Pith. sign in

REVIEW 6 major objections 5 minor 22 references

Hybrid Ensemble of Segmentation-Assisted Classification and GBDT for Skin Cancer Detection with Engineered Metadata and Synthetic Lesions from ISIC 2024 Non-Dermoscopic 3D-TBP Images

T0 review · 6 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that fusing image-model probabilities with a gradient-boosted tree on engineered metadata reaches a partial AUC of 0.1755 on the ISIC 2024 SLICE-3D skin cancer dataset.

desk verdict Plausible applied ensemble on SLICE-3D, but unresolved synthetic-data leakage undermines the headline pAUC. read the letter →

arxiv 2506.03420 v1 pith:CN67IFR3 submitted 2025-06-03 eess.IV cs.CVcs.LG

Muhammad Zubair Hasan , Fahmida Yasmin Rifat This is my paper
classification eess.IVcs.CVcs.LG
keywords skincancerdetectionISIC2024SLICE-3Dsegmentation-assistedclassificationgradient-boosteddecisiontreessyntheticlesionaugmentationpartialAUC3Dtotalbodyphotography
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

Skin cancer screening from non-dermoscopic 3D total-body-photography images is the setting, and the paper argues that the best classifier is not a single image model but a two-stage ensemble. Image models (EVA02 and a segmentation-assisted EdgeNeXt variant) produce soft probabilities; those probabilities are concatenated with 214 engineered metadata features and passed into an ensemble of gradient-boosted decision trees. On the ISIC 2024 SLICE-3D dataset, this configuration reaches a partial AUC of 0.1755 above an 80% true-positive rate, the highest among all tested configurations. The authors see this as evidence that structured patient metadata, synthetic malignant examples, and deep visual features are complementary for triage in telemedicine and low-resource settings.

What carries the argument

The engine is the two-stage fusion pipeline: EVA02 and EdgeNeXtSAC produce softmax probabilities, which are concatenated with 214 tabular features (raw metadata, one-hot encodings, engineered geometric and color descriptors, patient-normalized deviations, and external 3-class predictions) and fed into 45 GBDT models (LightGBM, XGBoost, CatBoost across 5 folds and 3 seeds). The segmentation-assisted component adds a UNet-style decoder and attention blocks to EdgeNeXt, trained with total loss $L_{\text{total}} = L_{\text{cls}} + L_{\text{cam}} + L_{\text{seg}}$, where the CAM loss aligns the class activation map with a ground-truth mask and the segmentation loss is a pixel-level BCE-plus-Dice loss. Synthetic malignant images and a diagnosis-informed relabeling of external data into nevus/melanoma/bkl feed the training set.

What would settle it

Run the EdgeNeXtSAC pipeline twice on identical folds: once with the paper's actual mask source and once with masks replaced by circular disks centered at the provided lesion coordinates with the provided lesion diameters, keeping everything else fixed. If the pAUC does not change materially, the segmentation loss and its masks are not doing the work attributed to them; if it drops, the mask quality is load-bearing.

Watch

Extended reading notes

Core claim

The paper's central claim is that the fusion step, not any single model, carries the performance. Image-only models peak at 0.1633 (EVA02 with synthetic data), a raw-metadata GBDT baseline scores 0.1500, and feature engineering lifts the GBDT to 0.1644; adding image-model probabilities to the GBDT pushes the ensemble to 0.1755. The segmentation-assisted EdgeNeXtSAC model improves on its plain counterpart (0.1439 vs. 0.1401 on real data) and further to 0.1576 with synthetic lesions, and its attention maps localize lesions more sharply than the baseline. The paper also reports that benign predictions are concentrated near confidence 0 while malignant predictions spread across the range, and that top GBDT features include engineered color and shape descriptors plus deep prediction scores.

Load-bearing premise

The load-bearing premise is that a ground-truth lesion mask exists for every training image, because the segmentation and CAM losses depend on it, yet the paper never states where the masks come from.

Editorial extensions

If this is right

  • The highest reported pAUC of 0.1755 is achieved only when GBDT features include image-model probabilities, so the paper's results imply that metadata-only and image-only pipelines leave performance on the table.
  • Synthetic malignant augmentation raises EVA02 from 0.1516 to 0.1633 and EdgeNeXtSAC from 0.1439 to 0.1576, implying that generated lesions help the models generalize to rare malignant cases.
  • Segmentation supervision improves both pAUC and localization: EdgeNeXtSAC outperforms plain EdgeNeXt, and its GradCAM++ overlays focus on lesion regions.
  • The confidence analysis implies that near-threshold false positives and low-confidence malignant predictions are the main error source, so threshold adjustment or calibration is a direct next step.

Reading between the lines

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

  • The paper never states how the ground-truth masks used in the segmentation and CAM losses were obtained; SLICE-3D provides coordinates and diameters, not pixel masks, so the masks must come from an unstated derivation. If that derivation is simple, such as circular masks, the segmentation gain may come from enforcing a shape prior rather than learning true lesion boundaries.
  • Because synthetic augmentation and the diagnosis-informed relabeling are introduced together, the individual contribution of Stable Diffusion data is not cleanly isolated; testing each separately would attribute the 0.1633 and 0.1576 gains.
  • The same fusion recipe, deep soft probabilities plus engineered tabular features into a GBDT, could be tested on other lesion datasets as a general way to combine structured and visual data in medical imaging.
  • The reported confidence distributions suggest that a well-chosen operating point or calibrated score, rather than a fixed 0.5 threshold, could recover some low-confidence malignant cases without retraining.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

6 major / 5 minor

Summary. This paper reports a two-stage hybrid pipeline for binary skin-lesion classification on the ISIC 2024 SLICE-3D dataset. The first stage consists of image classifiers: a fine-tuned EVA02 vision transformer, a second EVA02 trained as a 3-class classifier on relabeled external data, and a proposed segmentation-assisted classifier called EdgeNeXtSAC built on an EdgeNeXt backbone with BCE, CAM-Dice, and segmentation-Dice losses. The second stage is a GBDT ensemble (LightGBM, XGBoost, CatBoost) over 214 engineered metadata features plus the image-model probability outputs, with additional augmentation of the 393 real malignant images by 30,228 Stable Diffusion-generated synthetic malignant lesions. The central claim is that the full configuration achieves a partial AUC (pAUC) above 80% TPR of 0.1755, the highest among the evaluated configurations (Tables 2 and 3). The paper also includes confidence-distribution analysis, GradCAM++ visualizations, error examples, and feature-importance analysis.

Significance. If the reported pAUC is valid, the paper makes a useful empirical contribution: it evaluates on a large public non-dermoscopic dataset, uses the clinically relevant pAUC-above-80%-TPR metric, and provides ablations showing gains from synthetic augmentation, feature engineering, and fusion of image-model probabilities with GBDT. The segmentation-assisted classification idea is plausible and the visual localization results are suggestive. However, the current manuscript does not provide code, data splits, error bars, a precise description of how segmentation ground-truth masks were obtained, or a leakage-safe description of how synthetic images and image-model probability features were created relative to the cross-validation folds. These omissions are load-bearing because the headline pAUC is a single number whose validity depends on the absence of cross-validation leakage and on reproducible experimental setup. The paper therefore cannot yet be recommended for acceptance, but the core approach is defensible and the issues appear addressable in revision.

major comments (6)
  1. [Section 2.1 and Section 2.2.1] The synthetic-augmentation protocol is not specified in a way that rules out cross-validation leakage. The paper states that the training set is augmented with 30,228 Stable Diffusion-generated malignant images alongside 393 real malignant images, but it does not state whether the Stable Diffusion model was trained or fine-tuned on all 393 real malignant images, nor whether synthetic images were generated before or after the patient-level GroupKFold splits. If the generator was fit on all real malignant images and the augmented set was then split, the training folds could contain near-duplicates of held-out real lesions, inflating the reported improvements (e.g., EVA02 real-only 0.1516 vs. EVA02+Synth 0.1633 in Table 2) and the final pAUC of 0.1755 in Table 3. Please specify the exact temporal order, the generator's training set, how the class-conditional generation was conditioned, and provide a leakage-free control experiment in which synthetic images are generated separately for each training fold.
  2. [Section 2.2.3] The image-model probabilities used as GBDT input features appear to be produced by models trained on the same augmented data as the validation folds, but the paper never states that these probability features are out-of-fold. Section 2.2.3 describes a 3x5-fold GBDT setup with three random seeds, while the image models use a 3-fold Stratified GroupKFold setup (Section 2.2.1), so the relationship between the two splitting schemes is unclear. If the probability features are computed from in-fold predictions, the GBDT can exploit the image models' training predictions and the validation pAUC is optimistically biased. Please clarify exactly how each lesion's probability feature is generated relative to the image-model folds and the GBDT folds, and report results with out-of-fold probability features.
  3. [Equations (2)-(3)] The segmentation loss and the CAM loss require a binary ground-truth mask M_gt, but the paper never states where these masks come from. The ISIC 2024 SLICE-3D dataset provides lesion coordinates and diameters rather than pixel-level segmentation masks, so the derivation of M_gt is an unstated modeling choice that directly determines the behavior of the CAM-Dice loss and the segmentation supervision for EdgeNeXtSAC. Please describe the mask construction procedure in detail, including how lesion coordinates and diameters were converted to pixel masks, or acknowledge that this is a heuristic; otherwise the EdgeNeXtSAC results in Table 2 are not reproducible.
  4. [Section 2.2.1] The relabeling strategy contains a clinically incorrect mapping. The paper states that keratinocyte-related lesions including basal cell carcinoma are grouped under the class 'bkl' and describes 'bkl' as 'benign keratinocyte lesion'. Basal cell carcinoma is a malignant skin cancer, not a benign lesion. Mapping BCC to a benign class corrupts the external 3-class training labels and any downstream features derived from that classifier. Please correct the terminology and mapping, and state whether the external dataset's BCC labels were actually treated as malignant in the binary task or as 'bkl' in the 3-class task.
  5. [Tables 2 and 3] All pAUC values are reported as single numbers with no standard deviation, confidence intervals, or per-fold breakdown. Since the GBDT ensemble uses 3 seeds and 5 folds, and image models use 3 patient-grouped folds, the reported differences (e.g., 0.1401 vs. 0.1439 for EdgeNeXt and EdgeNeXtSAC, and 0.1516 vs. 0.1633 for EVA02 with and without synthetic data) may be within stochastic variation. Please report mean ± std across seeds/folds and, where possible, the per-fold pAUC values, so the reader can judge whether the claimed gains are statistically meaningful.
  6. [General reproducibility] The manuscript does not provide code, model checkpoints, exact hyperparameters, preprocessing details, or the construction of the external dataset used for the 3-class classifier. Given that the central claim is an empirical performance number on a public benchmark, this is a major omission. In a revision, the authors should release the code and the exact cross-validation split assignments, or at minimum provide enough detail (including seed values, learning rates, batch sizes, and the external data source) for an independent replication of Table 3.
minor comments (5)
  1. [Section 3.2, Eq. (5)] The pAUC formula is written as an integral of ROC(t) over 0.8 to 1.0, but the variable t is not defined and the direction of the ROC curve is ambiguous. Please define whether t denotes TPR or FPR and state the normalization explicitly, since the metric is central to the evaluation.
  2. [Table 1] The table lists 'Raw categorical (1-hot) 6' and separately 'One-hot Categorical Features 47'; the relationship between these two rows is unclear, and the claimed 174 engineered features versus the 214 total would benefit from a clearer accounting.
  3. [Figure 2] The pipeline diagram refers to 'EV A02-old' without defining it in the text; clarify whether this is the 3-class classifier or a separately trained earlier model.
  4. [Section 2.2.2] The text says '174 engineered features' are added to the original metadata to reach 214 total, but Table 1 appears to include raw numeric and raw categorical features in the total; please state which count refers to handcrafted additions versus original metadata.
  5. [References] Several references are missing venue or publication details (e.g., [12] and [14]); please standardize the bibliography.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identified: the reported pAUC values are empirical measurements on held-out folds, not outputs of a fitted equation or a self-citation chain.

full rationale

The paper is an empirical machine-learning study. Its central claim is a measured partial AUC of 0.1755 from a two-stage ensemble (Section 4.1, Table 3). This number is a validation-set measurement, not the result of a derivation whose conclusion is assumed in its inputs. The main pipeline components are: image classifiers (EVA02, EdgeNeXtSAC) trained on image data, a GBDT ensemble trained on engineered metadata plus image-model probability features, and Stable Diffusion synthetic augmentation. None of these components is defined in terms of the target pAUC or fitted directly to the final validation labels. The GBDT feature set includes softmax outputs from image models, but the paper explicitly states that both the image-model training and the GBDT training use patient-disjoint stratified GroupKFold splits (Sections 2.2.1 and 2.2.3), so the described procedure is designed to avoid the classic fitted-input circularity. Whether the out-of-fold bookkeeping is actually implemented correctly is not fully specified, and the source of the ground-truth segmentation masks M_gt for Equations (2) and (3) is never stated; these are reproducibility and evaluation-integrity risks, not circular reductions. The paper does not rely on self-citations for its load-bearing claims: all cited prior work is external, and no uniqueness theorem or ansatz is imported from the authors' own prior publications. The claimed pAUC improvements are presented as ablations over the paper's own configurations, and the conclusion is a comparative empirical statement rather than a theorem derived from its premises. Therefore no circular step can be exhibited from the paper's own equations or citations, and the appropriate circularity score is 0.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

The central claim (pAUC 0.1755) rests on empirical measurements that depend on several unverified choices: the hand-set regularization noise, the number of synthetic images, the oversampling ratio, the clinically questionable relabeling of external data, the assumed representativeness of synthetic images, and the undisclosed source of segmentation masks. These are not free parameters in a derivation, but they are load-bearing choices in the experimental setup.

free parameters (3)
  • Gaussian noise sigma = 0.1
    Injected into external prediction features to regularize the GBDT; chosen by hand and not justified by a principled rule (Section 2.2.2).
  • Synthetic augmentation count = 30,228
    Number of Stable Diffusion generated malignant lesions; the specific count appears chosen to balance classes and is not derived from any stated criterion (Section 2.1).
  • Oversampling ratio = 1:1
    Training batches are sampled with a 1:1 benign to malignant ratio; this is a modeling choice that affects model sensitivity (Section 2.2.1).
assumptions (3)
  • domain assumption The relabeling of external diagnoses into three classes is clinically valid.
    The auxiliary 3-class classifier depends on the mapping in Section 2.2.1, which groups basal cell carcinoma into the benign bkl class; this is clinically questionable and can corrupt the auxiliary features.
  • domain assumption Synthetic images from Stable Diffusion are representative of real malignant lesions.
    The augmentation is used to mitigate class imbalance (Section 2.1); if the synthetic distribution differs from real malignant lesions, the reported pAUC may not transfer to clinical data.
  • ad hoc to paper Ground-truth masks exist for the segmentation and CAM losses.
    The segmentation loss (Eq. 3) and CAM loss (Eq. 2) require masks M_gt, but no source is described; this is an unstated modeling choice that the paper introduces.
invented entities (1)
  • EdgeNeXtSAC
    purpose: Segmentation-assisted classifier combining an EdgeNeXt backbone with a decoder and CBAM attention.
    It is a new architecture variant introduced in this paper; performance is only shown on the internal validation split, with no external generalization evidence and no code release.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid Ensemble of Segmentation-Assisted Classification and GBDT for Skin Cancer Detection with Engineered Metadata and Synthetic Lesions from ISIC 2024 Non-Dermoscopic 3D-TBP Images." pith.science (2026). https://pith.science/paper/CN67IFR3

@misc{pith2026250603420,
  author       = {Pith},
  title        = {Pith review of: Hybrid Ensemble of Segmentation-Assisted Classification and GBDT for Skin Cancer Detection with Engineered Metadata and Synthetic Lesions from ISIC 2024 Non-Dermoscopic 3D-TBP Images},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CN67IFR3}},
  note         = {Machine review of arXiv:2506.03420}
}
read the original abstract

Skin cancer is among the most prevalent and life-threatening diseases worldwide, with early detection being critical to patient outcomes. This work presents a hybrid machine and deep learning-based approach for classifying malignant and benign skin lesions using the SLICE-3D dataset from ISIC 2024, which comprises 401,059 cropped lesion images extracted from 3D Total Body Photography (TBP), emulating non-dermoscopic, smartphone-like conditions. Our method combines vision transformers (EVA02) and our designed convolutional ViT hybrid (EdgeNeXtSAC) to extract robust features, employing a segmentation-assisted classification pipeline to enhance lesion localization. Predictions from these models are fused with a gradient-boosted decision tree (GBDT) ensemble enriched by engineered features and patient-specific relational metrics. To address class imbalance and improve generalization, we augment malignant cases with Stable Diffusion-generated synthetic lesions and apply a diagnosis-informed relabeling strategy to harmonize external datasets into a 3-class format. Using partial AUC (pAUC) above 80 percent true positive rate (TPR) as the evaluation metric, our approach achieves a pAUC of 0.1755 -- the highest among all configurations. These results underscore the potential of hybrid, interpretable AI systems for skin cancer triage in telemedicine and resource-constrained settings.

Figures

Figures reproduced from arXiv: 2506.03420 by the authors.

Figure 1
Figure 1. Sample images from the SLICE-3D dataset. These [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overall pipeline: synthetic malignant lesions are added to real data. Multiple image models (EVA02, EdgeNeXtSAC, EVA02- [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Segmentation-assisted classification architecture (EdgeNeXtSAC). The backbone (EdgeNeXt) feeds multi-scale features into a [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Left: Distribution of true positives (green) vs. false positives (red) across prediction confidence. Right: Overall predicted [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Comparison of lesion localization using GradCAM++ [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Failure case: despite localizing a lesion, the model over [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

22 extracted references · 21 canonical work pages

  1. [1]

    A. R. Baig, Q. Abbas, R. Almakki, M. E. A. Ibrahim, L. Al- Suwaidan, and A. E. S. Ahmed. Light-dermo: A lightweight pretrained convolution neural network for the diagnosis of multiclass skin lesions.Diagnostics, 13(3):385, 2023. 1

    work page 2023

  2. [2]

    Behara, E

    K. Behara, E. Bhero, and J. T. Agee. An improved skin lesion classification using a hybrid approach with active contour snake model and lightweight attention-guided capsule net- works.Diagnostics (Basel, Switzerland), 14(6):636, 2024. 1

    work page 2024

  3. [3]

    Cheng, J

    H. Cheng, J. Lian, and W. Jiao. Enhanced mobilenet for skin cancer image classification with fused spatial channel attention mechanism.Scientific Reports, 14:28850, 2024. 1

    work page 2024

  4. [4]

    A self-contrastive learning framework for skin cancer detection using histological images

    Roc ´ıo del Amor, Adri ´an Colomer, Sandra Morales, Cristian Pulgar´ın-Ospina, Liria Terradez, Jose Aneiros-Fernandez, and Valery Naranjo. A self-contrastive learning framework for skin cancer detection using histological images. In2022 IEEE International Conference on Image Processing (ICIP), pages 2291–2295, 2022. 1

    work page 2022

  5. [5]

    Hi-mvit: A lightweight model for explainable skin disease classifi- cation based on modified mobilevit.Digital Health, 9: 20552076231207197, 2023

    Yiming Ding, Zhi Yi, Ming Li, Jie Long, Shijie Lei, Yufei Guo, Ping Fan, Chao Zuo, and Yusen Wang. Hi-mvit: A lightweight model for explainable skin disease classifi- cation based on modified mobilevit.Digital Health, 9: 20552076231207197, 2023. 1

    work page 2023

  6. [6]

    Eva-02: A visual representation for neon genesis.Image and Vision Computing, 149:105171,

    Yuxin Fang, Quan Sun, Xinggang Wang, Tiejun Huang, Xin- long Wang, and Yue Cao. Eva-02: A visual representation for neon genesis.Image and Vision Computing, 149:105171,

  7. [7]

    S. A. Hamim, M. U. I. Tamim, M. F. Mridha, M. Safran, and D. Che. SmartSkin-XAI: An Interpretable Deep Learn- ing Approach for Enhanced Skin Cancer Diagnosis in Smart Healthcare.Diagnostics, 15(1):64, 2025. 1

    work page 2025

  8. [8]

    Sohel Rahman

    Nabil Ibtehaz and M. Sohel Rahman. Multiresunet : Re- thinking the u-net architecture for multimodal biomedical image segmentation.Neural Networks, 121:74–87, 2020. 4

    work page 2020

Show all 22 references
  1. [9]

    Mini- mal sourced and lightweight federated transfer learning mod- els for skin cancer detection.Scientific Reports, 15:2605,

    Vandana Khullar, Parul Kaur, Saurabh Gargrish, et al. Mini- mal sourced and lightweight federated transfer learning mod- els for skin cancer detection.Scientific Reports, 15:2605,

  2. [10]

    Kousis, I

    I. Kousis, I. Perikos, I. Hatzilygeroudis, and M. Virvou. Deep learning methods for accurate skin cancer recognition and mobile application.Electronics, 11(9):1294, 2022. 1

    2022

  3. [11]

    Kurtansky, Brian M

    Nicholas R. Kurtansky, Brian M. D’Alessandro, Maura C. Gillis, Brigid Betz-Stablein, Sara E. Cerminara, Rafael Gar- cia, Marcela Alves Girundi, Elisabeth Victoria Goessinger, Philippe Gottfrois, Pascale Guitera, Allan C. Halpern, Va- lerie Jakrot, Harald Kittler, Kivanc Kose, ...

    2024

  4. [12]

    A novel skin cancer assisted diagnosis method based on cap- sule networks with cbam

    Zhangli Lan, Songbai Cai, Jiqiang Zhu, and Yuantong Xu. A novel skin cancer assisted diagnosis method based on cap- sule networks with cbam. 2023. 1

    2023

  5. [13]

    Kumar Lilhore, S

    U. Kumar Lilhore, S. Simaiya, Y . K. Sharma, et al. A pre- cise model for skin cancer diagnosis using hybrid u-net and improved mobilenet-v3 with hyperparameters optimization. Scientific Reports, 14:4299, 2024. 1

    2024

  6. [14]

    Edgenext: Efficiently amalgamated cnn-transformer architecture for mobile vision applications, 2022

    Muhammad Maaz, Abdelrahman Shaker, Hisham Cholakkal, Salman Khan, Syed Waqas Zamir, Rao Muham- mad Anwer, and Fahad Shahbaz Khan. Edgenext: Efficiently amalgamated cnn-transformer architecture for mobile vision applications, 2022. 2

    2022

  7. [15]

    Rani Suresh Mohadikar and Chetan Ashokrao Dhule. Ad- vancement in melanoma detection: A comprehensive re- view on deep learning based classification approaches.Pro- ceedings of the 5th International Conference on Information Management & Machine Intelligence, 2023. 1

    2023

  8. [16]

    El-Behaidy

    Ensaf Hussein Mohamed and Wessam H. El-Behaidy. En- hanced skin lesions classification using deep convolutional networks. In2019 Ninth International Conference on Intel- ligent Computing and Information Systems (ICICIS), pages 180–188, 2019. 1

    2019

  9. [17]

    Skin cancer detection using deep learning—a review.Diagnostics, 13(11):1911,

    Mujtaba Naqvi, Syed Qamar Abbas Gilani, Tahir Syed, Oge Marques, and Heung-Il Choi Kim. Skin cancer detection using deep learning—a review.Diagnostics, 13(11):1911,

    1911

  10. [18]

    Garc´ıa-de-la Puente, Miguel L´opez-P´erez, La¨etitia Launet, and Valery Naranjo

    Natalia P. Garc´ıa-de-la Puente, Miguel L´opez-P´erez, La¨etitia Launet, and Valery Naranjo. Domain adaptation for unsuper- vised cancer detection: An application for skin whole slides images from an interhospital dataset. InMedical Image Com- puting and Computer Assisted Inte...

    2024

  11. [19]

    P. N. Srinivasu, J. G. SivaSai, M. F. Ijaz, A. K. Bhoi, W. Kim, and J. J. Kang. Classification of skin disease using deep learning neural networks with mobilenet v2 and lstm. Sensors, 21(8):2852, 2021. 1

    2021

  12. [20]

    Smart mobinet: A deep learning approach for accurate skin cancer diagnosis.Computers, Materials and Continua, 77(3):3533–3549, 2023

    Muhammad Suleman, Faizan Ullah, Ghadah Aldehim, Di- lawar Shah, Mohammad Abrar, Asma Irshad, and Sarra Ay- ouni. Smart mobinet: A deep learning approach for accurate skin cancer diagnosis.Computers, Materials and Continua, 77(3):3533–3549, 2023. 1

    2023

  13. [21]

    Dscc net: Multi-classification deep learning models for diagnosing of skin cancer using dermoscopic images.Cancers, 15(7):2179,

    Muhammad Tahir, Adeel Naeem, Haroon Malik, Junaid Tan- veer, Rubina A Naqvi, and Sang-Woong Lee. Dscc net: Multi-classification deep learning models for diagnosing of skin cancer using dermoscopic images.Cancers, 15(7):2179,

  14. [22]

    Cbam: Convolutional block attention module, 2018

    Sanghyun Woo, Jongchan Park, Joon-Young Lee, and In So Kweon. Cbam: Convolutional block attention module, 2018. 4

    2018

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.