Pith. sign in

REVIEW 3 major objections 5 minor 64 references

A scale-adaptive vision model trained on C. elegans confocal images links neuron morphology to dopamine-dependent behavior and flags the benzimidazole ring as a determinant of dopaminergic neurotoxicity.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 03:20 UTC pith:VPNFXZXM

load-bearing objection A substantial new benchmark and a plausible method; the benzimidazole-as-determinant claim is a hypothesis, not a result. the 3 major comments →

arxiv 2607.23183 v1 pith:VPNFXZXM submitted 2026-07-25 cs.CE cs.CV

A Scale-adaptive Vision Model Links C. elegans Neuronal Morphology to Behavior for Neurotoxicity Assessment

Haochao Ying , Shenchong Lv , Yutao Sun , Zijian Tu , Xufeng Jin , Yuyang Xu , Yizhe Wang , Wei Yang
show 3 more authors
Xiaomin Yue Jian Wu Peilin Yu
This is my paper
classification cs.CE cs.CV
keywords neurotoxicity assessmentCaenorhabditis elegansdopaminergic neuronsmasked image modelingself-supervised learningconfocal microscopymorphology-behavior linkbenzimidazole
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper tries to show that a dedicated vision model can turn confocal images of C. elegans dopaminergic neurons into quantitative, behaviorally meaningful neurotoxicity readouts. It introduces CeNeuMorph, a 27,117-image benchmark, and a scale-adaptive masked image modeling method that jointly varies resolution and patch size so that both fine lesions and gross deformations are learned. The authors report that fusing visual features with morphological descriptors predicts dopamine-dependent behavioral deficits, with R²=0.498 for the difference between on-food and off-food locomotion. They then screen 180 agrochemicals and identify the benzimidazole ring as a previously unrecognized structural determinant of dopaminergic neurotoxicity. If right, the work offers a scalable, objective alternative to hand-scored assays for environmental neurotoxicity assessment.

Core claim

The paper's central claim is that the information needed to assess dopaminergic neurotoxicity is present in confocal images of C. elegans cephalic neurons and can be extracted by a vision model built for the modality. To show this, the authors construct CeNeuMorph, a 27,117-image multi-grained benchmark with annotations for breakage, arborization, bending, beading, dendrite detection, and CEP/ADE soma segmentation, and pre-train a scale-adaptive masked autoencoder that randomly varies resolution and patch size under a fixed token budget. On this benchmark, the model reaches 88.44% accuracy on neuronal breakage detection and outperforms generalist and biomedical foundation models across class

What carries the argument

The load-bearing mechanism is scale-adaptive masked image modeling. Standard masked autoencoders operate on fixed patch grids, which either miss few-pixel dendritic beading or require prohibitive resolution; this method instead samples input resolution and patch size jointly during pre-training and truncates the token sequence to a fixed budget, so the encoder must reconstruct global neuron structure from a constant number of visible tokens. At fine-tuning, two patch scales share one encoder and are fused via exchanged classification tokens and a feature pyramid, letting local lesions and global deformations influence the same representation. The morphology-to-behavior link is carried by an

Load-bearing premise

The load-bearing assumption is that the benzimidazole ring—not other physicochemical properties, exposure kinetics, or the agrochemical-versus-pharmaceutical context of the test compounds—is what drives the observed dopaminergic neurotoxicity.

What would settle it

Screen a matched set of benzimidazole and isosteric non-benzimidazole analogues, such as indole- or pyridine-based compounds, with similar lipophilicity, solubility, and identical exposure concentration; if the non-benzimidazole isosteres produce the same dopaminergic injury and behavioral deficits, the ring is not the determinant.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Neuronal morphology assays in C. elegans can shift from hand-scored, subjective readouts to automated, reproducible image-based quantification at scale.
  • Morphological features can serve as a screening proxy for dopamine-dependent functional decline, so structurally damaged neurons can flag chemicals for behavioral follow-up.
  • The 180-compound screen identifies benzimidazole-containing agrochemicals, rather than clinical benzimidazole drugs, as the primary dopaminergic injury risk within that class, pointing to a safe-by-design substituent strategy.
  • The scale-adaptive pre-training recipe is transferable to other sparse fluorescence imaging domains where lesions span widely different spatial scales.

Where Pith is reading between the lines

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

  • Beyond the paper: the benzimidazole conclusion is directly testable by synthesizing isosteric ring replacements, such as indole or pyridine analogues, with matched lipophilicity and identical exposure; the current evidence is class-level association, not causation.
  • Beyond the paper: the moderate R² values for behavioral prediction (0.300–0.498) suggest the model is best used as a prioritization filter that narrows the chemical space for costly behavioral or mammalian validation.
  • Beyond the paper: applying the same scale-adaptive pipeline to whole-nervous-system imaging of all 302 C. elegans neurons, or to other transparent model organisms, could turn morphology-based neurotoxicity screening into a general environmental-health tool.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces CeNeuMorph, a 27,117-image confocal benchmark for C. elegans dopaminergic neuron morphology, and a scale-adaptive masked image modeling (MIM) framework that jointly samples resolutions and patch sizes under a fixed token budget. The authors report that the approach outperforms generalist and biomedical foundation models on classification, semantic segmentation, object detection, and instance segmentation tasks. They further fuse the learned visual features with TabPFN-encoded morphological descriptors to predict basal slowing response (BSR) behavior, reporting R²=0.498 on the fon−foff task. Applying the model to screen 180 agrochemicals, they identify carbendazim as the most neurotoxic compound and, comparing 29 benzimidazole-containing compounds against the rest of the panel, conclude that the benzimidazole ring is a structural determinant of dopaminergic neurotoxicity.

Significance. The CeNeuMorph benchmark and the scale-adaptive MIM pre-training strategy are potentially valuable contributions to automated neurotoxicity screening. The paper provides open code and data, and the reproducibility statements are a clear strength. If the performance claims were fully supported, the method would offer a practical alternative to manual morphological scoring. However, the headline biological conclusion—that the benzimidazole moiety is a determinant of dopaminergic neurotoxicity—is not yet secured because the comparison is confounded, and the benchmark claim of uniform superiority over all baselines is overstated. The technical core is sound and the dataset is a useful resource, but the causal and comparative claims need re-scoping or additional evidence.

major comments (3)
  1. [Abstract; Table 1] The abstract claims the model 'surpasses both generalist and biomedical foundation models across classification, segmentation and detection tasks.' Table 1 does not support this at the per-metric level: on Dendrite object detection, MixMAE achieves 91.56±1.10 versus Ours-480 at 89.66±1.15; on Arborization, DINOv2 achieves 83.60±0.95 versus Ours-480 at 83.36±0.86; and several differences are within one standard deviation. The claim should be revised to 'most tasks' or supported by paired statistical tests across the five folds. This is load-bearing because it is the central benchmarking assertion.
  2. [Section 2.4, Fig. 4k, Fig. 5; Discussion] The conclusion that the benzimidazole ring is a structural determinant of dopaminergic neurotoxicity is based on comparing 29 benzimidazole-containing compounds with the other screened compounds. These sets are not matched for molecular weight, lipophilicity, exposure kinetics, or bioactivity class, and the carbamate comparison in Fig. 4j is a class-level contrast rather than a structural knockout. Fig. A4 shows that clinical benzimidazoles are not prominently neurotoxic, so the ring is neither sufficient nor demonstrated necessary. Direct structural analogs, dose-response studies, or mechanistic perturbations are needed to support the causal language. As written, the evidence supports an enrichment of benzimidazole-containing compounds among hits, not a structural determinant.
  3. [Methods 4.5; Fig. 3b] The composite BSR injury score is defined as a 1:1:2 weighted sum of the three task predictions (Methods 4.5), but no sensitivity analysis is provided for these weights. Because the screening ranking in Fig. 3b is derived entirely from this composite, the ranking may change under alternative plausible weightings. Additionally, the morphological descriptors used in the fusion model include the same fine-tuned model's binary classification outputs, and the regression has R²=0.498 on Task 2; the screening results therefore inherit the predictive error of that model. Reporting confidence intervals or an uncertainty-aware ranking for the BSR scores would make the screening claims more robust.
minor comments (5)
  1. [Section 2.4 title] Typo: 'dopaminergic neuroroxicity' should be 'dopaminergic neurotoxicity'.
  2. [References [5] and [6]] References [5] (Lai et al.) and [6] (Kaletta et al.) have the same title. One of them is likely mis-cited; please correct.
  3. [Table 1] The row label 'DaviT' for MedImageInsight is inconsistent with the other architecture labels (e.g., 'ViT-B'). Please clarify the backbone architecture used.
  4. [Fig. 3b] The compound 'Faltan' appears in the ranked list; this is likely 'Folpet'. Also 'Thiophanate-Methyl' is listed but not discussed in the text.
  5. [Data Availability] The phrase 'horizontally indexed for academic use' is unclear. Please specify the license and access procedure.

Circularity Check

0 steps flagged

No significant circularity: the morphology–behavior link is an empirical regression fit, and the chemical-screen conclusion is an external-data association rather than a derivation from the model's own definitions.

full rationale

The paper's central predictive claim (R^2=0.498 in Task 2) is produced by a fitted regression/classification model in Section 2.2 and Methods 4.5; the behavioral labels (fon, foff, 0/0 status) are independently measured and are not reconstructed from the model's own outputs by construction. The morphological descriptors include predictions of the fine-tuned vision model, but these are used as features for a separate behavioral prediction task, so no target quantity is defined in terms of the predictor. The scale-adaptive MIM pre-training is initialized from public MAE weights [15] and benchmarked against external baselines and foundation models (Table 1); it does not rely on a same-author uniqueness theorem or ansatz smuggled through self-citation. The benzimidazole conclusion in Section 2.4 is an empirical comparison of 29 benzimidazole-containing compounds against the rest of the screened panel; whether the comparison is adequately matched is a causal-inference and confounding concern, not a circularity. The 1:1:2 composite BSR score is a hand-chosen aggregation, but the screening ranking is an output of the fitted model, not a quantity equal to its inputs by definition. The paper itself concedes that clinical benzimidazoles did not show prominent morphological alterations (Fig. A4), undercutting any reading that the ring alone determines toxicity; this is an internal validity caveat, not a circularity. No self-citation is load-bearing in the derivation chain. Therefore no circular step satisfying the quote-and-reduction standard was found.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 0 invented entities

The work does not introduce new physical entities or mathematical axioms. The main extra burden is in hand-set hyperparameters and domain assumptions about what the images and behavioral readouts actually measure.

free parameters (4)
  • BSR composite score weights = 1:1:2 for Tasks 1, 2, 3
    The integrated BSR injury score used for compound ranking and the central benzimidazole finding weights Task 3 double; this choice is ad hoc and not justified by data.
  • Masking token budget Lkeep = 180
    The constant token budget is set to 180 visible tokens; this is a design choice that controls compute and may affect representation quality.
  • Patch size and resolution sampling ranges = patch size [4,32], resolution [240,480] (or [112,224])
    These ranges are chosen by hand and are not optimized or ablated across alternatives.
  • Dual-scale fine-tuning patch sizes = 16 and 32
    The downstream fusion uses fixed fine and coarse patch sizes; other pairings are not tested.
axioms (4)
  • domain assumption Expert manual annotations of confocal images are correct and consistent ground truth.
    All downstream task labels (break, arborization, bend, bead, dendrite, soma) come from two experts' annotations; errors propagate through pretraining, fine-tuning, and screening.
  • domain assumption Basal Slowing Response is a dopamine-dependent behavioral readout captured by the two-point locomotion measurement (fon/foff and fon−foff).
    The paper relies on this to frame Task 1/2/3 as functional dopamine readouts; the Methods state only that BSR is a behavioral hallmark, citing prior work.
  • domain assumption 2D maximum intensity projections preserve the morphology features relevant for neurotoxicity.
    The dataset is built from Z-projections; the paper acknowledges this loses 3D context in the Discussion, yet the central claims are based on these 2D images.
  • domain assumption The 180-compound library, after excluding known PD-associated and nematicidal compounds, is a representative panel for discovering novel dopaminergic neurotoxicants.
    The filtering step in Methods 4.1 is a selection that could bias the screen toward or away from certain chemical classes, including benzimidazoles.

pith-pipeline@v1.3.0-alltime-deepseek · 19042 in / 8670 out tokens · 84613 ms · 2026-08-01T03:20:56.199825+00:00 · methodology

0 comments
read the original abstract

Neurological disorders are a leading cause of global disability and are increasingly linked to environmental chemical exposures. Yet neurotoxicity assessment still relies on hand-scored morphological readouts that are subjective and poorly predictive of behavioral outcomes. Caenorhabditis elegans provides a genetically tractable, 3R-compliant alternative, but quantifying neuronal phenotypes from confocal microscopy at scale remains computationally challenging: existing vision foundation models, trained on natural or radiological images, cannot resolve the sparse signals and multi-scale lesions of neuronal imaging. Here, we introduce a dedicated self-supervised vision model for C. elegans dopaminergic neurons, together with CeNeuMorph, a multi-grained confocal benchmark of 27,117 annotated images. Specifically, moving beyond standard Masked Autoencoders, we propose a scale-adaptive masked image modeling strategy that jointly learns representations across resolutions and patch sizes under a fixed token budget. By decoupling structural semantic learning from rigid grid constraints, the model effectively resolves the full spectrum of neurodegenerative lesions - ranging from fine dendritic beading to gross soma shrinkage - within a tractable computational framework. Finally, our model surpasses both generalist and biomedical foundation models across classification, segmentation and detection tasks. Fusing visual features with morphological descriptors enables prediction of dopamine-dependent behavioral deficits ($R^2=0.498$). Screening 180 agrochemicals, we identify the benzimidazole moiety as a previously unrecognized determinant of dopaminergic neurotoxicity. Together, the work demonstrates how scale-adaptive self-supervised learning can connect morphology to function for a scalable alternative to mammalian in vivo models for neurotoxicity assessment and drug discovery.

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

64 extracted references · 1 canonical work pages

  1. [1]

    J Glob Health 13, 04160 (2023)

    Huang, Y., Li, Y., Pan, H.: Global, regional, and national burden of neurological disorders in 204 countries and territories worldwide. J Glob Health 13, 04160 (2023)

    2023

  2. [2]

    Front Public Health 10, 952161 (2019)

    Ding, C., Wu, Y., Chen, X.: Global, regional, and national burden and attributable risk factors of neurological disorders: The global burden of disease study 1990-2019. Front Public Health 10, 952161 (2019)

    1990

  3. [3]

    Toxicol Pathol 48(1), 37–38 (2019)

    Brown, D.L., Staup, M., Swanson, C.: Stereology of the peripheral nervous system. Toxicol Pathol 48(1), 37–38 (2019)

    2019

  4. [4]

    Front Pharmacol 17, 1765786 (2026)

    Chakif, D., Furrer, J.: The impact of nutritional, environmental, and lifestyle fac- tors on neurological disorders: therapeutic implications and mechanistic insights. Front Pharmacol 17, 1765786 (2026)

    2026

  5. [5]

    Genome Res 10, 703–10 (2000)

    Lai, C.H., Chou, C.Y., Ch’ang, L.Y.: Identification of novel human genes evolu- tionarily conserved in caenorhabditis elegans by comparative proteomics. Genome Res 10, 703–10 (2000)

    2000

  6. [6]

    Genome Res 10, 703–10 (2000)

    Kaletta, T., Hengartner, M.O.: Identification of novel human genes evolutionarily conserved in caenorhabditis elegans by comparative proteomics. Genome Res 10, 703–10 (2000)

    2000

  7. [7]

    Ruszkiewicz, J., Endig, L., G¨ uver, E.: Life-cycle-dependent toxicities of mono- and bifunctional alkylating agents in the 3r-compliant model organism c. elegans. Cells 12(23), 2728 (2023)

    2023

  8. [8]

    elegans through stable robotic sample rotation and artificial intelligence-based 3-dimensional reconstruction

    Pan, P., Zhang, P., Premachandran, S.: High-resolution imaging and morphologi- cal phenotyping of c. elegans through stable robotic sample rotation and artificial intelligence-based 3-dimensional reconstruction. Research 7, 0513 (2024)

    2024

  9. [9]

    In: WormBook: The Online Review of C

    Hart, A.C.: Behavior. In: WormBook: The Online Review of C. Elegans Biology. WormBook, Pasadena, CA (2006). https://doi.org/10.1895/wormbook.1.87.1

    work page doi:10.1895/wormbook.1.87.1 2006

  10. [10]

    IEEE Reviews in Biomedical Engineering 18, 172–191 (2024)

    He, Y., Huang, F., Jiang, X., Nie, Y., Wang, M., Wang, J., Chen, H.: Foundation model for advancing healthcare: challenges, opportunities and future directions. IEEE Reviews in Biomedical Engineering 18, 172–191 (2024)

    2024

  11. [11]

    Nature Communications 15, 654 (2024)

    Ma, J., He, Y., Li, F., Han, L., You, C., Wang, B.: Segment anything in medical images. Nature Communications 15, 654 (2024)

    2024

  12. [12]

    Nejm Ai 2(1), 2400640 (2025)

    Zhang, S., Xu, Y., Usuyama, N., Xu, H., Bagga, J., Tinn, R., Preston, S., Rao, R., Wei, M., Valluri, N., et al.: A multimodal biomedical foundation model trained from fifteen million image–text pairs. Nejm Ai 2(1), 2400640 (2025)

    2025

  13. [13]

    Nature methods 15(12), 1090–1097 (2018)

    Weigert, M., Schmidt, U., Boothe, T., M¨ uller, A., Dibrov, A., Jain, A., Wilhelm, 19 B., Schmidt, D., Broaddus, C., Culley, S.,et al.: Content-aware image restoration: pushing the limits of fluorescence microscopy. Nature methods 15(12), 1090–1097 (2018)

    2018

  14. [14]

    In: 2024 IEEE International Conference on Computer Vision and Machine Intelligence (CVMI), pp

    Dutta, D., Sonowal, N., Barauh, R., Chetia, D., Kalita, S.K.: Recent advance- ments in microscopy image enhancement using deep learning: A survey. In: 2024 IEEE International Conference on Computer Vision and Machine Intelligence (CVMI), pp. 1–7 (2024)

    2024

  15. [15]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    He, K., Chen, X., Xie, S., Li, Y., Doll´ ar, P., Girshick, R.: Masked autoencoders are scalable vision learners. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 16000–16009 (2022)

    2022

  16. [16]

    In: International Conference on Computer Vision and Pattern Recognition (CVPR) (2022)

    Xie, Z., Zhang, Z., Cao, Y., Lin, Y., Bao, J., Yao, Z., Dai, Q., Hu, H.: Simmim: A simple framework for masked image modeling. In: International Conference on Computer Vision and Pattern Recognition (CVPR) (2022)

    2022

  17. [17]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Beyer, L., Izmailov, P., Kolesnikov, A., Caron, M., Kornblith, S., Zhai, X., Min- derer, M., Tschannen, M., Alabdulmohsin, I., Pavetic, F.: Flexivit: One model for all patch sizes. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14496–14506 (2023)

    2023

  18. [18]

    In: European Conference on Computer Vision, pp

    Li, Y., Mao, H., Girshick, R., He, K.: Exploring plain vision transformer back- bones for object detection. In: European Conference on Computer Vision, pp. 280–296 (2022)

    2022

  19. [19]

    arXiv preprint arXiv:2304.07193 (2023)

    Oquab, M., Darcet, T., Moutakanni, T., Vo, H., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., et al.: Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193 (2023)

    Pith/arXiv arXiv 2023

  20. [20]

    Nature 637(8045), 319–326 (2025)

    Hollmann, N., M¨ uller, S., Purucker, L., Krishnakumar, A., K¨ orfer, M., Hoo, S.B., Schirrmeister, R.T., Hutter, F.: Accurate predictions on small data with a tabular foundation model. Nature 637(8045), 319–326 (2025)

    2025

  21. [21]

    Science 276(5321), 2045–71 (1997)

    Polymeropoulos, M.H., Lavedan, C., Leroy, E.: Mutation in the alpha-synuclein gene identified in families with parkinson’s disease. Science 276(5321), 2045–71 (1997)

    2045

  22. [22]

    Lancet Neurol 24(11), 976–986 (2025)

    Dorsey, E.R., DeMiranda, B.R., Hussain, S.: Environmental toxicants and parkin- son’s disease: recent evidence, risks, and prevention opportunities. Lancet Neurol 24(11), 976–986 (2025)

    2025

  23. [23]

    Neurochem Int 60(8), 773–81 (2012) 20

    Lee, J.E., J.S., K., Ki, Y.W.: Fluazinam targets mitochondrial complex i to induce reactive oxygen species-dependent cytotoxicity in sh-sy5y cells. Neurochem Int 60(8), 773–81 (2012) 20

    2012

  24. [24]

    bioRxiv 23, 2025–1021683798 (2025)

    Huayta, J., Meyer, J.N.: Inhibition of mitochondrial complex iii causes dopamin- ergic neurodegeneration by redox stress in Caenorhabditis elegans. bioRxiv 23, 2025–1021683798 (2025)

    2025

  25. [25]

    J Vet Med Sci 85(4), 497–506 (2023)

    Kimura, M., Shoda, A., Murata, M.: Neurotoxicity and behavioral disorders induced in mice by acute exposure to the diamide insecticide chlorantraniliprole. J Vet Med Sci 85(4), 497–506 (2023)

    2023

  26. [26]

    Fitzmaurice, S.L. A.G. Rhodes, Lulla, A.: Aldehyde dehydrogenase inhibition as a pathogenic mechanism in parkinson disease. Proc Natl Acad Sci U S A 110(2), 636–41 (2013)

    2013

  27. [27]

    Environ Sci Technol 50(7), 3937–44 (2016)

    Tufi, S., Wassenaar, P.N., Osorio, V.H.: Pesticide mixture toxicity in surface water extracts in snails (lymnaea stagnalis) by an in vitro acetylcholinesterase inhibition assay and metabolomics. Environ Sci Technol 50(7), 3937–44 (2016)

    2016

  28. [28]

    Neurochem Res 47(7), 1956–1971 (2022)

    Ebedy, Y.A., Hassanen, E.I., Hussien, A.M.: Neurobehavioral toxicity induced by carbendazim in rats and the role of inos, cox-2, and nf- κb signalling pathway. Neurochem Res 47(7), 1956–1971 (2022)

    1956

  29. [29]

    elegans locomotory rate is mod- ulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway

    Sawin, E.R., Ranganathan, R., Horvitz, H.R.: C. elegans locomotory rate is mod- ulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26(3), 619–31 (2000)

    2000

  30. [30]

    Tanimoto, Y., Zhang, Y.G., Fei, X.: In actio optophysiological analyses reveal functional diversification of dopaminergic neurons in the nematode c. elegans. Sci Rep 6, 26297 (2016)

    2016

  31. [31]

    Redox Biol 81, 103538 (2025)

    Morton, K.S., George, A.J., Meyer, J.N.: Complex i superoxide anion produc- tion is necessary and sufficient for complex i inhibitor-induced dopaminergic neurodegeneration in caenorhabditis elegans. Redox Biol 81, 103538 (2025)

    2025

  32. [32]

    elegans and neurodegenera- tion In Caenorhabditis Elegans: Anatomy, Life Cycles and Biological Functions

    Martinez-Finley, E.J., Chakraborty, S., Caito, S.: C. elegans and neurodegenera- tion In Caenorhabditis Elegans: Anatomy, Life Cycles and Biological Functions. Adv Med Biol 44, 1–46 (2012)

    2012

  33. [33]

    Clark, A.S., Kalmanson, Z., Morton, K.: An unbiased, automated platform for scoring dopaminergic neurodegeneration in c. elegans. PLoS One 18(7), 0281797 (2023)

    2023

  34. [34]

    elegans intestine is neuroprotective

    Chikka, M.R., Anbalagan, C., Dvorak, K.: The mitochondria-regulated immune pathway activated in the c. elegans intestine is neuroprotective. Cell Rep 16(9), 2399–414 (2016)

    2016

  35. [35]

    Nat Commun 15(1), 3658 (2024) 21

    Stillman, Q.H., Joseph, J.A., Ahmed, J.: Protein mimetic 2d fast rescues alpha synuclein aggregation mediated early and post disease parkinson’s phenotypes. Nat Commun 15(1), 3658 (2024) 21

    2024

  36. [36]

    Annu Rev Pharmacol Toxicol 43, 521–44 (2003)

    Nass, R., Blakely, R.D.: The caenorhabditis elegans dopaminergic system: oppor- tunities for insights into dopamine transport and neurodegeneration. Annu Rev Pharmacol Toxicol 43, 521–44 (2003)

    2003

  37. [37]

    Mov Disord 33(12), 1928–1937 (2018)

    Lynch, W.B., Tschumi, C.W., Sharpe, A.L.: Progressively disrupted somatoden- dritic morphology in dopamine neurons in a mouse parkinson’s model. Mov Disord 33(12), 1928–1937 (2018)

    1928

  38. [38]

    Mov Disord 38(2), 256–266 (2023)

    Ledonne, A., Massaro, C.M., Paldino, E.: Morpho-functional changes of nigral dopamine neurons in an alpha-synuclein model of parkinson’s disease. Mov Disord 38(2), 256–266 (2023)

    2023

  39. [39]

    Environ Health Perspect 120, 340–7 (2012)

    Mark M., Brouwer, M., Kromhout, H.: Is pesticide use related to parkinson dis- ease? some clues to heterogeneity in study results. Environ Health Perspect 120, 340–7 (2012)

    2012

  40. [40]

    Int J Environ Res Public Health 16, 337 (2019)

    Gunnarsson, L.G., Bodin, L.: Occupational exposures and neurodegenerative diseases-a systematic literature review and meta-analyses. Int J Environ Res Public Health 16, 337 (2019)

    2019

  41. [41]

    Neurotoxicology 80, 76–78 (2020)

    Sanchez, C.L., Souders, C.L., Pena-Delgado, C.J.: Neurotoxicity assessment of tri- azole fungicides on mitochondrial oxidative respiration and lipids in differentiated human sh-sy5y neuroblastoma cells. Neurotoxicology 80, 76–78 (2020)

    2020

  42. [42]

    Nat Commun 14(1), 2803 (2023)

    Paul, K.C., Krolewski, R.C., Lucumi, M.E.: A pesticide and ipsc dopaminergic neuron screen identifies and classifies parkinson-relevant pesticides. Nat Commun 14(1), 2803 (2023)

    2023

  43. [43]

    Environ Int 208, 110087 (2026)

    Arellano, M., Barnhill, L.M., Kim, A.M.: Identification of pesticides associated with an increased risk of parkinson’s disease using a multi-screen approach. Environ Int 208, 110087 (2026)

    2026

  44. [44]

    Chem Biol Drug Des 86(1), 19–65 (2015)

    Keri, R.S., Hiremathad, A., Budagumpi, S.: Comprehensive review in current developments of benzimidazole-based medicinal chemistry. Chem Biol Drug Des 86(1), 19–65 (2015)

    2015

  45. [45]

    Molecules 29, 1218 (2024)

    Bai, S., Zhang, M., Tang, S.: Research progress on benzimidazole fungicides: A review. Molecules 29, 1218 (2024)

    2024

  46. [46]

    Silva: Neurotoxicity and teratogenicity induced by carbendazim and ametryn in zebrafish: Implications for environmental and biological health

    Teixeira, J.R., Tavares, L.A.M., da, A.P. Silva: Neurotoxicity and teratogenicity induced by carbendazim and ametryn in zebrafish: Implications for environmental and biological health. Environ Toxicol Pharmacol 122, 104947 (2026)

    2026

  47. [47]

    Neurobehav Toxicol Teratol 4(6), 779–87 (1982)

    Miller, D.B.: Neurotoxicity of the pesticidal carbamates. Neurobehav Toxicol Teratol 4(6), 779–87 (1982)

    1982

  48. [48]

    Toxicology 91(1), 29–36 22 (1994)

    Machemer, L.H., Pickel, M.: Carbamate insecticides. Toxicology 91(1), 29–36 22 (1994)

    1994

  49. [49]

    Friedman, Gagne, M., Loo, L.H.: Utility of in vitro bioactivity as a lower bound estimate of in vivo adverse effect levels and in risk-based prioritization

    Pau, K. Friedman, Gagne, M., Loo, L.H.: Utility of in vitro bioactivity as a lower bound estimate of in vivo adverse effect levels and in risk-based prioritization. Toxicol Sci 173(1), 202–225 (2020)

    2020

  50. [50]

    Aging Dis 16(5), 2937–2952 (2024)

    Wang, S., Jiang, Y., Yang, A.: The expanding burden of neurodegenerative diseases: An unmet medical and social need. Aging Dis 16(5), 2937–2952 (2024)

    2024

  51. [51]

    elegans neurons jettison protein aggregates and mitochondria under neurotoxic stress

    Melentijevic, I., Toth, M.L., Arnold, M.L.: C. elegans neurons jettison protein aggregates and mitochondria under neurotoxic stress. Nature 542(7641), 367–371 (2017)

    2017

  52. [52]

    2022 IEEE International Conference on Image Processing (ICIP), 966–970 (2022) https://doi.org/10.1109/ICIP46576

    Akyon, F.C., Altinuc, S.O., Temizel, A.: Slicing aided hyper inference and fine-tuning for small object detection. 2022 IEEE International Conference on Image Processing (ICIP), 966–970 (2022) https://doi.org/10.1109/ICIP46576. 2022.9897990

    arXiv 2022

  53. [53]

    In: International Conference on Learning Representations (2021)

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., Houlsby, N.: An image is worth 16x16 words: Transformers for image recognition at scale. In: International Conference on Learning Representations (2021)

    2021

  54. [54]

    https://github.com/open-mmlab/mmsegmentation (2020)

    Contributors, M.: MMSegmentation: OpenMMLab Semantic Segmentation Tool- box and Benchmark. https://github.com/open-mmlab/mmsegmentation (2020)

    2020

  55. [55]

    arXiv preprint arXiv:1906.07155 (2019)

    Chen, K., Wang, J., Pang, J., Cao, Y., Xiong, Y., Li, X., Sun, S., Feng, W., Liu, Z., Xu, J., Zhang, Z., Cheng, D., Zhu, C., Cheng, T., Zhao, Q., Li, B., Lu, X., Zhu, R., Wu, Y., Dai, J., Wang, J., Shi, J., Ouyang, W., Loy, C.C., Lin, D.: MMDetection: Open mmlab detection toolbox and benchmark. arXiv preprint arXiv:1906.07155 (2019)

    Pith/arXiv arXiv 1906

  56. [56]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Chen, X., Xie, S., He, K.: An empirical study of training self-supervised vision transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 9640–9649 (2021)

    2021

  57. [57]

    In: European Conference on Computer Vision, pp

    Assran, M., Caron, M., Misra, I., Bojanowski, P., Bordes, F., Vincent, P., Joulin, A., Rabbat, M., Ballas, N.: Masked siamese networks for label-efficient learning. In: European Conference on Computer Vision, pp. 456–473 (2022)

    2022

  58. [58]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Liu, J., Huang, X., Zheng, J., Liu, Y., Li, H.: Mixmae: Mixed and masked autoen- coder for efficient pretraining of hierarchical vision transformers. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 6252–6261 (2023)

    2023

  59. [59]

    In: International 23 Conference on Learning Representations (2023)

    Zhang, X., Tian, Y., Xie, L., Huang, W., Dai, Q., Ye, Q., Tian, Q.: Hivit: A sim- pler and more efficient design of hierarchical vision transformer. In: International 23 Conference on Learning Representations (2023)

    2023

  60. [60]

    Image and Vision Computing149, 105171 (2024)

    Fang, Y., Sun, Q., Wang, X., Huang, T., Wang, X., Cao, Y.: Eva-02: A visual representation for neon genesis. Image and Vision Computing149, 105171 (2024)

    2024

  61. [61]

    Transactions on Machine Learning Research Journal (2024)

    Oquab, M., Darcet, T., Moutakanni, T., Vo, H., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., et al.: Dinov2: Learning robust visual features without supervision. Transactions on Machine Learning Research Journal (2024)

    2024

  62. [62]

    arXiv preprint arXiv:2502.14786 (2025)

    Tschannen, M., Gritsenko, A., Wang, X., Naeem, M.F., Alabdulmohsin, I., Parthasarathy, N., Evans, T., Beyer, L., Xia, Y., Mustafa, B., et al.: Siglip 2: Multilingual vision-language encoders with improved semantic understanding, localization, and dense features. arXiv preprint arXiv:2502.14786 (2025)

    Pith/arXiv arXiv 2025

  63. [63]

    Nature communications 15(1), 654 (2024)

    Ma, J., He, Y., Li, F., Han, L., You, C., Wang, B.: Segment anything in medical images. Nature communications 15(1), 654 (2024)

    2024

  64. [64]

    arXiv preprint arXiv:2410.06542 (2024) 24 Appendix A Extended Data Fig

    Codella, N.C., Jin, Y., Jain, S., Gu, Y., Lee, H.H., Abacha, A.B., Santamaria- Pang, A., Guyman, W., Sangani, N., Zhang, S., et al.: Medimageinsight: An open-source embedding model for general domain medical imaging. arXiv preprint arXiv:2410.06542 (2024) 24 Appendix A Extended Data Fig. A1 Qualitative results of our proposed framework. a, Reconstructed c...

    Pith/arXiv arXiv 2024