm SMD UV LED

    I was planning on using a temp sensor from Microchip, the:

    Microchip Technology MCP9805T-BE/MC

    Board Mount Temperature Sensors Ser output temp sensor

    But Mouser doesn't have them in stock :( So now I have to find another.

  • Time to Vote!

    David07/29/2015 at 14:05 0 comments

    Should I use Loop-mediated isothermal amplification (LAMP) or Helicase-dependent amplification (HDA) for the core technique? Please, post your comments and ideas!

  • Slow progress

    David07/29/2015 at 10:18 0 comments

    So, to build my first prototype, I have to build my first CNC machine, so I can machine the PCR tube-holder out of aluminium.

    I have now got my spindle, and just have to mount it to my CNC framework, and model the parts in 3D etc etc etc.

View all 9 project logs

Enjoy this project?



2016ee187 wrote 07/10/2020 at 09:43 point

can you share the code and the circuit diagram with us please

  Are you sure? yes | no

joseph.oleniczak wrote 11/07/2017 at 15:03 point

Do you happen to sell your DNA-LAMP Diagnostic Devices?  I would be interested in purchasing one.    

  Are you sure? yes | no

Pat Cho wrote 04/05/2017 at 23:25 point

Do you have any more updates of this project? Thanks!

  Are you sure? yes | no

Sastri wrote 01/12/2016 at 14:05 point

  Are you sure? yes | no

Sastri wrote 01/12/2016 at 14:02 point

Thanks David, May I request you for more elaboration, if possible...Thanks in advance

  Are you sure? yes | no

Sastri wrote 01/10/2016 at 17:10 point

Just wanted to check the progress of the project. Any updaets???

  Are you sure? yes | no

David wrote 01/11/2016 at 09:47 point

The boards are assembled, but I have had much more interest in the PCR machine. I plan on having the PCR machines available in kit-form for purchase in a few months.

  Are you sure? yes | no

Luke Gary wrote 06/09/2015 at 02:33 point

I'll be watching this one closely! 

Will you be doing relative quantification through some mode of fluorescence or conductivity?

Have you looked into cheap ways to quantify in real time without complex probe chemistry?

What reaction vessel will you be using? Standard PCR tubes or a pipe of some kind?

  Are you sure? yes | no

David wrote 06/09/2015 at 11:23 point

I still plan on using a loop system. That way I don't need to change temperatures dynamically. Double-stranded fluorescent DNA dyes seem to be the best solution, as they can often be excite with non-damaging blue light. I'm hoping I can use a LED as a wavelength-specific light sensor to cut down the cost of the detector, but I plan of having the sensor perpendicular to the exciter, and go right through the telfon tubing.

  Are you sure? yes | no

Luke Gary wrote 06/24/2015 at 16:48 point

Be Careful about that! Often the emission wavelength of a given dye is only a few tens of nanometers shifted from the excitation wavelength. So little so, that the spectral output of the LED is still too wide, you may try a laser pointer instead of an LED since LED based systems still require expensive optics to filter out the excitation band, provided that lasers output wavelength is compliant wiht the dye you would like to use. Although there are some cheaper materials to use at the cost of degraded performance. 

  Are you sure? yes | no

David wrote 07/29/2015 at 14:09 point

I'm thinking about LAMP PCR, and using turbidity to detect the reaction, via scattering to a sensor at 90 degrees to the light source. Is requires more primers and a special polymerase, but its isothermal :)

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates