pith. sign in

arxiv: 2604.13004 · v2 · submitted 2026-04-14 · 📡 eess.IV

Inexpensive Optical Projection Tomography on a Mobile Phone Platform

Gennifer T. Smith , James M. Sikes , Nicholas Dwork This is my paper

Pith reviewed 2026-05-10 13:52 UTC · model grok-4.3

classification 📡 eess.IV
keywords optical projection tomographymobile phone microscopylow-cost 3D imagingzebrafish phantomfiltered backprojectionportable tomographyresolution testing
0
0 comments X

The pith

A mobile phone with $50 in add-ons performs optical projection tomography at 3.91 micrometer resolution.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that a functional optical projection tomography system can be assembled from an iPhone camera, a commercial microscope lens attachment, a stepper motor, LED lights, and 3D-printed parts. Projection images are captured while the sample rotates, converted to attenuation maps, corrected for uneven illumination, and reconstructed slice-by-slice with filtered backprojection before stacking into volumes. The resulting system resolves fine details in a fabricated zebrafish phantom, including the spine. A reader would care because this removes the need for specialized lab equipment, opening 3D optical microscopy to classrooms, field sites, and low-resource laboratories.

Core claim

The completed system achieved a resolution of 3.91 μm and produced volumetric reconstructions in which anatomical features of the zebrafish phantom, including the spine, were clearly visible, using only an iPhone camera, low-cost lens attachment, stepper motor, LED illumination, and custom 3D-printed components with total added cost around 50 dollars.

What carries the argument

The low-cost OPT setup that rotates the sample under phone-camera view, converts projections to attenuation images, applies field nonuniformity correction, and reconstructs each slice by filtered backprojection before stacking.

If this is right

  • Volumetric 3D images of small biological samples become obtainable without dedicated laboratory tomography instruments.
  • Filtered backprojection on corrected phone-camera projections can resolve anatomical structures at the scale of a few micrometers.
  • Portable, battery-powered 3D microscopy becomes feasible for education and field applications.
  • Low-cost phantom fabrication by embedding fixed larvae in UV-cured resin provides a repeatable test object for system validation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same hardware approach might adapt to other small transparent specimens such as embryos or plant sections.
  • On-device reconstruction software could turn the phone into a self-contained 3D imager without external computers.
  • Calibration routines developed here could transfer to other phone-based optical systems that need consistent projection geometry.

Load-bearing premise

The low-cost phone camera, lens attachment, and custom corrections produce projection data of sufficient quality and consistency for filtered backprojection to yield accurate 3D volumes without major artifacts or distortions.

What would settle it

Reconstruction of the zebrafish phantom that fails to resolve the spine or shows large geometric distortions or missing features compared to the known phantom structure.

Figures

Figures reproduced from arXiv: 2604.13004 by Gennifer T. Smith, James M. Sikes, Nicholas Dwork.

Figure 1
Figure 1. Figure 1: a) Example of crimp bead-sealed tube under UV light. Note that there isn’t any zebrafish in this tube. b) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A schematic of the arrangement of components of the tomographic microscope. The electronics includes an Ardunio Nano, TMC2209 driver, and connections to power the LEDs. The light gray region shows the portion that is encased in a housing to hold all components in the proper location and to block external light. 2.2.1 Motor and Electronics We used a NEMA 17 Stepper Motor (0.9 degree step size, 1.5 A) combin… view at source ↗
Figure 3
Figure 3. Figure 3: Images of the checkerboard pattern used to identify the intrinsic camera parameters of the camera: a) an [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Images of the 1951 Air Force Target captured with an iPhone camera a) without and b) with the additional [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Top row: examples of the luminance projection images captured by the iPhone. Bottom row: corresponding [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: a) Outside view of 3D-printed box. b) Disassembled 3D-printed box to show how it can lay flat for shipping. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: a) Motor attachment used to attach the sample tube to the NEMA motor shaft and keep the sample upright. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Image of the 1951 Air Force target captured by the smartphone microscope with a zoomed in region showing [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Reconstructions of individual slices of the zebrafish phantom. a) A projection image of the zebrafish phantom [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Visualizations of the reconstructed volume using 3D Viewer within the the FIJI ImageJ application. [PITH_FULL_IMAGE:figures/full_fig_p009_10.png] view at source ↗
read the original abstract

This work presents an inexpensive optical projection tomography (OPT) system built on a mobile phone platform for three-dimensional optical microscopy. The system uses an iPhone camera together with a low-cost commercial microscope lens attachment, a stepper motor for sample rotation, LED illumination, and custom 3D-printed components, with a total component cost of approximately 50 US dollars excluding the phone. To support system evaluation, we also developed a low-cost method for fabricating a zebrafish phantom by embedding fixed larvae in UV-cured resin. Camera calibration was performed using a checkerboard target, and effective magnification was estimated with images of a 1951 Air Force resolution target. Projection images acquired during sample rotation were converted to attenuation images and corrected for field nonuniformity. Each slice was reconstructed with filtered backprojection and the resulting slices were stacked into a 3D volume. The completed system achieved a resolution of 3.91 $\mu m$ and produced volumetric reconstructions in which anatomical features of the zebrafish phantom, including the spine, were clearly visible. These results demonstrate that mobile-phone-based OPT can provide accessible, portable, and low-cost 3D microscopy, with potential utility for education, field work, and resource-limited settings.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript describes construction of a low-cost optical projection tomography (OPT) system using an iPhone camera, commercial microscope lens attachment, stepper motor, LED illumination, and 3D-printed parts (total component cost ~$50 excluding phone). It reports a low-cost zebrafish phantom fabrication method, checkerboard camera calibration, Air Force target magnification estimation, projection-to-attenuation conversion with field nonuniformity correction, slice-by-slice filtered backprojection, and stacking into 3D volumes. The central claims are a measured resolution of 3.91 μm and clear visibility of anatomical features including the spine in the reconstructed zebrafish phantom volumes.

Significance. If the reported resolution and feature visibility are confirmed to be free of reconstruction artifacts, the work would establish that portable, sub-5 μm 3D optical microscopy is achievable with consumer-grade hardware and minimal cost, directly supporting applications in education, field biology, and resource-limited laboratories.

major comments (1)
  1. [Abstract and reconstruction procedure] Abstract and reconstruction procedure: no description is given of how the stepper-motor rotation axis (center of rotation) was located, calibrated, or corrected prior to filtered backprojection. Sub-pixel misalignment in FBP is known to produce streaking or blurring artifacts that can either obscure or fabricate linear structures such as a zebrafish spine; the claim that anatomical features are 'clearly visible' therefore rests on an unverified assumption of perfect alignment.
minor comments (2)
  1. [Abstract] The abstract states that effective magnification was estimated with a 1951 Air Force resolution target, but does not report the specific group/element used or the line-pair frequency at which contrast fell to a defined threshold (e.g., 10 % or Rayleigh criterion).
  2. [Results] No quantitative metrics (e.g., contrast-to-noise ratio, edge sharpness, or comparison against a commercial OPT system) are mentioned to support the visual assessment that features are 'clearly visible'.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their detailed and constructive review. The major comment highlights an important omission in the description of our reconstruction procedure. We address this point below and commit to revisions that strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and reconstruction procedure] Abstract and reconstruction procedure: no description is given of how the stepper-motor rotation axis (center of rotation) was located, calibrated, or corrected prior to filtered backprojection. Sub-pixel misalignment in FBP is known to produce streaking or blurring artifacts that can either obscure or fabricate linear structures such as a zebrafish spine; the claim that anatomical features are 'clearly visible' therefore rests on an unverified assumption of perfect alignment.

    Authors: We agree that the manuscript lacks a description of how the center of rotation was located, calibrated, or corrected, and that this is a substantive gap. Sub-pixel misalignment in filtered backprojection can indeed introduce streaking or blurring that might affect interpretation of linear features such as the spine. In the revised manuscript we will add a dedicated subsection under the reconstruction procedure that details the alignment protocol, including the method used to determine the rotation axis, any verification steps performed with test projections, and the approach taken to minimize residual misalignment before applying filtered backprojection. This addition will enable readers to evaluate the likelihood of reconstruction artifacts and will support the reported visibility of anatomical structures. revision: yes

Circularity Check

0 steps flagged

No circularity: purely experimental hardware demonstration with no derivation chain

full rationale

The paper reports construction of a low-cost OPT system (iPhone + lens + stepper motor + 3D-printed parts), standard camera calibration on a checkerboard, magnification estimation on an Air Force target, projection acquisition with field nonuniformity correction, slice-by-slice filtered backprojection, and stacking into volumes. All results are direct physical measurements and standard reconstruction; no equations, fitted parameters renamed as predictions, self-citations, or ansatzes are invoked to derive the claimed 3.91 μm resolution or feature visibility. The work is self-contained against external benchmarks (resolution target, phantom imaging) with no reduction of outputs to inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard tomographic reconstruction principles and optical calibration methods drawn from prior literature, with no new free parameters, invented entities, or ad-hoc axioms introduced beyond the choice of commercial components.

axioms (1)
  • domain assumption Filtered backprojection can accurately reconstruct 3D volumes from 2D attenuation projections when field nonuniformity is corrected and the sample is rotated through sufficient angles.
    Invoked in the description of converting projections to attenuation images and performing slice reconstruction.

pith-pipeline@v0.9.0 · 5513 in / 1330 out tokens · 36548 ms · 2026-05-10T13:52:08.249462+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

  1. [1]

    Three-dimensional imaging

    Fatima A Merchant and Alberto Diaspro. Three-dimensional imaging. InMicroscope image processing, pages 247–317. Elsevier, 2023

    work page 2023

  2. [2]

    Introduction to confocal microscopy.Journal of Investigative Dermatology, 132(12):1–5, 2012

    Adaobi Nwaneshiudu, Christiane Kuschal, Fernanda H Sakamoto, R Rox Anderson, Kathryn Schwarzenberger, and Roger C Young. Introduction to confocal microscopy.Journal of Investigative Dermatology, 132(12):1–5, 2012

    work page 2012

  3. [3]

    Optical coherence tomography.science, 254(5035):1178–1181, 1991

    David Huang, Eric A Swanson, Charles P Lin, Joel S Schuman, William G Stinson, Warren Chang, Michael R Hee, Thomas Flotte, Kenton Gregory, Carmen A Puliafito, et al. Optical coherence tomography.science, 254(5035):1178–1181, 1991

    work page 1991

  4. [4]

    Optical projection tomography.Annu

    James Sharpe. Optical projection tomography.Annu. Rev. Biomed. Eng., 6(1):209–228, 2004

    work page 2004

  5. [5]

    Image processing assisted algorithms for optical projection tomography.IEEE Transactions on Medical Imaging, 31(1):1–15, 2011

    Abbas Cheddad, Christoffer Svensson, James Sharpe, Fredrik Georgsson, and Ulf Ahlgren. Image processing assisted algorithms for optical projection tomography.IEEE Transactions on Medical Imaging, 31(1):1–15, 2011

    work page 2011

  6. [6]

    Correction of artefacts in optical projection tomography.Physics in Medicine & Biology, 50(19):4645–4665, 2005

    Johnathon R Walls, John G Sled, James Sharpe, and R Mark Henkelman. Correction of artefacts in optical projection tomography.Physics in Medicine & Biology, 50(19):4645–4665, 2005

    work page 2005

  7. [7]

    OptiJ: Open- source optical projection tomography of large organ samples.Scientific reports, 9(1):15693, 2019

    Pedro P Vallejo Ramirez, Joseph Zammit, Oliver Vanderpoorten, Fergus Riche, Francois-Xavier Bl´e, Xiao-Hong Zhou, Bogdan Spiridon, Christopher Valentine, Simeon E Spasov, Pelumi W Oluwasanya, et al. OptiJ: Open- source optical projection tomography of large organ samples.Scientific reports, 9(1):15693, 2019

    work page 2019

  8. [8]

    Review of tissue simulating phantoms for optical spectroscopy, imaging and dosimetry.Journal of biomedical optics, 11(4):041102–041102, 2006

    Brian W Pogue and Michael S Patterson. Review of tissue simulating phantoms for optical spectroscopy, imaging and dosimetry.Journal of biomedical optics, 11(4):041102–041102, 2006

    work page 2006

  9. [9]

    Fabrication of healthy and disease-mimicking retinal phantoms with tapered foveal pits for optical coherence tomography.Journal of Biomedical Optics, 20(8):085004, 2015

    Gary CF Lee, Gennifer T Smith, Monica Agrawal, Theodore Leng, and Audrey K Ellerbee. Fabrication of healthy and disease-mimicking retinal phantoms with tapered foveal pits for optical coherence tomography.Journal of Biomedical Optics, 20(8):085004, 2015

    work page 2015

  10. [10]

    Automated, depth-resolved estimation of the attenuation coefficient from optical coherence tomography data.IEEE transactions on medical imaging, 34(12):2592–2602, 2015

    Gennifer T Smith, Nicholas Dwork, Daniel O’Connor, Uzair Sikora, Kristen L Lurie, John M Pauly, and Au- drey K Ellerbee. Automated, depth-resolved estimation of the attenuation coefficient from optical coherence tomography data.IEEE transactions on medical imaging, 34(12):2592–2602, 2015

    work page 2015

  11. [11]

    Multi- modal 3D cancer-mimicking optical phantom.Biomedical optics express, 7(2):648–662, 2016

    Gennifer T Smith, Kristen L Lurie, Dimitar V Zlatev, Joseph C Liao, and Audrey K Ellerbee Bowden. Multi- modal 3D cancer-mimicking optical phantom.Biomedical optics express, 7(2):648–662, 2016. 10 APREPRINT- APRIL17, 2026

    work page 2016

  12. [12]

    Alex Y Lin, Yifu Ding, Daniel J Vanselow, Spencer R Katz, Maksim A Yakovlev, Darin P Clark, David Mandrell, Jean E Copper, Damian B van Rossum, and Keith C Cheng. Rigid embedding of fixed and stained, whole, millimeter-scale specimens for section-free 3d histology by micro-computed tomography.Journal of visualized experiments: JoVE, (140):58293, 2018

    work page 2018

  13. [13]

    Cambridge university press, 2003

    Richard Hartley and Andrew Zisserman.Multiple view geometry in computer vision. Cambridge university press, 2003

    work page 2003

  14. [14]

    Prentice Hall PTR, 2004

    Gerard Medioni and Sing Bing Kang.Emerging topics in computer vision. Prentice Hall PTR, 2004

    work page 2004

  15. [15]

    SIAM, 2001

    Avinash C Kak and Malcolm Slaney.Principles of computerized tomographic imaging. SIAM, 2001. 11

    work page 2001