Pith. sign in

REVIEW 4 major objections 4 minor 52 references

Automated HER2 scoring with uncertainty quantification using lensfree holography and deep learning

T0 review · 4 major / 4 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read HER2 scoring of breast cancer tissue can be performed from lens-free holograms by a $980 device with deep learning, reaching 84.9% four-class accuracy and 94.8% binary accuracy on a blinded 412-core test set.

desk verdict First lensfree-HER2 scoring with genuine novelty, but the headline accuracy is conditional on an unquantified quality filter and the propagation-distance selection needs clarification. read the letter →

arxiv 2601.18219 v1 pith:ECGPZER4 submitted 2026-01-26 physics.med-ph cs.CVcs.LG

Che-Yung Shen , Xilin Yang , Yuzhu Li , Leon Lenk , Aydogan Ozcan This is my paper · ORCID
classification physics.med-phcs.CVcs.LG
keywords HER2scoringlensfreeholographydigitalpathologydeeplearninguncertaintyquantificationMonteCarlodropoutimmunohistochemistrybreastcancer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper establishes that HER2 scoring — the immunohistochemistry test that decides whether breast cancer patients qualify for trastuzumab — can be performed from lens-free holograms captured by a compact $980 device, with deep learning replacing the microscope. On a blinded test set of 412 unique tissue cores, the system achieves 84.9% four-class accuracy (0, 1+, 2+, 3+) and 94.8% binary accuracy, matching or slightly exceeding a brightfield whole-slide scanner run through the same neural pipeline. The authors show that the hologram's digitally back-propagated complex field, fed as a six-channel RGB tensor to an EfficientNet-B0 ensemble, carries enough diagnostic information to classify HER2 expression without a physical objective lens. They further show that Monte Carlo dropout uncertainty estimates allow the system to abstain on low-confidence predictions, correcting 30.4% of errors while discarding only 7.2% of correct ones.

What carries the argument

The load-bearing element is lensfree inline holography paired with digital back-propagation: a compact laser-illuminated CMOS sensor records interference patterns of the stained tissue without any objective lens, and the angular-spectrum method propagates the intensity back toward the sample plane to synthesize a complex field. This six-channel RGB complex field — not a reconstructed image — is the direct input to an EfficientNet-B0 classifier, bypassing phase retrieval. The paper's other machinery is a confidence-aware five-network ensemble with a positive-class override, and a Monte Carlo dropout uncertainty score (FOM = baseline confidence divided by the standard deviation of 256 stochast

What would settle it

Apply the trained ensemble to an uncurated, consecutive series of HER2 IHC slides (no exclusions for staining quality or artifacts) and compare the resulting four-class accuracy and rejection rate to the reported 84.9% and 30.4%; if the accuracy drops materially or the rejection rate balloons, the curated-test estimate does not generalize. Alternatively, quantify how many cores were discarded in data preparation — if it exceeds roughly 15% of all cores, the headline accuracy is optimistic.

Watch

Extended reading notes

Core claim

The central claim is that the diagnostic information needed for HER2 IHC classification survives in lens-free inline holograms, provided the network is trained on a digitally back-propagated complex-field representation. The system records diffraction patterns of DAB-stained tissue under red, green, and blue laser illumination on a monochrome CMOS sensor, then uses angular-spectrum propagation (at a fixed digital distance Z3=2.4 mm) to turn each hologram into a six-channel tensor (real and imaginary parts per color). An ensemble of five EfficientNet-B0 classifiers, with a HER2-positive-sensitive fusion rule, yields 80.8% four-class accuracy; applying class-specific Monte Carlo dropout uncert

Load-bearing premise

Everything rests on the assumption that the unreported number of tissue cores removed for poor staining or imaging artifacts is small enough not to distort accuracy, and that the pathologist-consensus labels are correct ground truth even though HER2 1+/2+ boundaries are notoriously subjective.

Editorial extensions

If this is right

  • If correct, the result means a ~$980, lens-free device can deliver HER2 scoring accuracy comparable to a commercial brightfield scanner, removing the optics cost and bulk from the bottleneck of digital pathology.
  • Clinical workflows could adopt the FOM threshold as an explicit 'abstain' rule: slides below threshold are re-read by a pathologist, which corrects 30.4% of misclassifications while discarding only 7.2% of correct ones.
  • The blue-channel-only result (75.0% four-class, 92.5% binary) indicates a single-wavelength version could cut hardware cost further while retaining most binary clinical value.
  • The strong dependence on digital propagation distance (80.8% at Z3=2.4 mm vs 56.3% at Z3=0) shows the complex-field representation is what carries the classification signal, not the raw hologram.
  • Because the system needs no mechanical focusing or high-precision alignment, it is compatible with decentralized, low-infrastructure settings — the stated motivation of the work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the 84.9% figure is measured on a curated dataset from which an unquantified number of poor-staining or artifact-laden cores were removed; a deployment on consecutive uncurated slides would likely lower accuracy, and the system's real tolerance to staining variability is untested.
  • Editorial inference: the near-parity of blue-only illumination with full RGB suggests that DAB phase contrast at short wavelengths is the dominant signal; this could be validated by acquiring the same cores under blue light and checking whether the accuracy gap between blue and RGB shrinks as staining intensity varies.
  • Editorial inference: the same uncertainty-abstention scheme could be transferred to other equivocal IHC biomarkers (e.g., PD-L1, Ki-67) or to other low-cost imaging platforms, since the FOM is model-agnostic.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. This manuscript reports an automated HER2 immunohistochemistry scoring pipeline that combines a compact lensfree holographic microscope with an EfficientNet-B0 classifier and Monte Carlo dropout uncertainty quantification. The system images ~1,250 mm² of tissue in ~15 minutes at an estimated hardware cost below $1,000 (excluding the tunable laser). On a blinded, patient-separated test set of 412 tissue microarray cores, the authors report 4-class HER2 testing accuracy of 84.9% and binary accuracy of 94.8% after FOM-based rejection of low-confidence predictions, with a 30.4% correction rate and 7.2% loss rate. The paper also compares performance against a brightfield whole-slide scanner baseline and analyzes monochrome (single-wavelength) holography and the effect of digital propagation distance.

Significance. If the reported accuracies hold on unselected clinical material, this would be a meaningful demonstration that a low-cost, lensfree imaging platform can approach the scoring performance of conventional brightfield scanners, with potential value for resource-limited settings. The use of a blinded test set with patient-level separation, the inclusion of an uncertainty-based rejection mechanism, and the direct comparison with a brightfield scanner are notable strengths. However, the headline accuracy is conditional on an unquantified quality-control filter and on threshold choices that require clearer reporting; these gaps currently limit the strength of the claim.

major comments (4)
  1. [Materials and Methods, 'Data preparation and labeling'] The paper states that 'We discarded tissue cores with poor staining quality ... and cores affected by imaging artifacts,' but it never reports how many cores were discarded. With 15 TMA slides of ~100–150 cores each, the raw pool is 1,500–2,250 cores, while only 1,273 + 412 = 1,685 cores remain. If the discarded fraction is substantial, the 84.9% test accuracy is measured on a selected subset and overstates performance on unselected slides. Please report the exact exclusion count/rate and, ideally, a sensitivity analysis or a description of the excluded cases. This is the most load-bearing missing number for the central claim.
  2. [Materials and Methods, 'Uncertainty quantification and filtering using MC dropout' and Results, 'Uncertainty quantificat] The FOM thresholds [12.1, 16.2, 11.6, 21.2] were 'determined on the validation set to achieve >30% correction rate,' and the test-set correction rate is reported as 30.4%. Thus the headline correction rate is the test realization of the tuning target. Although the thresholds were apparently applied blindly to the test set, this makes the 30.4% figure an expected consequence of the selection rule rather than an independent discovery. Please report validation-set correction/loss rates, describe the threshold-selection procedure more explicitly (e.g., the grid searched, the selection criterion), and provide confidence intervals for test-set correction and loss rates.
  3. [Results, 'Impact of the digital propagation distance on the automated HER2 scoring performance'] Figure 6 shows that Z3 = 2.4 mm was chosen as the operating point and that it gives the highest test accuracy among the tested distances. The manuscript does not state whether Z3 was selected using a validation set or whether it was selected by inspecting test-set performance. If Z3 was tuned on the test set, the reported 84.9%/94.8% accuracies are optimistically biased, and the comparison with Z3 = 0 mm or Z3 = 2.84 mm is not a valid out-of-sample comparison. Please clarify the selection procedure for Z3 and, if it was test-set-selected, report a validation-based or nested evaluation.
  4. [Materials and Methods, 'Data preparation and labeling' and Discussion] The ground-truth labels are described as 'pathologist-verified ... by at least 3 certified pathologists,' but no consensus rule is given (e.g., majority vote, adjudication, or exact agreement requirement). Given the known inter-observer variability in HER2 IHC scoring, especially at the 1+/2+ boundary (ref. 40), the label noise could materially affect the reported class accuracies, particularly the 2+ class (68.0% before uncertainty rejection). Please specify the labeling protocol and, if possible, report inter-observer agreement or a label-noise sensitivity analysis.
minor comments (4)
  1. [Results, 'Automated HER2 scoring using lensfree holography' and Fig. S3/S4] The brightfield comparison reports 82.8%/84.3% 4-class accuracies, but no confidence intervals or significance tests are provided. Given n=412, the difference between 84.9% (lensfree) and 82.8% (brightfield) is within sampling noise. Please add confidence intervals or a paired test to support the 'comparable' claim.
  2. [Results, 'Uncertainty quantification for HER2 scoring'] The text reports rejection rates for correctly classified cores by class, but not the class-wise acceptance counts for misclassified cores. Adding the accepted/rejected confusion-matrix counts in a supplementary table would help readers assess whether the rejection rule is clinically sensible.
  3. [Table 1] The Basler CMOS sensor is listed with a unit price of $584 but a total of $689. Please reconcile this discrepancy or clarify whether the total includes additional components (e.g., cables or lens mounts).
  4. [General] A few typographical issues: the abstract states '~1,250 mm^2' while the Methods say '~12.5 cm²' (equivalent, but should be consistent); Fig. 6 caption appears truncated in the manuscript text; and Eq. (1) uses subscripts/superscripts that are hard to parse. Please correct these.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central results are direct empirical evaluations on a held-out test set, with disclosed validation-based threshold selection.

full rationale

The paper's central claim—84.9% four-class and 94.8% binary HER2 accuracy from lensfree holography with MC-dropout-based rejection—is an empirical result on 412 blinded patient cores separated from training/validation. The ensemble and uncertainty-quantification pipelines are not defined in terms of the test outcome. The FOM thresholds were selected on the validation set to target a >30% correction rate, and the paper explicitly states that these thresholds were then 'blindly applied to our test images.' Reporting the resulting 30.4% test correction rate is standard held-out evaluation after hyperparameter selection; it is not a fitted input renamed as a prediction. The comparison to brightfield microscopy uses a prior dataset from the same group (ref. 34) as an external benchmark, but the lensfree result is not derived from that dataset and the benchmark does not bear the load of the claim. The unquantified exclusion of poor-quality/artifactual cores is a legitimate external-validity concern, but it is a data-selection issue, not a circular derivation. No equation in the paper reduces to its own inputs, and no load-bearing premise rests solely on a self-citation. Therefore the derivation chain is self-contained with respect to circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities; its contribution is the application of known optics (lensfree holography) and known ML (EfficientNet ensemble, MC dropout) to HER2 IHC scoring. The honest inventory is the five fitted operating parameters (Z3, FOM thresholds, R_drop, N_MC, M) that shape the headline numbers, plus domain assumptions about label quality, artifact exclusion, patient separation, and transfer learning that the accuracy claims rest on.

free parameters (5)
  • Class-specific FOM thresholds = [12.1, 16.2, 11.6, 21.2] for HER2 [0,1+,2+,3+]; second set [11.5,12.0,10.2,16.3]
    Tuned on the validation set to achieve >30% correction rate while maximizing coverage; the reported 30.4% correction rate on test is the realization of this tuning target.
  • Digital propagation distance Z3 = 2.4 mm (physical sample-to-sensor distance Z2 = 2.84 mm)
    Z3=2.4 mm is the accuracy peak of Fig 6, whose caption reports 'HER2 testing accuracy' vs Z3; all main results use this value, but the paper does not state that it was fixed on a validation split rather than selected on test performance.
  • MC dropout rate R_drop = 0.70
    Hand-chosen, unusually high; it shapes both the FOM statistics and the accepted-prediction set.
  • MC sampling count N_MC = 256
    Hand-chosen; directly affects sigma_S in the FOM, hence which predictions are rejected.
  • Ensemble size M = 5
    Hand-chosen; the benefit of the HER2-positive-sensitive fusion rule in Eq. 1 depends on M.
assumptions (6)
  • domain assumption Angular-spectrum back-propagation of raw intensity, without phase retrieval, produces a six-channel complex-field representation sufficient for HER2 classification.
    The paper explicitly disclaims that the field 'does not necessarily reconstruct the actual complex field representation' (Results, workflow) yet the entire classification input depends on this representation being discriminative.
  • domain assumption The pathologist-verified ground-truth spreadsheet (verified by at least 3 certified pathologists) is correct.
    Methods, Data preparation: the consensus rule is unreported, and HER2 IHC inter-observer variability is itself substantial (ref 40), especially at the 1+/2+ boundary.
  • domain assumption Excluding cores with poor staining, background, folds, or imaging artifacts does not bias the accuracy estimate for the intended use population.
    Methods, Data preparation: the discarded-core count is unreported; the headline accuracy is conditional on this filter.
  • domain assumption Train/test cores come from disjoint patients.
    Methods, Data preparation: 'strict patient-level separation' is claimed, but TMA slides typically contain multiple cores per patient, and no check is described that duplicate cores from one patient do not straddle the train/test split.
  • domain assumption ImageNet-pretrained EfficientNet-B0 weights transfer to six-channel holographic inputs.
    Methods, Architecture: the first convolution is modified 3->6 channels and ImageNet-initialized; the 78.2% single-model baseline depends on this transfer.
  • standard math Binomial accuracy estimates from n=412 test cores are stable enough to support cross-configuration comparisons.
    The binomial SE on 84.9% is about 1.8%; the paper reports no confidence intervals, yet prose treats differences such as 80.8% vs 82.8% as meaningful.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Automated HER2 scoring with uncertainty quantification using lensfree holography and deep learning." pith.science (2026). https://pith.science/paper/ECGPZER4

@misc{pith2026260118219,
  author       = {Pith},
  title        = {Pith review of: Automated HER2 scoring with uncertainty quantification using lensfree holography and deep learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ECGPZER4}},
  note         = {Machine review of arXiv:2601.18219}
}
read the original abstract

Accurate assessment of human epidermal growth factor receptor 2 (HER2) expression is critical for breast cancer diagnosis, prognosis, and therapy selection; yet, most existing digital HER2 scoring methods rely on bulky and expensive optical systems. Here, we present a compact and cost-effective lensfree holography platform integrated with deep learning for automated HER2 scoring of immunohistochemically stained breast tissue sections. The system captures lensfree diffraction patterns of stained HER2 tissue sections under RGB laser illumination and acquires complex field information over a sample area of ~1,250 mm^2 at an effective throughput of ~84 mm^2 per minute. To enhance diagnostic reliability, we incorporated an uncertainty quantification strategy based on Bayesian Monte Carlo dropout, which provides autonomous uncertainty estimates for each prediction and supports reliable, robust HER2 scoring, with an overall correction rate of 30.4%. Using a blinded test set of 412 unique tissue samples, our approach achieved a testing accuracy of 84.9% for 4-class (0, 1+, 2+, 3+) HER2 classification and 94.8% for binary (0/1+ vs. 2+/3+) HER2 scoring with uncertainty quantification. Overall, this lensfree holography approach provides a practical pathway toward portable, high-throughput, and cost-effective HER2 scoring, particularly suited for resource-limited settings, where traditional digital pathology infrastructure is unavailable.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 1 canonical work pages

  1. [1]

    P. W. Hamilton, P. Bankhead, Y. Wang, R. Hutchinson, D. Kieran, D. G. McArt, J. James, M. Salto- Tellez, Digital pathology and image analysis in tissue biomarker research. Methods 70, 59–73 (2014)

    2014

  2. [2]

    S. W. Jahn, M. Plass, F. Moinfar, Digital Pathology: Advantages, Limitations and Emerging Perspectives. J. Clin. Med. 9, 3697 (2020)

    2020

  3. [3]

    Rajaganesan, R

    S. Rajaganesan, R. Kumar, V. Rao, T. Pai, N. Mittal, A. Sahay, S. Menon, S. Desai, Comparative Assessment of Digital Pathology Systems for Primary Diagnosis. J. Pathol. Inform. 12, 25 (2021)

    2021

  4. [4]

    Farahani, A

    N. Farahani, A. V. Parwani, L. Pantanowitz, Whole slide imaging in pathology: advantages, limitations, and emerging perspectives. Pathol. Lab. Med. Int. 7, 23–33 (2015)

    2015

  5. [5]

    Kusta, M

    O. Kusta, M. Bearman, R. Gorur, T. Risør, J. B. Brodersen, K. Hoeyer, Speed, accuracy, and efficiency: The promises and practices of digitization in pathology. Soc. Sci. Med. 345, 116650 (2024)

    2024

  6. [6]

    M. J. Fanous, P. Casteleiro Costa, Ç. Işıl, L. Huang, A. Ozcan, Neural network-based processing and reconstruction of compromised biophotonic image data. Light Sci. Appl. 13, 231 (2024). 13

    2024

  7. [7]

    Bishara, H

    W. Bishara, H. Zhu, A. Ozcan, Holographic opto-fluidic microscopy. Opt. Express 18, 27499–27510 (2010)

    2010

  8. [8]

    Tseng, O

    D. Tseng, O. Mudanyali, C. Oztoprak, S. O. Isikman, I. Sencan, O. Yaglidere, A. Ozcan, Lensfree microscopy on a cellphone. Lab. Chip 10, 1787–1792 (2010)

    2010

Show all 52 references
  1. [10]

    Greenbaum, W

    A. Greenbaum, W. Luo, T. -W. Su, Z. Göröcs, L. Xue, S. O. Isikman, A. F. Coskun, O. Mudanyali, A. Ozcan, Imaging without lenses: achievements and remaining challenges of wide -field on -chip microscopy. Nat. Methods 9, 889–895 (2012)

    2012

  2. [11]

    Greenbaum, Y

    A. Greenbaum, Y. Zhang, A. Feizi, P. -L. Chung, W. Luo, S. R. Kandukuri, A. Ozcan, Wide -field computational imaging of pathology slides using lens -free on-chip microscopy. Sci. Transl. Med. 6, 267ra175-267ra175 (2014)

    2014

  3. [12]

    Ozcan, E

    A. Ozcan, E. McLeod, Lensless Imaging and Sensing. Annu. Rev. Biomed. Eng. 18, 77–102 (2016)

    2016

  4. [13]

    T. Liu, Y. Li, H. C. Koydemir, Y. Zhang, E. Yang, M. Eryilmaz, H. Wang, J. Li, B. Bai, G. Ma, A. Ozcan, Rapid and stain-free quantification of viral plaque via lens-free holography and deep learning. Nat. Biomed. Eng. 7, 1040–1052 (2023)

    2023

  5. [14]

    Y. Wu, A. Ozcan, Lensless digital holographic microscopy and its applications in biomedicine and environmental monitoring. Methods 136, 4–16 (2018)

    2018

  6. [15]

    Madabhushi, G

    A. Madabhushi, G. Lee, Image analysis and machine learning in digital pathology: Challenges and opportunities. Med. Image Anal. 33, 170–175 (2016)

    2016

  7. [16]

    Rivenson, Z

    Y. Rivenson, Z. Göröcs, H. Günaydin, Y. Zhang, H. Wang, A. Ozcan, Deep learning microscopy. Optica 4, 1437–1443 (2017)

    2017

  8. [17]

    M. K. K. Niazi, A. V. Parwani, M. N. Gurcan, Digital pathology and artificial intelligence. Lancet Oncol. 20, e253–e261 (2019)

    2019

  9. [18]

    Rivenson, T

    Y. Rivenson, T. Liu, Z. Wei, Y. Zhang, K. de Haan, A. Ozcan, PhaseStain: the digital staining of label-free quantitative phase microscopy images using deep learning. Light Sci. Appl. 8, 23 (2019)

    2019

  10. [19]

    Rivenson, A

    Y. Rivenson, A. Ozcan, Deep learning accelerates whole slide imaging for next -generation digital pathology applications. Light Sci. Appl. 11, 300 (2022)

    2022

  11. [20]

    B. Bai, H. Wang, Y. Li, K. de Haan, F. Colonnese, Y. Wan, J. Zuo, N. B. Doan, X. Zhang, Y. Zhang, J. Li, X. Yang, W. Dong, M. A. Darrow, E. Kamangar, H. S. Lee, Y. Rivenson, A. Ozcan, Label-Free Virtual HER2 Immunohistochemical Staining of Breast Tissue u sing Deep Learning. B...

    2022

  12. [21]

    X. Wang, S. Yang, J. Zhang, M. Wang, J. Zhang, W. Yang, J. Huang, X. Han, Transformer -based unsupervised contrastive learning for histopathological image classification. Med. Image Anal. 81, 102559 (2022). 14

    2022

  13. [22]

    H. Chen, L. Huang, T. Liu, A. Ozcan, Fourier Imager Network (FIN): A deep neural network for hologram reconstruction with superior external generalization. Light Sci. Appl. 11, 254 (2022)

    2022

  14. [23]

    B. Bai, X. Yang, Y. Li, Y. Zhang, N. Pillar, A. Ozcan, Deep learning -enabled virtual histological staining of biological samples. Light Sci. Appl. 12, 57 (2023)

    2023

  15. [24]

    H. Chen, L. Huang, T. Liu, A. Ozcan, eFIN: Enhanced Fourier Imager Network for Generalizable Autofocusing and Pixel Super -Resolution in Holographic Imaging. IEEE J. Sel. Top. Quantum Electron. 29, 1–10 (2023)

    2023

  16. [25]

    R. J. Chen, T. Ding, M. Y. Lu, D. F. K. Williamson, G. Jaume, A. H. Song, B. Chen, A. Zhang, D. Shao, M. Shaban, M. Williams, L. Oldenburg, L. L. Weishaupt, J. J. Wang, A. Vaidya, L. P. Le, G. Gerber, S. Sahai, W. Williams, F. Mahmood, Towards a general -purpose foundation mod...

    2024

  17. [26]

    Vorontsov, A

    E. Vorontsov, A. Bozkurt, A. Casson, G. Shaikovski, M. Zelechowski, K. Severson, E. Zimmermann, J. Hall, N. Tenenholtz, N. Fusi, E. Yang, P. Mathieu, A. van Eck, D. Lee, J. Viret, E. Robert, Y. K. Wang, J. D. Kunz, M. C. H. Lee, J. H. Bernhard, R. A. Godr ich, G. Oakley, E. Mi...

    2024

  18. [27]

    Zhang, L

    Y. Zhang, L. Huang, N. Pillar, Y. Li, H. Chen, A. Ozcan, Pixel super -resolved virtual staining of label-free tissue using diffusion models. Nat. Commun. 16, 5016 (2025)

    2025

  19. [28]

    Choudhury, J

    D. Choudhury, J. M. Dolezal, E. Dyer, S. Kochanny, S. Ramesh, F. M. Howard, J. R. Margalus, A. Schroeder, J. Schulte, M. C. Garassino, J. N. Kather, A. T. Pearson, Developing a low -cost, open- source, locally manufactured workstation and computational pipe line for automated ...

    2024

  20. [29]

    M. J. Fanous, C. M. Seybold, H. Chen, N. Pillar, A. Ozcan, BlurryScope enables compact, cost - effective scanning microscopy for HER2 scoring using deep learning on blurry images. Npj Digit. Med. 8, 506 (2025)

    2025

  21. [30]

    Loibl, L

    S. Loibl, L. Gianni, HER2-positive breast cancer. The Lancet 389, 2415–2429 (2017)

    2017

  22. [31]

    Zubair, S

    M. Zubair, S. Wang, N. Ali, Advanced Approaches to Breast Cancer Classification and Diagnosis. Front. Pharmacol. 11 (2021)

    2021

  23. [32]

    Smolarz, A

    B. Smolarz, A. Z. Nowak, H. Romanowicz, Breast Cancer —Epidemiology, Classification, Pathogenesis and Treatment (Review of Literature). Cancers 14, 2569 (2022)

    2022

  24. [33]

    S. Wu, M. Yue, J. Zhang, X. Li, Z. Li, H. Zhang, X. Wang, X. Han, L. Cai, J. Shang, Z. Jia, X. Wang, J. Li, Y. Liu, The Role of Artificial Intelligence in Accurate Interpretation of HER2 Immunohistochemical Scores 0 and 1+ in Breast Cancer. Mod. Pathol. 36, 100054 (2023)

    2023

  25. [34]

    S. Y. Selcuk, X. Yang, B. Bai, Y. Zhang, Y. Li, M. Aydin, A. F. Unal, A. Gomatam, Z. Guo, D. M. Angus, G. Kolodney, K. Atlan, T. K. Haran, N. Pillar, A. Ozcan, Automated HER2 Scoring in Breast Cancer Images Using Deep Learning and Pyramid Sampling. BME Front. 5, 0048 (2024). 15

    2024

  26. [35]

    Matsushima, T

    K. Matsushima, T. Shimobaba, Band -Limited Angular Spectrum Method for Numerical Simulation of Free-Space Propagation in Far and Near Fields. Opt. Express 17, 19662–19673 (2009)

    2009

  27. [36]

    Mudanyali, D

    O. Mudanyali, D. Tseng, C. Oh, S. O. Isikman, I. Sencan, W. Bishara, C. Oztoprak, S. Seo, B. Khademhosseini, A. Ozcan, Compact, lightweight and cost -effective microscope based on lensless incoherent holography for telemedicine applications. Lab. Chip 10, 1417–1428 (2010)

    2010

  28. [37]

    Efficientnet: Rethinking model scaling for convolutional neural networks

    M. Tan, Q. Le, “Efficientnet: Rethinking model scaling for convolutional neural networks” in International Conference on Machine Learning (PMLR, 2019; https://proceedings.mlr.press/v97/tan19a.html?ref=ji), pp. 6105–6114

    2019

  29. [38]

    Y. Gal, Z. Ghahramani, Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning. arXiv arXiv:1506.02142 [Preprint] (2016). https://doi.org/10.48550/arXiv.1506.02142

    2016 doi arXiv

  30. [39]

    Goncharov, R

    A. Goncharov, R. Ghosh, H. -A. Joung, D. D. Carlo, A. Ozcan, Autonomous Uncertainty Quantification for Computational Point -of-care Sensors. arXiv arXiv:2512.21335 [Preprint] (2025). https://doi.org/10.48550/arXiv.2512.21335

    2025 doi arXiv

  31. [40]

    T. A. Thomson, M. M. Hayes, J. J. Spinelli, E. Hilland, C. Sawrenko, D. Phillips, B. Dupuis, R. L. Parker, HER -2/neu in breast cancer: interobserver variability and performance of immunohistochemistry with 4 antibodies compared with fluorescent in situ hy bridization. Mod. Pa...

    2001

  32. [41]

    X. Lin, Y. Rivenson, N. T. Yardimci, M. Veli, Y. Luo, M. Jarrahi, A. Ozcan, All -optical machine learning using diffractive deep neural networks. Science 361, 1004–1008 (2018)

    2018

  33. [42]

    Y. Luo, D. Mengu, N. T. Yardimci, Y. Rivenson, M. Veli, M. Jarrahi, A. Ozcan, Design of task - specific optical systems using broadband diffractive neural networks. Light Sci. Appl. 8, 112 (2019)

    2019

  34. [43]

    M. Veli, D. Mengu, N. T. Yardimci, Y. Luo, J. Li, Y. Rivenson, M. Jarrahi, A. Ozcan, Terahertz pulse shaping using diffractive surfaces. Nat. Commun. 12, 37 (2021)

    2021

  35. [44]

    Ç. Işıl, D. Mengu, Y. Zhao, A. Tabassum, J. Li, Y. Luo, M. Jarrahi, A. Ozcan, Super-resolution image display using diffractive decoders. Sci. Adv. 8, eadd3433 (2022)

    2022

  36. [45]

    J. Li, T. Gan, Y. Zhao, B. Bai, C.-Y. Shen, S. Sun, M. Jarrahi, A. Ozcan, Unidirectional imaging using deep learning–designed materials. Sci. Adv. 9, eadg1505 (2023)

    2023

  37. [46]

    C.-Y. Shen, J. Li, T. Gan, Y. Li, M. Jarrahi, A. Ozcan, All-optical phase conjugation using diffractive wavefront processing. Nat. Commun. 15, 4989 (2024)

    2024

  38. [47]

    J. Li, D. Mengu, Y. Luo, Y. Rivenson, A. Ozcan, Class -specific differential detection in diffractive optical neural networks improves inference accuracy. Adv. Photonics 1, 046001 (2019)

    2019

  39. [48]

    J. Li, D. Mengu, N. T. Yardimci, Y. Luo, X. Li, M. Veli, Y. Rivenson, M. Jarrahi, A. Ozcan, Spectrally encoded single-pixel machine vision using diffractive networks. Sci. Adv. 7, eabd7690 (2021)

    2021

  40. [49]

    M. S. S. Rahman, J. Li, D. Mengu, Y. Rivenson, A. Ozcan, Ensemble learning of diffractive optical networks. Light Sci. Appl. 10, 14 (2021). 16

    2021

  41. [50]

    B. Bai, Y. Li, Y. Luo, X. Li, E. Çetintaş, M. Jarrahi, A. Ozcan, All-optical image classification through unknown random diffusers using a single-pixel diffractive network. Light Sci. Appl. 12, 69 (2023)

    2023

  42. [51]

    https://www.tissuearray.com/tissue - arrays

    Human Paraffin Embedded Tissue Array, TissueArray.Com. https://www.tissuearray.com/tissue - arrays

  43. [52]

    B. Zoph, Q. V. Le, Neural Architecture Search with Reinforcement Learning. arXiv arXiv:1611.01578 [Preprint] (2017). https://doi.org/10.48550/arXiv.1611.01578

    2017 doi arXiv

  44. [53]

    Loshchilov, F

    I. Loshchilov, F. Hutter, Decoupled Weight Decay Regularization. arXiv arXiv:1711.05101 [Preprint] (2019). https://doi.org/10.48550/arXiv.1711.05101. 17 Fig. 1. Schematics of the customized lensfree holography system and the digital HER2 classification model. (A) Photo of the ...

    2019 doi arXiv

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.