Deep Learning with Self-Attention and Enhanced Preprocessing for Precise Diagnosis of Acute Lymphoblastic Leukemia from Bone Marrow Smears in Hemato-Oncology

Bishowjit Paul; Md.Mahbubul Haque; Md. Maruf

REVIEW 4 major objections 2 minor 65 references

Deep Learning with Self-Attention and Enhanced Preprocessing for Precise Diagnosis of Acute Lymphoblastic Leukemia from Bone Marrow Smears in Hemato-Oncology

T0 review · 4 major / 2 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read An enhanced VGG19 with multi-head self-attention and focal-loss training classifies acute lymphoblastic leukemia from bone marrow smear images at 99.25% accuracy, beating a ResNet101 baseline at 98.62%.

desk verdict A small empirical delta (99.25% vs 98.62%) that cannot be evaluated: the abstract omits the evaluation protocol, and the full text is garbled. read the letter →

arxiv 2508.17216 v1 pith:TQCIFVA2 submitted 2025-08-24 cs.CV cs.LG

Md. Maruf , Md.Mahbubul Haque , Bishowjit Paul This is my paper · ORCID

classification cs.CVcs.LG

keywords acutelymphoblasticleukemiabonemarrowsmeardeeplearningmulti-headself-attentionVGG19focallossmedicalimageclassification

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

The paper is trying to establish that a relatively light convolutional network—VGG19—can outperform a heavier ResNet101 on the clinical task of diagnosing acute lymphoblastic leukemia from bone marrow smear images, once it is augmented with a multi-head self-attention block and trained with Focal Loss. The proposed pipeline standardizes image quality through preprocessing and uses attention to capture long-range relationships among cellular features. If correct, the result means attention-augmented CNNs give more discriminative representations of leukemic morphology while remaining computationally efficient. That would make automated ALL recognition and subtyping a practical support tool in diagnostic workflows.

What carries the argument

The central mechanism is the multi-head self-attention block inserted into a VGG19 convolutional backbone, together with Focal Loss as the training objective. Self-attention lets each image region attend to all other regions, modeling long-range dependencies and contextual relationships among cellular features; Focal Loss reweights the loss so the model focuses on hard, underrepresented examples rather than easy majority-class ones. A preprocessing pipeline standardizes image quality before these components operate.

What would settle it

Re-run the same preprocessing and attention-augmented model on an external, multi-institution bone marrow smear dataset with a patient-level split and per-subtype labels. The central claim is supported if accuracy stays near 99% and the MHSA variant still beats ResNet101; it is undermined if the margin collapses or accuracy falls below the ResNet101 baseline.

Watch

Extended reading notes

Core claim

The paper's central claim is that inserting a multi-head self-attention (MHSA) block into a VGG19 backbone, combined with a robust preprocessing pipeline and training under Focal Loss, yields a model that reaches 99.25% accuracy for acute lymphoblastic leukemia diagnosis from bone marrow smear images. This surpasses a strong ResNet101 baseline that reaches 98.62% accuracy on the same task. The attention block is presented as the key design element: it models long-range dependencies and contextual relationships among cellular features that plain convolutional filters may miss, while Focal Loss mitigates class imbalance by down-weighting easy examples. The authors argue that this combination p

Load-bearing premise

The load-bearing premise is that the evaluation dataset is representative of real clinical bone marrow smears: the paper reports one accuracy number without describing the train/test split, patient-level separation, or external validation, so the claim of clinical utility rests on that representativeness.

Editorial extensions

If this is right

A VGG19+MHSA model trained with Focal Loss can serve as a high-accuracy screening and subtyping tool for ALL from bone marrow smears.
Attention-augmented CNNs offer a computationally cheaper alternative to larger architectures while matching or exceeding their accuracy.
Preprocessing plus loss rebalancing contributes meaningful gains beyond the choice of backbone architecture.
The method has the potential to accelerate diagnostic workflows and support clinician decision-making in hematology settings.

Reading between the lines

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

The reported 0.63-percentage-point gain over ResNet101 is a single accuracy number; without confidence intervals, repeated runs, or statistical significance testing, I would not treat that margin as settled.
If the improvement reproduces on external data, the same recipe—pretrained CNN plus self-attention plus focal loss—is a low-cost thing to try on other cytology and hematologic malignancy classification tasks where class imbalance and context-dependent morphology matter.
A direct ablation of the preprocessing step, on and off, would clarify how much of the gain comes from image standardization versus the attention block itself.

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.

Desk Editor's Note

read the letter

Two things to know before you spend time on this: the central claim is one accuracy number with no evaluation details, and the full text is unreadable (mojibake), so I could only work from the abstract. That limits how much credit or blame I can assign, but the abstract itself is enough to see where the soft spots are.

What is legitimate: the authors combine known, sensible pieces—VGG19 plus a multi-head self-attention block, trained with focal loss—and they benchmark against a strong ResNet101 baseline. They also describe a preprocessing pipeline and claim computational efficiency. These are reasonable engineering choices, and reporting a baseline is more than many papers in this space do. No code, data, or metrics beyond accuracy are given, so there is nothing formally reproducible to lean on.

The load-bearing problem is the evaluation. The abstract gives no train/test split, no patient-level separation, no repeated seeds, and no confidence intervals. The reported 0.63-point advantage over ResNet101 is smaller than run-to-run variance for CNNs near saturation, so it may be noise. If images from the same patient landed in both training and test sets, the accuracy figure is inflated. The authors say class imbalance motivated focal loss, yet they report only overall accuracy—on an imbalanced test set, a model can hit 99% by perfecting the majority class while missing rare ALL subtypes. The clinical closing claim ('reliable decision-making') is not supported by a single accuracy number without per-class sensitivity and specificity.

Novelty is also thin: dozens of prior papers report >95% accuracy on similar public ALL datasets, and the specific combination of attention and focal loss is established in medical imaging. This is a legitimate but incremental engineering paper, not a new result or a new dataset.

Who benefits from reading it: someone tracking common failure modes in clinical ML evaluations, or a reviewer wanting a checklist of what to demand from ALL classification papers. It does not deserve a serious referee in its current form—not because the idea is bad, but because the manuscript as submitted is unreadable and the abstract lacks the evidence needed to assess even the primary claim. If the authors resubmit with the full text intact, a proper evaluation protocol, and at least sensitivity/specificity, it could be a minor conference or journal paper. As it stands, desk rejection is the right call.

Recommendation: don't cite it, don't discuss it in reading group; if you're an editor, send it back for a complete rewrite and proper evaluation before any peer review.

Referee Report

4 major / 2 minor

Summary. The manuscript proposes an automated ALL diagnosis framework combining a preprocessing pipeline, a VGG19 backbone augmented with a multi-head self-attention block, and Focal Loss training. The abstract reports 99.25% accuracy for VGG19+MHSA versus 98.62% for a ResNet101 baseline on bone marrow smear images. However, the full text is severely corrupted by character-encoding errors, leaving the Methods, Results, and Tables unreadable; only the abstract and fragmented figure/table labels are intelligible. The central empirical claim therefore cannot be checked against the actual reported experiments.

Significance. If the claimed evaluation were robust, the contribution would be a modest but useful engineering result: attention augmentation plus Focal Loss improves near-saturated accuracy over a strong CNN baseline. The manuscript has a clear architecture comparison and an explicit rationale for using Focal Loss. However, the paper currently provides no reproducible code, no dataset description, no explicit split protocol, no confidence intervals, and no per-class metrics. The reported 0.63-point advantage is too small to interpret without information about run-to-run variance, patient-level separation, and class balance. The clinical generalization claim in the abstract is unsupported by the evidence provided.

major comments (4)

[Abstract and Results (Tables 1-4)] The entire evaluation is summarized by a single accuracy number (99.25% vs 98.62%) with no dataset size, train/val/test split, patient-level separation, number of random seeds, or confidence intervals. Near saturation, a 0.63-point difference is within typical run-to-run variation for CNNs. The paper must state the exact split protocol, whether cells from the same patient appear in both training and test sets, and report mean and standard deviation over multiple runs, or the claimed advantage over ResNet101 is not statistically grounded.
[Results / class-imbalance discussion] Focal Loss is introduced to mitigate class imbalance, yet only aggregate accuracy is reported. On an imbalanced test set, a model can achieve high overall accuracy while performing poorly on rare ALL subtypes. The manuscript needs a confusion matrix and per-class sensitivity/specificity, especially for any underrepresented subtype, before the 'precise diagnosis' claim can be accepted.
[Full text (all sections after the abstract)] The body of the paper is unreadable due to character corruption; equations, figure captions, and table cell values are garbled placeholders. This is not a presentation issue: it prevents verification of the network architecture, preprocessing steps, hyperparameters, dataset provenance, and the numerical results themselves. The authors must resubmit a readable version with a complete Methods and Results section. Without this, no reviewer can assess the validity of the reported accuracy.
[Abstract, clinical claim] The abstract's closing statement that the tool can 'support reliable decision-making in clinical settings' goes beyond the measured numbers. No external validation, site diversity, or domain-shift analysis is described. The clinical claim should be removed or clearly qualified until such evidence is provided.

minor comments (2)

[Title/Abstract] The title promises 'precise diagnosis' and the abstract mentions 'precise subtyping,' but no subtype-level results are reported in the available text. The authors should either report subtype-level metrics or adjust the wording to reflect overall classification performance.
[General] No dataset name, count of images/patients, or data availability statement is legible in the supplied text. Please include these details in the revision, as they are essential for reproducibility and for assessing representativeness.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: empirical benchmark study with no derivation chain that reduces to its inputs.

full rationale

This paper is an empirical classification study, not a derivation. The central claim is that VGG19+MHSA with Focal Loss reaches 99.25% accuracy on a bone marrow smear dataset, compared to a ResNet101 baseline at 98.62%. That accuracy is a measured test-set outcome, not a quantity constructed from the input or from a fitted parameter renamed as a prediction. There is no equation-to-equation dependency, no self-citation used to justify the core result, and no uniqueness theorem or imported ansatz that makes the conclusion true by definition. The preprocessing, attention block, and Focal Loss are model choices whose effect is evaluated empirically rather than derived from the target accuracy. Concerns about the evaluation protocol—such as missing split details, patient-level separation, or confidence intervals—are correctness and reporting risks, not circularity. The paper is self-contained as an experimental claim; it does not define its outcome into existence or rely on a citation chain to force its conclusion.

Assumptions & free parameters 2 free parameters · 2 assumptions · 0 invented entities

The central claim depends on unspecified training hyperparameters and on the assumption that the evaluation dataset is correctly labeled and representative. No new entities are introduced.

free parameters (2)

Focal loss gamma = not reported
The focal loss hyperparameter that trades off easy vs hard examples is a hand-chosen value that directly affects the learned model and the resulting accuracy.
Learning rate and training schedule = not reported
Training hyperparameters are selected by the authors and influence the final weights.

assumptions (2)

domain assumption The dataset labels are correct (ground truth reflects true ALL status)
The accuracy metric assumes the reference labels used for training and evaluation are accurate pathologist annotations, which is not discussed in the abstract.
domain assumption Bone marrow smear images in the evaluation set are representative of clinical images
The abstract claims clinical decision support potential, which assumes distributional similarity between the test set and real-world smears from different scanners and centers.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Deep Learning with Self-Attention and Enhanced Preprocessing for Precise Diagnosis of Acute Lymphoblastic Leukemia from Bone Marrow Smears in Hemato-Oncology." pith.science (2026). https://pith.science/paper/TQCIFVA2

@misc{pith2026250817216,
  author       = {Pith},
  title        = {Pith review of: Deep Learning with Self-Attention and Enhanced Preprocessing for Precise Diagnosis of Acute Lymphoblastic Leukemia from Bone Marrow Smears in Hemato-Oncology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TQCIFVA2}},
  note         = {Machine review of arXiv:2508.17216}
}

read the original abstract

Acute lymphoblastic leukemia (ALL) is a prevalent hematological malignancy in both pediatric and adult populations. Early and accurate detection with precise subtyping is essential for guiding therapy. Conventional workflows are complex, time-consuming, and prone to human error. We present a deep learning framework for automated ALL diagnosis from bone marrow smear images. The method combines a robust preprocessing pipeline with convolutional neural networks (CNNs) to standardize image quality and improve inference efficiency. As a key design, we insert a multi-head self-attention (MHSA) block into a VGG19 backbone to model long-range dependencies and contextual relationships among cellular features. To mitigate class imbalance, we train with Focal Loss. Across evaluated architectures, the enhanced VGG19+MHSA trained with Focal Loss achieves 99.25% accuracy, surpassing a strong ResNet101 baseline (98.62%). These results indicate that attention-augmented CNNs, coupled with targeted loss optimization and preprocessing, yield more discriminative representations of leukemic cell morphology. Our approach offers a highly accurate and computationally efficient tool for automated ALL recognition and subtyping, with potential to accelerate diagnostic workflows and support reliable decision-making in clinical settings.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

65 extracted references · 63 canonical work pages

[1]

T cell acute lymphoblastic leukemia (t-all): new insights into the cellular origins and infiltration mechanisms common and unique among hematologic malignancies

Eduardo Vadillo, Elisa Dorantes-Acosta, Rosana Pelayo, and Michael Schnoor. T cell acute lymphoblastic leukemia (t-all): new insights into the cellular origins and infiltration mechanisms common and unique among hematologic malignancies. Blood reviews , 32(1):36--51, 2018

work page 2018
[2]

A hematologist-level deep learning algorithm (bmsnet) for assessing the morphologies of single nuclear balls in bone marrow smears: algorithm development

Yi-Ying Wu, Tzu-Chuan Huang, Ren-Hua Ye, Wen-Hui Fang, Shiue-Wei Lai, Ping-Ying Chang, Wei-Nung Liu, Tai-Yu Kuo, Cho-Hao Lee, Wen-Chiuan Tsai, et al. A hematologist-level deep learning algorithm (bmsnet) for assessing the morphologies of single nuclear balls in bone marrow smears: algorithm development. JMIR medical informatics , 8(4):e15963, 2020

work page 2020
[3]

Timely diagnosis of acute lymphoblastic leukemia using artificial intelligence-oriented deep learning methods

Sorayya Rezayi, Niloofar Mohammadzadeh, Hamid Bouraghi, Soheila Saeedi, Ali Mohammadpour, et al. Timely diagnosis of acute lymphoblastic leukemia using artificial intelligence-oriented deep learning methods. Computational Intelligence and Neuroscience , 2021, 2021

work page 2021
[4]

Leukemia, lymphoma, and myeloma

Alan Rosmarin. Leukemia, lymphoma, and myeloma. Cancer: Prevention, early detection, treatment and recovery , pages 299--316, 2019

work page 2019
[5]

Customized deep learning classifier for detection of acute lymphoblastic leukemia using blood smear images

Niranjana Sampathila, Krishnaraj Chadaga, Neelankit Goswami, Rajagopala P Chadaga, Mayur Pandya, Srikanth Prabhu, Muralidhar G Bairy, Swathi S Katta, Devadas Bhat, and Sudhakara P Upadya. Customized deep learning classifier for detection of acute lymphoblastic leukemia using blood smear images. In Healthcare , volume 10, page 1812. MDPI, 2022

work page 2022
[6]

Pattern recognition of acute lymphoblastic leukemia (all) using computational deep learning

Nasmin Jiwani, Ketan Gupta, Giovanni Pau, and Mohammad Alibakhshikenari. Pattern recognition of acute lymphoblastic leukemia (all) using computational deep learning. IEEE Access , 11:29541--29553, 2023

work page 2023
[7]

a ger, and Joachim Sch \

Felix M Onyije, Ann Olsson, Dan Baaken, Friederike Erdmann, Martin Stanulla, Daniel Wollschl \"a ger, and Joachim Sch \"u z. Environmental risk factors for childhood acute lymphoblastic leukemia: an umbrella review. Cancers , 14(2):382, 2022

work page 2022
[8]

Genetic biomarkers and their clinical implications in b-cell acute lymphoblastic leukemia in children

Monika Lejman, Aleksandra Cha upnik, Zuzanna Chilimoniuk, and Maciej Dobosz. Genetic biomarkers and their clinical implications in b-cell acute lymphoblastic leukemia in children. International Journal of Molecular Sciences , 23(5):2755, 2022

work page 2022

Show all 65 references

[9]

Asparaginase in the treatment of acute lymphoblastic leukemia in adults: Current evidence and place in therapy

Krishna R Juluri, Chloe Siu, and Ryan D Cassaday. Asparaginase in the treatment of acute lymphoblastic leukemia in adults: Current evidence and place in therapy. Blood and Lymphatic Cancer: Targets and Therapy , pages 55--79, 2022

2022
[10]

Cancer statistics, 2023

Rebecca L Siegel, Kimberly D Miller, Nikita Sandeep Wagle, and Ahmedin Jemal. Cancer statistics, 2023. Ca Cancer J Clin , 73(1):17--48, 2023

2023
[11]

Use of car t-cell for acute lymphoblastic leukemia (all) treatment: a review study

Mohsen Sheykhhasan, Hamed Manoochehri, and Paola Dama. Use of car t-cell for acute lymphoblastic leukemia (all) treatment: a review study. Cancer gene therapy , 29(8-9):1080--1096, 2022

2022
[12]

A novel deep learning framework (dlf) for classification of acute lymphoblastic leukemia

Sunita Chand and Virendra P Vishwakarma. A novel deep learning framework (dlf) for classification of acute lymphoblastic leukemia. Multimedia Tools and Applications , 81(26):37243--37262, 2022

2022
[13]

Peripheral bone relapse of paediatric tcf3-hlf positive acute lymphoblastic leukaemia during haematopoietic stem cell transplantation: A case report

M \'a t \'e Horv \'a th, Gabriella Kert \'e sz, Csaba Kassa, Vera Goda, Kata Csord \'a s, Lidia Hau, Anna K \"o v \'e r, Anita Str \'e hn, Orsolya Horv \'a th, Kriszti \'a n K \'a llay, et al. Peripheral bone relapse of paediatric tcf3-hlf positive acute lymphoblastic leukaemi...

1919
[14]

Automated detection of b cell and t cell acute lymphoblastic leukaemia using deep learning

KK Anilkumar, VJ Manoj, and TM Sagi. Automated detection of b cell and t cell acute lymphoblastic leukaemia using deep learning. Irbm , 43(5):405--413, 2022

2022
[15]

Interactions of b-lymphocytes and bone cells in health and disease

Danka Gr c evi \'c , Archana Sanjay, and Joseph Lorenzo. Interactions of b-lymphocytes and bone cells in health and disease. Bone , 168:116296, 2023

2023
[16]

Flt3 signaling in b lymphocyte development and humoral immunity

Kay L Medina. Flt3 signaling in b lymphocyte development and humoral immunity. International journal of molecular sciences , 23(13):7289, 2022

2022
[17]

T cell development and t cell receptor structure

Tobili Y Sam-Yellowe and Tobili Y Sam-Yellowe. T cell development and t cell receptor structure. Immunology: overview and laboratory manual , pages 105--116, 2021

2021
[18]

Comparative features and outcomes between paediatric t-cell and b-cell acute lymphoblastic leukaemia

David T Teachey and Ching-Hon Pui. Comparative features and outcomes between paediatric t-cell and b-cell acute lymphoblastic leukaemia. The Lancet Oncology , 20(3):e142--e154, 2019

2019
[19]

Artificial intelligence with multi-functional machine learning platform development for better healthcare and precision medicine

Zeeshan Ahmed, Khalid Mohamed, Saman Zeeshan, and XinQi Dong. Artificial intelligence with multi-functional machine learning platform development for better healthcare and precision medicine . Database , 2020:baaa010, 03 2020

2020
[20]

Deep learning in medical image analysis

Heang-Ping Chan, Ravi K Samala, Lubomir M Hadjiiski, and Chuan Zhou. Deep learning in medical image analysis. Deep Learning in Medical Image Analysis: Challenges and Applications , pages 3--21, 2020

2020
[21]

Zubair Ahmad, Shabina Rahim, Maha Zubair, and Jamshid Abdul-Ghafar. Artificial intelligence (ai) in medicine, current applications and future role with special emphasis on its potential and promise in pathology: present and future impact, obstacles including costs and acceptan...

2021
[22]

A systematic survey of computer-aided diagnosis in medicine: Past and present developments

Juri Yanase and Evangelos Triantaphyllou. A systematic survey of computer-aided diagnosis in medicine: Past and present developments. Expert Systems with Applications , 138:112821, 2019

2019
[23]

Redefining radiology: a review of artificial intelligence integration in medical imaging

Reabal Najjar. Redefining radiology: a review of artificial intelligence integration in medical imaging. Diagnostics , 13(17):2760, 2023

2023
[24]

A scoping review of transfer learning research on medical image analysis using imagenet

Mohammad Amin Morid, Alireza Borjali, and Guilherme Del Fiol. A scoping review of transfer learning research on medical image analysis using imagenet. Computers in biology and medicine , 128:104115, 2021

2021
[25]

Bo-allcnn: Bayesian-based optimized cnn for acute lymphoblastic leukemia detection in microscopic blood smear images

Ghada Atteia, Amel A Alhussan, and Nagwan Abdel Samee. Bo-allcnn: Bayesian-based optimized cnn for acute lymphoblastic leukemia detection in microscopic blood smear images. Sensors , 22(15):5520, 2022

2022
[26]

Mutual information based hybrid model and deep learning for acute lymphocytic leukemia detection in single cell blood smear images

Krishna Kumar Jha and Himadri Sekhar Dutta. Mutual information based hybrid model and deep learning for acute lymphocytic leukemia detection in single cell blood smear images. Computer methods and programs in biomedicine , 179:104987, 2019

2019
[27]

A deep learning-based approach for the diagnosis of acute lymphoblastic leukemia

Adnan Saeed, Shifa Shoukat, Khurram Shehzad, Ijaz Ahmad, Ala’Abdulmajid Eshmawi, Ali H Amin, and Elsayed Tag-Eldin. A deep learning-based approach for the diagnosis of acute lymphoblastic leukemia. Electronics , 11(19):3168, 2022

2022
[28]

Leukemia diagnosis in blood slides using transfer learning in cnns and svm for classification

Luis HS Vogado, Rodrigo MS Veras, Flavio HD Araujo, Romuere RV Silva, and Kelson RT Aires. Leukemia diagnosis in blood slides using transfer learning in cnns and svm for classification. Engineering Applications of Artificial Intelligence , 72:415--422, 2018

2018
[29]

Development and evaluation of a leukemia diagnosis system using deep learning in real clinical scenarios

Min Zhou, Kefei Wu, Lisha Yu, Mengdi Xu, Junjun Yang, Qing Shen, Bo Liu, Lei Shi, Shuang Wu, Bin Dong, et al. Development and evaluation of a leukemia diagnosis system using deep learning in real clinical scenarios. Frontiers in Pediatrics , 9:693676, 2021

2021
[30]

An aggregated-based deep learning method for leukemic b-lymphoblast classification

Payam Hosseinzadeh Kasani, Sang-Won Park, and Jae-Won Jang. An aggregated-based deep learning method for leukemic b-lymphoblast classification. Diagnostics , 10(12):1064, 2020

2020
[31]

Method for diagnosis of acute lymphoblastic leukemia based on vit-cnn ensemble model

Zhencun Jiang, Zhengxin Dong, Lingyang Wang, and Wenping Jiang. Method for diagnosis of acute lymphoblastic leukemia based on vit-cnn ensemble model. Computational Intelligence and Neuroscience , 2021, 2021

2021
[32]

Acute lymphoblastic leukemia (all) image dataset

Mehrad Aria, Mustafa Ghaderzadeh, Davood Bashash, Hassan Abolghasemi, Farkhondeh Asadi, and Azamossadat Hosseini. Acute lymphoblastic leukemia (all) image dataset. Kaggle , 2021

2021
[33]

A review: Data pre-processing and data augmentation techniques

Kiran Maharana, Surajit Mondal, and Bhushankumar Nemade. A review: Data pre-processing and data augmentation techniques. Global Transitions Proceedings , 3(1):91--99, 2022

2022
[34]

Cnn variants for computer vision: History, architecture, application, challenges and future scope

Dulari Bhatt, Chirag Patel, Hardik Talsania, Jigar Patel, Rasmika Vaghela, Sharnil Pandya, Kirit Modi, and Hemant Ghayvat. Cnn variants for computer vision: History, architecture, application, challenges and future scope. Electronics , 10(20):2470, 2021

2021
[35]

Attention-guided deep convolutional neural networks for skin cancer classification

Arshiya Aggarwal, Nisheet Das, and Indu Sreedevi. Attention-guided deep convolutional neural networks for skin cancer classification. In 2019 Ninth International Conference on Image Processing Theory, Tools and Applications (IPTA) , pages 1--6. IEEE, 2019

2019
[36]

1-dimensional polynomial neural networks for audio signal related problems

Habib Ben Abdallah, Christopher J Henry, and Sheela Ramanna. 1-dimensional polynomial neural networks for audio signal related problems. Knowledge-Based Systems , 240:108174, 2022

2022
[37]

A transdisciplinary review of deep learning research and its relevance for water resources scientists

Chaopeng Shen. A transdisciplinary review of deep learning research and its relevance for water resources scientists. Water Resources Research , 54(11):8558--8593, 2018

2018
[38]

Transfer learning

Lisa Torrey and Jude Shavlik. Transfer learning. In Handbook of research on machine learning applications and trends: algorithms, methods, and techniques , pages 242--264. IGI global, 2010

2010
[39]

Frequency disentangled residual network

Satya Rajendra Singh, Roshan Reddy Yedla, Shiv Ram Dubey, Rakesh Sanodiya, and Wei-Ta Chu. Frequency disentangled residual network. arXiv preprint arXiv:2109.12556 , 2021

2021 arXiv
[40]

Training very deep neural networks: Rethinking the role of skip connections

Oyebade K Oyedotun, Kassem Al Ismaeil, and Djamila Aouada. Training very deep neural networks: Rethinking the role of skip connections. Neurocomputing , 441:105--117, 2021

2021
[41]

Continuous-in-depth neural networks

Alejandro F Queiruga, N Benjamin Erichson, Dane Taylor, and Michael W Mahoney. Continuous-in-depth neural networks. arXiv preprint arXiv:2008.02389 , 2020

2008 arXiv
[42]

A content-based image retrieval system for the diagnosis of lymphoma using blood micrographs: An incorporation of deep learning with a traditional learning approach

M Roy Reena and PM Ameer. A content-based image retrieval system for the diagnosis of lymphoma using blood micrographs: An incorporation of deep learning with a traditional learning approach. Computers in Biology and Medicine , 145:105463, 2022

2022
[43]

Pyramid graph networks with connection attentions for region-based one-shot semantic segmentation

Chi Zhang, Guosheng Lin, Fayao Liu, Jiushuang Guo, Qingyao Wu, and Rui Yao. Pyramid graph networks with connection attentions for region-based one-shot semantic segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 9587--9595, 2019

2019
[44]

Pruning depthwise separable convolutions for mobilenet compression

Cheng-Hao Tu, Jia-Hong Lee, Yi-Ming Chan, and Chu-Song Chen. Pruning depthwise separable convolutions for mobilenet compression. In 2020 International Joint Conference on Neural Networks (IJCNN) , pages 1--8. IEEE, 2020

2020
[45]

A new image classification approach via improved mobilenet models with local receptive field expansion in shallow layers

Wei Wang, Yiyang Hu, Ting Zou, Hongmei Liu, Jin Wang, and Xin Wang. A new image classification approach via improved mobilenet models with local receptive field expansion in shallow layers. Computational Intelligence and Neuroscience , 2020, 2020

2020
[46]

Shunt connection: An intelligent skipping of contiguous blocks for optimizing mobilenet-v2

Brijraj Singh, Durga Toshniwal, and Sharan Kumar Allur. Shunt connection: An intelligent skipping of contiguous blocks for optimizing mobilenet-v2. Neural Networks , 118:192--203, 2019

2019
[47]

P2op—plant pathology on palms: A deep learning-based mobile solution for in-field plant disease detection

Sivasubramaniam Janarthan, Selvarajah Thuseethan, Sutharshan Rajasegarar, and John Yearwood. P2op—plant pathology on palms: A deep learning-based mobile solution for in-field plant disease detection. Computers and Electronics in Agriculture , 202:107371, 2022

2022
[48]

When mobilenetv2 meets transformer: A balanced sheep face recognition model

Xiaopeng Li, Jinzhi Du, Jialin Yang, and Shuqin Li. When mobilenetv2 meets transformer: A balanced sheep face recognition model. Agriculture , 12(8):1126, 2022

2022
[49]

Deep learning approach in gregg shorthand word to english-word conversion

Dionis A Padilla, Nicole Kim U Vitug, and Julius Benito S Marquez. Deep learning approach in gregg shorthand word to english-word conversion. In 2020 IEEE 5th International Conference on Image, Vision and Computing (ICIVC) , pages 204--210. IEEE, 2020

2020
[50]

Convolutional neural network and its pretrained models for image classification and object detection: A survey

Biswajit Jena, Gopal Krishna Nayak, and Sanjay Saxena. Convolutional neural network and its pretrained models for image classification and object detection: A survey. Concurrency and Computation: Practice and Experience , 34(6):e6767, 2022

2022
[51]

Micro-architecture design exploration template for automl case study on squeezesemauto

Chantana Chantrapornchai, Supasit Kajkamhaeng, and Phattharaphon Romphet. Micro-architecture design exploration template for automl case study on squeezesemauto. Scientific Reports , 13(1):10642, 2023

2023
[52]

Image classification based on inception-v3 and a mixture of handcrafted features

A Shubha Rao and K Mahantesh. Image classification based on inception-v3 and a mixture of handcrafted features. In Distributed Computing and Optimization Techniques: Select Proceedings of ICDCOT 2021 , pages 527--537. Springer, 2022

2021
[53]

Cnn based fundus images classification for glaucoma identification

Thisara Shyamalee and Dulani Meedeniya. Cnn based fundus images classification for glaucoma identification. In 2022 2nd International Conference on Advanced Research in Computing (ICARC) , pages 200--205. IEEE, 2022

2022
[54]

Damage detection using in-domain and cross-domain transfer learning

Zaharah A Bukhsh, Nils Jansen, and Aaqib Saeed. Damage detection using in-domain and cross-domain transfer learning. Neural Computing and Applications , 33(24):16921--16936, 2021

2021
[55]

Towards indoor suctionable object classification and recycling: Developing a lightweight ai model for robot vacuum cleaners

Qian Huang. Towards indoor suctionable object classification and recycling: Developing a lightweight ai model for robot vacuum cleaners. Applied Sciences , 13(18):10031, 2023

2023
[56]

A survey on deep learning in medical image analysis

Geert Litjens et al. A survey on deep learning in medical image analysis. Medical Image Analysis , 42:60--88, 2017

2017
[57]

Gan-based synthetic medical image augmentation for increased cnn performance in liver lesion classification

Maayan Frid-Adar et al. Gan-based synthetic medical image augmentation for increased cnn performance in liver lesion classification. Neurocomputing , 321:321--331, 2018

2018
[58]

Diffusion models for medical imaging

Armin Kazerouni et al. Diffusion models for medical imaging. NeurIPS Medical Imaging Workshop , 2022

2022
[59]

Big data analytics for preventive medicine

Muhammad Imran Razzak et al. Big data analytics for preventive medicine. Neural Computing and Applications , 32:923--939, 2020

2020
[60]

Eric J. Topol. High-performance medicine: the convergence of human and artificial intelligence. Nature Medicine , 25:44--56, 2019

2019
[61]

Scalable ai deployment using cloud computing

Ramesh Krishnan and Brian Williams. Scalable ai deployment using cloud computing. IEEE Cloud Computing , 7(2):56--63, 2020

2020
[62]

Searching for mobilenetv3

Andrew Howard et al. Searching for mobilenetv3. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 1314--1324, 2019

2019
[63]

Opportunities and obstacles for deep learning in biology and medicine

Travers Ching et al. Opportunities and obstacles for deep learning in biology and medicine. Journal of the Royal Society Interface , 15(141), 2018

2018
[64]

What do we need to build explainable ai systems for the medical domain? Journal of Artificial Intelligence Research , 71:593--626, 2021

Andreas Holzinger et al. What do we need to build explainable ai systems for the medical domain? Journal of Artificial Intelligence Research , 71:593--626, 2021

2021
[65]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

T cell acute lymphoblastic leukemia (t-all): new insights into the cellular origins and infiltration mechanisms common and unique among hematologic malignancies

Eduardo Vadillo, Elisa Dorantes-Acosta, Rosana Pelayo, and Michael Schnoor. T cell acute lymphoblastic leukemia (t-all): new insights into the cellular origins and infiltration mechanisms common and unique among hematologic malignancies. Blood reviews , 32(1):36--51, 2018

work page 2018

[2] [2]

A hematologist-level deep learning algorithm (bmsnet) for assessing the morphologies of single nuclear balls in bone marrow smears: algorithm development

Yi-Ying Wu, Tzu-Chuan Huang, Ren-Hua Ye, Wen-Hui Fang, Shiue-Wei Lai, Ping-Ying Chang, Wei-Nung Liu, Tai-Yu Kuo, Cho-Hao Lee, Wen-Chiuan Tsai, et al. A hematologist-level deep learning algorithm (bmsnet) for assessing the morphologies of single nuclear balls in bone marrow smears: algorithm development. JMIR medical informatics , 8(4):e15963, 2020

work page 2020

[3] [3]

Timely diagnosis of acute lymphoblastic leukemia using artificial intelligence-oriented deep learning methods

Sorayya Rezayi, Niloofar Mohammadzadeh, Hamid Bouraghi, Soheila Saeedi, Ali Mohammadpour, et al. Timely diagnosis of acute lymphoblastic leukemia using artificial intelligence-oriented deep learning methods. Computational Intelligence and Neuroscience , 2021, 2021

work page 2021

[4] [4]

Leukemia, lymphoma, and myeloma

Alan Rosmarin. Leukemia, lymphoma, and myeloma. Cancer: Prevention, early detection, treatment and recovery , pages 299--316, 2019

work page 2019

[5] [5]

Customized deep learning classifier for detection of acute lymphoblastic leukemia using blood smear images

Niranjana Sampathila, Krishnaraj Chadaga, Neelankit Goswami, Rajagopala P Chadaga, Mayur Pandya, Srikanth Prabhu, Muralidhar G Bairy, Swathi S Katta, Devadas Bhat, and Sudhakara P Upadya. Customized deep learning classifier for detection of acute lymphoblastic leukemia using blood smear images. In Healthcare , volume 10, page 1812. MDPI, 2022

work page 2022

[6] [6]

Pattern recognition of acute lymphoblastic leukemia (all) using computational deep learning

Nasmin Jiwani, Ketan Gupta, Giovanni Pau, and Mohammad Alibakhshikenari. Pattern recognition of acute lymphoblastic leukemia (all) using computational deep learning. IEEE Access , 11:29541--29553, 2023

work page 2023

[7] [7]

a ger, and Joachim Sch \

Felix M Onyije, Ann Olsson, Dan Baaken, Friederike Erdmann, Martin Stanulla, Daniel Wollschl \"a ger, and Joachim Sch \"u z. Environmental risk factors for childhood acute lymphoblastic leukemia: an umbrella review. Cancers , 14(2):382, 2022

work page 2022

[8] [8]

Genetic biomarkers and their clinical implications in b-cell acute lymphoblastic leukemia in children

Monika Lejman, Aleksandra Cha upnik, Zuzanna Chilimoniuk, and Maciej Dobosz. Genetic biomarkers and their clinical implications in b-cell acute lymphoblastic leukemia in children. International Journal of Molecular Sciences , 23(5):2755, 2022

work page 2022