• Jeremy Rutman

Taking an X-Ray to Nanox Imaging

Updated: Oct 19

Patent Portfolio Evaluation - An Analysis of the Israeli Unicorn


The Nanox claim of breakthrough technology in a digital Xray source appears to hinge on low-cost sources with improved stability and lifetime, which are capable of multispectral operation.

The ~100W Nanox source at $100 in mass production would indeed be cheaper than a modern ~10kW CT tube at ~$100K but we believe this is an apples-to-oranges comparison of a high-power (ca. 50kW) continuous-operation device to a low-power (100W) intermittent-operation device. Alibaba sources 1000W tubes at $100 as well and therefore it appears there is no basis for the Nanox claims on economic grounds unless some argument could be made that the Nanox sources are somehow superior (e.g. in terms of multispectral capability, lifetime or the like) - but we have found no such claims in the patent portfolio, nor any clear comparisons showing better performance.

While several of the Nanox patent applications deal with issues of lifetime and stability, the paucity of any measured data in these applications leaves one wondering whether these patents have ever gotten past the conceptual stage, although It is possible that the company hasn’t revealed such data in order to make replication more difficult.

Nano-X’s patent portfolio does not appear to be strong at this point. Two patents have been accepted, while a number have been abandoned. There are no claims directly involving device lifetime in any of the Nanox patents. In our opinion the Nanox portfolio provides rather thin protection for a device that is claimed to be a breakthrough . We find no one-on-one comparisons between the Nanox tube and a conventional tube of the same power. Most of the Nanox patents appear to derive from work prior to Nanox, done at Sony.

The Nanox business model may nonetheless succeed - their approach is to sell machines cheaply (conceivably, at $10,000 for a loss) en masse (e.g. hundreds at a time) to large entities (e.g. governments) and lock customers into multi-year service contracts. Thus the actual cost and technology of the machine may be largely irrelevant, as long as it provides usable images for some range of patients.

This is a video of the Nanox prototype installed in Hadassah hospital, apparently going through some early tests.

Interestingly there appears to have been an earlier version of this video including a 3D reconstruction that got scrubbed after someone pointed out some discrepancies.

And here's another promo showing a bit of detail on the Nanox sources:

Background on Xray sources

Conventional X-ray sources are similar to incandescent light-bulbs, with a hot filament. Unlike the light-bulb this filament releases electrons and hence serves as a cathode. The electrons are accelerated onto an anode (usually made of a high-melting-point metal like tungsten) using high voltage (e.g. tens of kilovolts) where the violent collision produces X-rays - at a low efficiency of ~1%, meaning 99% of the input power is turned into heat at the anode.

conventional xray tube above, schematic below [0]

Other means for producing free electrons can be used; for instance a sharp needle-tip held at a low voltage will create a large electric field that can cause electrons to be stripped from it and again accelerated towards an anode; this is called ‘field effect emission’. The production of such needle tips in large arrays using microlithographic techniques has been explored since the 1970's.

State of the Art - Pre-Nanox

The original field emitter array was the Spindt array, in which the individual field emitters are small sharp molybdenum cones. Each is deposited inside a cylindrical void in an oxide film, with a counterelectrode deposited on the top of the film. The counterelectrode (called the "gate") contains a separate circular aperture for each conical emitter. The device is named after Charles A. Spindt, who developed this technology at SRI International, publishing the first article describing a single emitter tip microfabricated on a wafer in 1968. Spindt, Shoulders and Heynick filed U.S. 3755704 in 1970 for a vacuum device comprising an array of emitter tips. Fig. 1 of this patent below shows the tips 16 ‘hiding’ in holes 15 which protect them and serve to shape the electric field.

The Nanox device appears similar: the following image is from a press release:

Nano-Spindt Array [7]

Nanox is hardly alone or even early here: such an array of needle-tip sources is shown below from a 2009 press release for Radius Health which may be the forerunner of Adaptix (see next figure).

Similarly the Adaptix flat panel source started their work in 2009:

Generally what we are looking for in an X-ray source is high current density (which means our light-bulb is brighter, for better images) and a small spot size (which means less blur for higher resolution and sharper images) in addition to the obvious (long lifetime and low cost) . Current density of several amps per cm^2 and spot size of a few mm or less are good values. At higher power the spot size will be larger ; a spotisze of 1mm at a power of 80kW is good for a conventional tube. Unfortunately, Nanox has not released any data on their device specs, neither current density, power, nor spot size, thus we can't compare their values to those in the market. Some other things to keep in mind are typical current involved. A typical conventional tube current , it seems, is of order 10-100mA. Here they speak of 300mA coming from an entire array of 50,000 tips, or 6μA per tip. So the notion of turning on tips in sequence (as in the Nanox promo video at the head of this article) to produce an image is not feasable - you'd need to turn on large numbers of tips to get a useful current.

Sanity Check - Keeping Cool

The key advantage that Nanox is claiming is cost of the Xray source. In the image above they compare a $150K conventional device with their $100 emitter array. We believe this is an apples-to-oranges comparison of a high-power to a low-power tube. The frame rate is directly affected by power (which directly affects Xray intensity).

One issue here is that to produce tomographic images quickly, the X-ray source must be rather powerful, e.g. ~50kW for a commercial CT tube that might cost $150,000 new. This high power is required to take multiple (like a hundred) images over a relatively short time period (like 20seconds). A single view e.g.for dental purposes, on the other hand, can use a single long exposure (e.g. 5s) and therefore can make do with less power (e.g. 1000W input for the $100 Alibaba dental source in the image below).

A second issue is that using high power for a long time will cause heat to build up at the anode, where 99% of the energy is turned into heat. So an intermittent-use device can get away without special cooling provisions, while a high-power continuous-use device will have to get rid of the heat at the same rate it comes in or wind up melting.

A quick-and-dirty check will give some feeling for what's involved:

A tube run at 10kW with a tungsten anode (heat capacity 0.14 J/gC) of 100g would (if uncooled) heat up by the following amount during a time t:

dT=Q/mc = 10kW*t/(.14*100) 

The anode might last up to about 50kJ or 5 seconds before melting (tungsten melts at 3420C).

In fact the heat load so great in a CT tube that a lot of innovation goes into heat control - the industry uses a special term 'HU' or heat unit, rates the tubes in terms of heat capacity, and the high-end tubes use methods such as direct-oil cooling, spiral-groove bearings, and rotating or liquid metal anodes all to deal with heat - this being one reason that the high power tubes cost so much.

From the 2nd video (above), there is a bit of text being read, stressing the point that the Nanox tube doesn't require a rotating anode to avoid excessive heating:

"Electrons pulsating from each cone hit a different spot on the anode which also keeps the anode cool without the need to rotate it."

target (artist's conception) [7]

source (beam is prob. added by hand) [7]

There are several problems apparent here:

  1. The more these target spots are separated, the larger the effective spot size and thus the more blurred the resultant image (which will be receiving light from a distributed rather than a point source; for a regular photograph this would be fine, but the Xray is a shadowgram...) The nanox source appears to be about 2mm x 2mm , while a CT spotsize of 1mm for a 10KW beam is apparently reasonable, and smaller spot sizes are used for research work. If you focus the spots to a central point to avoid blur (some of the Nanox patents deal with focusing), you're back to locally heating a tiny area with a large current.

  2. The distribution of heat over a larger area will give a similar result as a rotating anode over the short term, but for longer term continuous use, the device will have to reject heat at the same rate it is input; the heat capacity of the anode is limited and it will begin to heat up if heat is not rejected (e.g. using an oil flow or the like thermally coupled to the anode)...in other words, the same problems of limited heat capacity obtain for the Nanox device as for conventional devices.

This point gets somewhat misunderstood - for instance take this quote from SeekingAlpha:

X-rays rely on heating a metal filament to high temperatures to produce electrons that are accelerated to an anode across of it. This process generates substantial heat at the anode, and the cooling system required to cool the anode costs a substantial amount of money and equipment to deploy. NNOX believes that by directing electrons to small holes on a semiconductor, this will reduce heat gain and lead to a cheaper and smaller device.

The heat is indeed generated at the anode (where the electrons hit at great speed) since the 'braking radiation' or brehmsstrahlung that produces the X-ray is an energetically inefficient process. However, Nanox hasn't found a way around this, at least not in any of their patents, and it would seem that unless they are using an entirely different physical process than brehmsstrahoung, they cannot avoid this heat production. The heated cathode filament of a conventional tube, on the other hand, is a relatively minor energy input into the entire heat load; by going to a cold cathode you save only a few watts, while still dealing with kilowatts worth of energy output as heat on the anode side. A Nanox device of a given power would still have to reject the same amount of heat as a conventional tube, since it generated X-rays by the same inefficient process of braking radiation.

Also, the holes are at the cathode; the electrons are directed FROM small holes in the semiconductor cathode TO the anode (see images above).

A conventional 1000W tube suitable for dental Xrays can be obtained for ~100$ :

50KV 20mA conventional dental X-ray source from everybody's favorite, Alibaba

From the Nanox image below [7] we find their input power (for this image, at any rate) is 40KV*2.5mA=100W or one tenth the power of a conventional tube of similar price. Even if the Nanox device were, let's say, twice as efficient as the conventional, it would still be providing 1/5 the Xray power for the same price - less bang for the buck.

The main Nanox claim thus appears to be without merit - even if they can produce a 100W tube for $100 this will on its own not raise the bar in any way.

There are a few ways out of the heat impasse that I can imagine:

  1. The nanox device is somehow producing x-rays more efficiently. However the nanox device still uses electrons accelerated into  an anode, and therefore is using the same braking radiation mechanism as the conventional tube, and, it would seem, necessarily has the same low X-ray generation efficiency as a conventional tube.

  2. The use of N tubes allows each to be used 1/N of the time and cool down during off periods. This is fine but would still require high-power devices to produce images in the same short timeframe as the CT tube they are making the cost comparison to. And the fact is that cheap lower-power tubes are already available, with ten times the power Nanox shows in the shoulder scan above, for the same price.

And we should keep in mind that the technology may be irrelevant, and the real play here is an alternate business model wherein large numbers of machines are sold to governments or other large entities at cost or less in a subscription model, who they pay per-time or per-use fees for cloud-based analyses and upkeep.

Pros and Cons of the Field Emitter Array The use of an emitter array may carry several drawbacks and advantages:

  • The tips won’t wear as much if each is only used a small amount of the time. On the other hand each alone produces an unusably small current; large numbers of tips must be used simultaneously.

  • The source direction of the X-rays may be varied quickly (e.g. by quick turn on/off of particular sources), allowing for 3D information to be rapidly obtained without requiring a rotating gantry or 3D detector.

  • Thermally emitted electrons are emitted in random directions, while field emission electrons are directed forward, which may be primarily responsible for producing highly resolved x-ray images that have been obtained using field emission [1].

  • Field emission electrons have a narrow energy spread, which may also be behind the high resolution of electron-beam instruments [1][2]

  • It is difficult to maintain uniform current density in a field emitter array [3]. Field emission depends on the local environment near the tip. Stable field emission generally requires this environment to be stable, which will be the case in ultra-high vacuum, (UHV), while conventional x-ray tubes are operated in non-UHV. In non-UHV, field emission is very difficult to control, inter alia due to ‘‘cathode sputtering,’’ wherein residual gas molecules get ionized (to positive ions) and bombard the emitter. [6] This is why lifetime is a crucial factor for field-emission - it's a recognized achilles heel of the method.

Tomography vs. Tomosynthesis

The high-resolution CT images currently enjoyed by medical professionals is the result of many individual projections, that are used to reconstruct the scanned object or person in 2D or 3D. The fewer the projections, the worse the resolution:

Image reconstruction as function of number of angles used [4]

Nanox is using a small number of angles (e.g. coming from 12 tubes) so the backprojection images obtained would be 'information poor', either having low resolution and/or limited depth of field. The technique of reconstruction from a limited number angles not spanning 180 degrees is called tomosynthesis [5] while tomography uses the full 180 degrees.

Nanox Patents - A Detailed Analysis

“There were many attempts to create a digital, Cold Cathode X-ray source over the past few decades with little to no success.” ,aid Hitoshi Masuya, CEO of Nanox Japan, “while some companies have made achievements using carbon nano tubes as a basis for field emission X-ray with similar approach to the one used by Nanox, to the best of our knowledge no company have achieved a commercially stable source that can be embedded inside a medical imaging system and operate with an acceptable lifespan. We are proud at our achievement and look forward to beginning to revolutionizing medical imaging in the world.” Taking this statement at face value, the key improvements of Nanox are stability and lifetime – we keep this in mind in the following patent analysis.

There are 13 patent families assigned to Nanox Imaging LTD, Nanox Imaging PLC (GB) and Nanox Japan Inc, filed from Aug. 2012 to Jan. 2019. We’ve found two patents that are granted (highlighted in green in the table below), several that have been abandoned (highlighted in red), and the rest are pending. Most of the inventions appear to come from the Nanox Japan site.

We proceed to look at each of these in turn.

1. WO2014027294A3/US20150206698A1 Image Capture Device

This patent is an early one and includes a display - so it clearly was intended for a field emission TV of some sort and not an Xray device. The device uses an array of field emission electron sources, with a stratified resistive layer between the field emission type electron source and the cathode. A one- or two-layer focuser is mentioned as well as use of multiple energy ranges (by changing acceleration voltage) and intensity (by combining several emitters). However the claims deal only with a) using a spacer between source and display, and b) a method for preventing gate-cathode leakage current.

There is a potential problem of current leakage between the gate electrode 10 and the cathode 70 (see Fig. 25A above). Direct discharge between the cathode 70 and the gate electrode 10 may be prevented or at least limited through the introduction of a resistive interlayer 85A, but current leakage may still occur to along the surface path 86A adjacent to the electron source aperture. To prevent this, the pathlength can be increased as in 86B; this can be accomplished practically by using a sandwich structure as in Fig.25C below

The resistive layer has two different sections 854A and 854B with different resistivities (and resistivity here probably refers to resistance to etching or milling, not electrical resistance), allowing for a corrugated sidewall when the device is etched or milled) with increased path length and correspondingly decreased leakage current.

The claims also include extra high-resistance layers in section 80, to prevent high current density as current flows from the cathode to the conical emitter 9.

The specific use of silicon carbide as a resistor material is claimed (which may be useful e.g. to resist the radiation environment) but no information is given as to an increased lifetime or stability due to use of this material.

This patent has some good material (reduction of leakage current, reduction of current density near emitter, both by use of sandwiched layers). It does not deal with stability nor lifetime, however.

2. WO2013136299A1 US20190189383A1 Devices Having an Electron Emitting Structure - claims a cathode with several 'active zones' that can be controlled individually, all firing at a shared anode. These can be concentric, such that the focus size can be adjusted. On the other hand it is pretty well known how to focus electron beams from years of CRT and electron microscope development so its not clear if this has any practical use. The ability to fire from different areas would come in handy if the sources were firing from appreciably different angles as seen from the target; however the actual source size appears to be about 2-3 mm. and if the subject is at a distance of at least e.g. 20cm from the source then you reach a angle subtended of about 0.6 degrees maximum, which is unlikely to be useful for any tomosynthesis scheme.

The application also claims doing away with the grid electrode that is otherwise used in many situations. There are further some graphs of spot size, one of which is shown below:

A spot size of 100microns (as in the figure above) is unremarkable in the context of X-ray sources, with mini-focus tubes providing 50 microns, and micro-focus reaching 5 microns spot size (see ref); at any rate the full specification would be spotsize at given intensity or current.

We also note that these results derive from simulation and not measurement, further cementing the impression that the project was still at concept stage at the time of this patent submission.

US20190189383A1 is somewhat modified from the parent WO2013136299A1 .

3. WO2014009832A1 “Imaging device with election source array” - discloses an array of electron-emitting elements where each pixel unit comprises an AND gate, a field emission type ("FET") electron source and a pixel selection transistor. The disclosure further relates to an image capture device with an electron receiving construct and an electron emitting construct facing each other, separated by a set of spacers. Thus this is largely a reboot of WO2013136299A1 above. There is a bit more tech here, with claim 1 having a row scanning driver, a column scanning driver for driving the emitter array. The row/column addressing here however is a standard for many array devices such as LED screens, CMOS sensors, and the like. This claim allows one to address the cones individually - but with a single tip you would have a tiny current apparently around 0.1uA (micro-amp). A typical conventional tube current , it seems, is of order 100mA , and here https://avs.scitation.org/doi/10.1116/1.1849189they speak of 300mA coming from the whole array of 50,000 tips. So you'd have to use all or a large chunk of tips to get a readable image on the detector. A possible advantage of being able to use subsets of tips independently would be that then you can switch the source direction .But if these chip arrays are too small and too far away from the target then using emitters from opposite sides of the chip will hardly affect the source direction .

4. WO2015079393A1/US20190221398A1 Electron Emitting Construct Configured With Ion Bombardment Resistant [sic]

This patent is for design of an x-ray emitter device configured to prevent the emitters from being damaged by ion bombardment in high-voltage applications. The invention avoids ion bombardment damage in high-voltage applications, by means of setting a non-emitter zone surrounded by or set between the emitter areas. There is also an angled target anode or a stepped target anode (see right image 5A below) to further reduce the ion bombardment damage; angled targets are the standard in thermionic Xray tubes however.

Since this does appear to be about lifetime, lets look at the claims:

Claim 1: A electron emitting construct comprising: an array of field emission type electron sources and a plurality of control contacts configured for controlling said electron sources; a focus electrode configured for applying a voltage above said array; and a shield disposed over said control contacts.

Specifically claiming that part of the focus electrode serves as a shield may be quite useful, if hard to get through the USPTO on grounds of obviousness (remember this is not an accepted patent but rather an application).

However - in none of the figures is the lifetime shown (e.g. as a function of shield thickness or hole diameter) or reduction in damage assessed - leading us to wonder if this has gotten past the idea stage or not.

5A. US20180005796A1 - X-ray Tube and a Controller Thereof

This application discloses a cold cathode using a special focusing structure which has a plurality of focal point areas that can be controlled independently, with voltages applied to the focusing structures controlling the position and shape of the beam impact as in the following figure.

However as the focusing structure appears to be conventional (a split annular ring) this does not strike us as a particular breakthrough and in any case has no clear bearing on lifetime or reliability.

5B. US20180075997A1 X-ray Tube and a Controller Thereof - this application is for a system using a plurality of cathode parts that can individually be turned on/off, and where the whole cathode structure can be rotated. These applications were abandoned, then recently (Sept. 8 2020) revived with an amendment to the claims. This amendment is for a control method that rotates with the rotation of the cathode . In our opinion this is a relatively minor advance given that rotating anode or cathode X-ray sources have been in use for decades.

6. US20170301505A1 X-ray tube and a conditioning method thereof

This is an attempt to deal with cathode deterioration due to sputtering:

"when a cold cathode source is used as an electron emission source, there is a problem that electron emission is easily affected by the degree of vacuum...because the electron emission is sensitive to a surface state of the cathode compared to a hot cathode. Particularly, it is known that in a Spindt-type cold cathode array using a molybdenum (Mo) material, a current decrease occurs due to generation of oxidizing gas in a vacuum tube being in an operating state (see J. Vac. Sci. Technol. B16, 2859 (1998)... Thus, for some situations, there is a problem that decrease in anode current occurs..."

The solution given here is to get the tube to a working condition with one part of the anode, then once its up and running, use a second section of the anode to run. So the cathode here has two regions which can independently be turned on/off. This patent has been abandoned. A petition to revive was recently (29 June 2020) sent and awaits an answer.

7. WO2019151249A1 Control method of x-ray imaging device - this patent attempts to reduce the number of X-ray tubes configuring a distributed X-ray source. This “involves a driving step S1 for sequentially driving the multiple X-ray tubes and a moving step S3 for moving the multiple X-ray tubes. After execution of the moving step S3, the driving step S1 is executed again.” This is a rather obvious method for operation of a distributed source and doesn’t relate to the basic technology of the emitters.

8. WO2019151248A1 Cold cathode x-ray tube and control method therefor - this appears to be a follow-on to US20170301505A1 , but this time with nanostructured emitters rather than two macroscopic emission sections.

“Provided is a cold cathode X-ray tube capable of preventing time-dependent decrease in anode current such that the cold cathode x-ray tube stably operates for a long period of time. A cold cathode X-ray tube (1) comprises: an electron discharge portion (10) including an electron discharging element using a cold cathode; an anode portion (11) disposed so as to be opposed to the electron discharge portion (10); a target (12) disposed on a part of the surface of the anode portion (11); a casing (15) in which the electron discharge portion (10), the anode portion (11), and the target (12) are disposed; and a hydrogen generation portion (14) made of material that generates hydrogen when electrons collide therewith and disposed at a part of the surface present inside the casing (15) other than the surface of the target (12).”

The hydrogen generation here is notable; this is claimed to be “effective in preventing ... a decrease in anode current” according to a nonpatent reference, and thus the hydrogen release is an attempt to stabilize the electron currents. However the current stabilization is not actually shown as measured (which would reassure the reader that the device had actually been built and tested) but rather sketched schematically, and hence we are left with the impression that this is a piece of guesswork.

9. WO2019151251A1 Method for controlling x-ray tube and device for controlling x-ray tube

The invention is directed towards tomography using distributed X-ray sources, providing stabilization by using feedback, e.g. with a lookup table for relating the desired emission current to the applied base current; the emission current is detected and fed back to the emitter gate, thus stabilizing it. While in principle this sounds like it should work , again there are no measurements shown of current stabilization, leading one to wonder whether this remains an idea on paper or has actually been tried. Furthermore feedback is one of the basics of electric engineering and thus getting this application through a national patent office may be an uphill battle.

10. WO2019151250A1 X-ray imaging device and synthesis method of tomosynthesis image

The invention provides another synthesis method for generating tomographic images. Multiple distributed X-ray sources, each having multiple X-ray tubes arranged at a fixed pitch P1 are provided. The distributed X-ray sources are arranged separated by a space SP, and the distance P2 between two X-ray tubes that are adjacent but separated by the space SP is greater than the pitch P1.

This is a ‘junk patent’ in my opinion – dealing not with basic technology of detectors but matters of spacing existing detectors.

11. WO2019244874A1 Cold cathode electron source and x-ray generator equipped with same

This provides an easily replaceable cathode; “an X-ray generator according to an aspect of the present invention comprises: a vacuum container having an opening; an anode at least a portion of which is located inside the vacuum container; a flange which is detachably attached to the vacuum container and secured to the vacuum container to close the opening thereof; and a cathode secured to the flange and including a cold cathode electron source located inside the vacuum container. This makes it possible to easily and replace just a cathode section.”

This deals not with tube lifetime or stability but rather a maintenance issue.

12. WO2020141435A1 System and method for providing a digitally switchable x-ray sources

This has very generally worded claims for digitally switching a low voltage driving circuit to activate the field emission cathode, thereby generating a pulse of x-rays. The xray source and detection device (here a scintillator) are synchronized (so that e.g. the detector is only detecting while the Xrays are on).

There doesn’t seem to be much 'meat' here; a low voltage is used to switch on the field emitters and a high voltage to accelerate the emitted electrons onto the anode, as one would expect in normal operation of a field emission device. The synchronization of source and detector is almost surely covered in the thermionic emission patent literature, but may possibly get approved here since the same method is being applied to the field emission source.

13. WO2020158644A1 Tomosynthesis device and method for driving same

The patent aims to provide a compact tomography device using several cold cathode X-ray sources arranged in a plane, where the X-ray sources are caused to emit in turn. Shutter signals and the activation signals may be synchronized to produce required x-ray detection profiles.

This application deals not with basic issues such as novel emission sources or geometry but rather provides a simplistic method for use of a planar array of sources, namely turning them on sequentially. The actual reconstruction of an image from this method isn't dealt with - this would involve some known mathematics as in the inverse Radon transform but the particular implementation here would have been nice to see; as it is this seems like a 'quick and dirty' application. One dependent claim allows for use of different energies in the different sources.


The Nanox claim of breakthrough technology in a digital Xray source appears to hinge on low-cost sources with improved stability and lifetime, which are capable of multispectral operation.

The ~100W Nanox source at $100 in mass production would indeed be cheaper than a modern ~50kW CT tube at ~$100K but this is an apples-to-oranges comparison. Alibaba sources 1000W tubes at $100 and therefore it appears there is no basis for the Nanox claims on economic grounds.

Nanox claims $100 sources allowing $10,000 systems - which does seem questionable given that a detector alone (which Nanox is not claiming to develop) can go far upwards of $50,000.

If the Nanox.arc uses several fixed-location sources and translates the arc as in their video then they would save a bit of mechanics compared to a regular CT machine.

The Nanox claim of improvements to a digital Xray source appear to hinge on price, improved stability and lifetime. While several of the Nanox patent applications deal with issues of lifetime and stability, the paucity of measured data in these applications leaves one wondering whether these patents have ever gotten past the conceptual stage. At the same time, it is possible that the company hasn’t revealed such data in order to make replication more difficult.

Let’s compare the Nanox results to a competitor. If we check e.g. US10524743B2 to Adaptix, we find a method for determining pitch of emitter elements involving the solution to the inequality

This is a far more substantive claim than any we have found in the Nanox applications – it is specific and fulfills the patent criteria of novelty, inventiveness, and usefulness.

Nano-X’s patent portfolio does not appear to be strong at this point. Two patents have been accepted, while a number have been abandoned – never a good sign. Of the accepted patents one claims a shield structure forming part of a focus electrode, which may in principle help with device lifetime, and the other introduces methods for reducing cathode-emitter leakage (which affects neither lifetime nor stability as far as we can determine). There are no claims directly involving device lifetime in any of the Nanox patents, and there is one application involving simple feedback loop for current stabilization. In our opinion this is rather thin protection for a device that is claimed to be a breakthrough in terms of stability and lifetime.

The Nanox business model may nonetheless succeed - their approach is to sell machines cheaply (conceivably, at $10,000 for a loss) en masse (e.g. hundreds at a time) to large entities (e.g. governments) and lock customers into multi-year service contracts. Thus the actual cost and technology of the machine may be largely irrelevant, as long as it provides usable images for some range of patients.


This analysis is inherently limited by several factors:


a. Unpublished documents obviously cannot be located;

b. Documents in languages not in the scope of the search will be missed excepting those with abstracts within scope of the search;

c. Searches are performed within time constraints that limit the amount of material that can be analyzed.


The assessment is limited in scope to analysis of the technical merits of the patents and applications found, while the actual utility of a given patent is often out of proportion to its actual technical merit and can be heavily influenced by extraneous factors involving expedience, personal judgements, and others. The technical merit appearing in this report is judged subjectively by an experienced patent attorney limited in his assessment to information publicly available at the time of the assessment.

Hadassah hospital comment on Nanox activity there:

"Hadassah Hospital's partnership agreement was signed with the company [Nanox] in September 2019. Within the terms of the agreement it was agreed that Hadassah would provide services to the company and there would be a possibility of joint development in the future.

In accordance with the agreement, at the end of December 2019, the company's prototype device was transferred to Hadassah for several weeks, during which models made of plastic materials, simulating the human body (not real patients), were scanned. After several weeks the device was taken by the company, before humans were scanned.

The company received images scanned by the prototype device that the company had developed, as well regular images of the models scanned by Hadassah equipment."


[1] Rev. Sci. Instrum., Vol. 75, No. 5, May 2004

[2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4083558/

[3] https://avs.scitation.org/doi/full/10.1116/1.1849189

[4] http://xrayphysics.com/ctsim.html

[5] https://en.wikipedia.org/wiki/Tomosynthesis

[6] Rev. Sci. Instrum., Vol. 83, 094704 (2012) Quasi-monochromatic field-emission x-ray, Babacar Diopa and Vu Thien Binhb

USPTO, expacenet






Interestingly, a novel tunable Xray source was recently (Sept. 2020) reported by Kaminer, using Van der Waals heterostructures and other atomic superlattices irradiated by moderately relativistic electrons.


©2023 by Rutman IP. Page design by Eye Light Design