{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,10]],"date-time":"2026-10-10T01:51:55Z","timestamp":1791597115779,"version":"4.3.4"},"reference-count":41,"publisher":"MDPI AG","issue":"19","license":[{"start":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T00:00:00Z","timestamp":1791417600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T00:00:00Z","timestamp":1791417600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Int. J. Mol. Sci.","IJMS"],"abstract":"<jats:p>Automated analysis of cystoscopic images may support consistent characterization of bladder tissue patterns, but reported performance is highly sensitive to patient composition, data partitioning, and evaluation methodology. Objective: This study provides an exploratory comparison of convolutional neural networks, Transformer-based models, and a proposed Local\u2013Global Hybrid Transformer (LGH-Transformer) for four-class bladder tissue image classification. The dataset comprised 1754 frames and used the publicly released 1257\/308\/189 training\u2013validation\u2013held-out test split. The annotation-defined held-out test composition was HGC = 74, LGC = 53, NST = 37, and NTL = 25 (n = 189). Images were resized to 224 \u00d7 224 pixels, normalized, and augmented only in the training subset. Adam and RMSprop were evaluated. Accuracy and macro-averaged precision, recall, and F1-score were calculated from the integer-count evaluation matrices. The highest held-out accuracy, 187\/189 (0.9894), was reached by seven model\u2013optimizer configurations: DenseNet201\u2013RMSprop, XCiT Medium 24 P16\u2013Adam, ConvNeXt Base\u2013Adam, CoAtNet\u2013Adam, ConvNeXt Base\u2013RMSprop, CoAtNet\u2013RMSprop, and the proposed LGH-Transformer\u2013RMSprop; MobileNet\u2013Adam and Xcep-tion\u2013RMSprop each yielded 185\/189 (0.9788), and LGH-Transformer\u2013Adam yielded 186\/189 (0.9841). On a 189-frame test set, differences in one to two frames do not establish statistical superiority, so these configurations are reported as a top-accuracy tier rather than as a ranked superiority claim. Within this tier, the LGH-Transformer had the smallest computational footprint (\u224818.2 M parameters and \u22484.1 GFLOPs, versus 20.0\u201388.6 M parameters and 4.3\u201316.2 GFLOPs for the other top-tier models). Component-wise ablation experiments showed accuracy decreases of 0.0106, 0.0053, and 0.0053 after removing the CNN tokenizer, local attention, and global attention, respectively. The LGH-Transformer reaches the top accuracy tier of this 18-architecture benchmark at the lowest parameter and FLOP budget within that tier, supporting an internally consistent, efficiency-oriented comparison within the available cohort. The absence of task-aligned external validation and the small number of held-out patients require the findings to be interpreted as exploratory internal validation rather than evidence of clinical generalizability.<\/jats:p>","DOI":"10.3390\/ijms27198929","type":"journal-article","created":{"date-parts":[[2026,10,10]],"date-time":"2026-10-10T01:40:17Z","timestamp":1791596417000},"page":"8929","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["LGH-Transformer: A Hybrid CNN\u2013Transformer Model for Multi-Class Bladder Tissue Classification from Cystoscopic Images"],"prefix":"10.3390","volume":"27","author":[{"given":"Hasan An\u0131l","family":"Kurt","sequence":"first","affiliation":[{"name":"Department of Urology, Faculty of Medicine, \u00c7anakkale Onsekiz Mart University, 17100 \u00c7anakkale, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-9299-7141","authenticated-orcid":false,"given":"Sabire","family":"K\u0131l\u0131\u00e7arslan","sequence":"additional","affiliation":[{"name":"Department of Medical System Biology, Graduate School of Sciences, \u00c7anakkale Onsekiz Mart University, 17100 \u00c7anakkale, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4303-9717","authenticated-orcid":false,"given":"Meliha Merve","family":"\u00c7i\u00e7ekliyurt","sequence":"additional","affiliation":[{"name":"Department of Medical Biology, Faculty of Medicine, \u00c7anakkale Onsekiz Mart University, 17100 \u00c7anakkale, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9483-4425","authenticated-orcid":false,"given":"Serhat","family":"Kili\u00e7arslan","sequence":"additional","affiliation":[{"name":"Department of Software Engineering, Faculty of Engineering, Bandirma Onyedi Eyl\u00fcl University, 10200 Bal\u0131kesir, Turkey"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,10,8]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"58","DOI":"10.1038\/s41572-023-00468-9","article-title":"Bladder cancer","volume":"9","author":"Hansel","year":"2023","journal-title":"Nat. Rev. Dis. Primers"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"216","DOI":"10.6004\/jnccn.2024.0024","article-title":"NCCN Guidelines\u00ae Insights: Bladder Cancer, Version 3.2024","volume":"22","author":"Flaig","year":"2024","journal-title":"J. Natl. Compr. Cancer Netw."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"531","DOI":"10.1016\/j.hoc.2021.02.004","article-title":"Diagnosis and Staging of Bladder Cancer","volume":"35","author":"Ahmadi","year":"2021","journal-title":"Hematol. Clin. North Am."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"3056","DOI":"10.21037\/tau.2019.09.38","article-title":"Transurethral resection of bladder tumour (TURBT)","volume":"9","author":"Kim","year":"2020","journal-title":"Transl. Androl. Urol."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"115","DOI":"10.1016\/j.euf.2023.08.002","article-title":"The Diagnostic Accuracy of Cystoscopy for Detecting Bladder Cancer in Adults Presenting with Haematuria: A Systematic Review from the European Association of Urology Guidelines Office","volume":"10","author":"Devlies","year":"2024","journal-title":"Eur. Urol. Focus"},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Guldhammer, C.S., V\u00e1squez, J.L., Kristensen, V.M., Norus, T., Nadler, N., Jensen, J.B., and Azawi, N. (2023). Cystoscopy Accuracy in Detecting Bladder Tumors: A Prospective Video-Confirmed Study. Cancers, 16.","DOI":"10.3390\/cancers16010160"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"17","DOI":"10.1016\/j.euros.2021.06.011","article-title":"Time to Turn on the Blue Lights: A Systematic Review and Meta-analysis of Photodynamic Diagnosis for Bladder Cancer","volume":"31","author":"Veeratterapillay","year":"2021","journal-title":"Eur. Urol. Open Sci."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"307","DOI":"10.1016\/j.purol.2023.03.003","article-title":"Photodynamic cystoscopy for bladder cancer diagnosis and for NMIBC follow-up: An overview of systematic reviews and meta-analyses","volume":"33","author":"Soorojebally","year":"2023","journal-title":"Prog. Urol."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"890","DOI":"10.1111\/bju.16471","article-title":"Advanced optical imaging techniques for bladder cancer detection and diagnosis: A systematic review","volume":"134","author":"Remmelink","year":"2024","journal-title":"BJU Int."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"1030","DOI":"10.1089\/end.2020.0919","article-title":"Cystoscopic Imaging for Bladder Cancer Detection Based on Stepwise Organic Transfer Learning with a Pretrained Convolutional Neural Network","volume":"35","author":"Ikeda","year":"2021","journal-title":"J. Endourol."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"96","DOI":"10.1186\/s12938-023-01157-0","article-title":"A survey of Transformer applications for histopathological image analysis: New developments and future directions","volume":"22","author":"Atabansi","year":"2023","journal-title":"Biomed. Eng. Online"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"20639","DOI":"10.1038\/s41598-023-47992-y","article-title":"The classification of the bladder cancer based on Vision Transformers (ViT)","volume":"13","author":"Khedr","year":"2023","journal-title":"Sci. Rep."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"e36144","DOI":"10.1016\/j.heliyon.2024.e36144","article-title":"Development of deep learning model for diagnosing muscle-invasive bladder cancer on MRI with vision transformer","volume":"10","author":"Kurata","year":"2024","journal-title":"Heliyon"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"11629","DOI":"10.1038\/s41598-021-91081-x","article-title":"Deep learning-based classification of blue light cystoscopy imaging during transurethral resection of bladder tumors","volume":"11","author":"Ali","year":"2021","journal-title":"Sci. Rep."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Devi, M.S., Pandian, J.A., Umanandhini, D., and Vishal, G.P. (2024). Endoscopic Bladder Tissue Classification Using Seventeen Layered Deep Convolutional Neural Network. 2024 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC), IEEE. Available online: https:\/\/ieeexplore.ieee.org\/abstract\/document\/10507996\/.","DOI":"10.1109\/ASSIC60049.2024.10507996"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Amaouche, M., Karrakchou, O., Ghogho, M., El Ghazzaly, A., Alami, M., and Ameur, A. (2024). Redefining cystoscopy with AI: Bladder cancer diagnosis using an efficient hybrid CNN-transformer model. 2024 IEEE International Conference on Image Processing (ICIP), IEEE. Available online: https:\/\/ieeexplore.ieee.org\/abstract\/document\/10647282\/.","DOI":"10.1109\/ICIP51287.2024.10647282"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"103348","DOI":"10.1016\/j.media.2024.103348","article-title":"Will Transformers change gastrointestinal endoscopic image analysis? A comparative analysis between CNNs and Transformers, in terms of performance, robustness and generalization","volume":"99","author":"Kusters","year":"2025","journal-title":"Med. Image Anal."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Chou, C.-K., Nguyen, H.-T., Wang, Y.-K., Chen, T.-H., Wu, I.-C., Huang, C.-W., and Wang, H.-C. (2023). Preparing well for esophageal endoscopic detection using a hybrid model and transfer learning. Cancers, 15.","DOI":"10.3390\/cancers15153783"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"95","DOI":"10.25299\/itjrd.2025.17867","article-title":"CNN-based classification of bladder tissue lesions from endoscopy images","volume":"9","author":"Lutviana","year":"2025","journal-title":"IT J. Res. Dev."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"vbag087","DOI":"10.1093\/bioadv\/vbag087","article-title":"DeepBCTPred: Deep learning-based prediction of bladder cancer tissues from endoscopic images","volume":"6","author":"Nafi","year":"2026","journal-title":"Bioinform. Adv."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"El-Aziz, A.A.A., Mahmood, M.A., and El-Ghany, S.A. (2025). EfficientNet-B3-Based Automated Deep Learning Framework for Multiclass Endoscopic Bladder Tissue Classification. Diagnostics, 15.","DOI":"10.3390\/diagnostics15192515"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"1722","DOI":"10.1007\/s10278-024-01228-1","article-title":"Comparative Bladder Cancer Tissues Prediction Using Vision Transformer","volume":"38","author":"Sunnetci","year":"2025","journal-title":"J. Imaging Inform. Med."},{"key":"ref_23","unstructured":"Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., and Gelyy, S. (2021). An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. arXiv."},{"key":"ref_24","first-page":"3965","article-title":"Coatnet: Marrying convolution and attention for all data sizes","volume":"34","author":"Dai","year":"2021","journal-title":"Adv. Neural-Form. Process. Syst."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Liu, Z., Mao, H., Wu, C.-Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022). A ConvNet for the 2020s. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE.","DOI":"10.1109\/CVPR52688.2022.01167"},{"key":"ref_26","unstructured":"Mehta, S., and Rastegari, M. (2022). MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer. arXiv."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"2822","DOI":"10.1109\/TBME.2023.3265679","article-title":"Semi-Supervised Bladder Tissue Classification in Multi-Domain Endoscopic Images","volume":"70","author":"Lazo","year":"2023","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Zhang, Z. (2018). Improved adam optimizer for deep neural networks. 2018 IEEE\/ACM 26th International Symposium on Quality of Service (IWQoS), IEEE. Available online: https:\/\/ieeexplore.ieee.org\/abstract\/document\/8624183\/.","DOI":"10.1109\/IWQoS.2018.8624183"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Zou, F., Shen, L., Jie, Z., Zhang, W., and Liu, W. (2019). A sufficient condition for convergences of adam and rmsprop. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, IEEE. Available online: http:\/\/openaccess.thecvf.com\/content_CVPR_2019\/html\/Zou_A_Sufficient_Condition_for_Convergences_of_Adam_and_RMSProp_CVPR_2019_paper.html.","DOI":"10.1109\/CVPR.2019.01138"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"G\u00fcne\u015f, A. (2026). Efficient Uncertainty Quantification in Medical Imaging via Mamba State Space Models. Tomography, 12.","DOI":"10.3390\/tomography12070096"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"e70380","DOI":"10.1111\/tgis.70380","article-title":"Linear-Time Generative Representation Learning for Few-Shot Hyperspectral Image Classification","volume":"30","author":"Taskin","year":"2026","journal-title":"Trans. GIS"},{"key":"ref_32","unstructured":"Simonyan, K., and Zisserman, A. (2015). Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Chollet, F. (2017). Xception: Deep learning with depthwise separable convolutions. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE. Available online: https:\/\/ieeexplore.ieee.org\/abstract\/document\/8099678\/.","DOI":"10.1109\/CVPR.2017.195"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K.Q. (2017). Densely connected convolutional networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE. Available online: http:\/\/openaccess.thecvf.com\/content_cvpr_2017\/html\/Huang_Densely_Connected_Convolutional_CVPR_2017_paper.html.","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref_35","unstructured":"Howard, A.G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., and Adam, H. (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. arXiv."},{"key":"ref_36","unstructured":"Tan, M., and Le, Q. (2021). Efficientnetv2: Smaller models and faster training. International Conference on Machine Learning, PMLR. Available online: http:\/\/proceedings.mlr.press\/v139\/tan21a.html."},{"key":"ref_37","unstructured":"Touvron, H., Cord, M., Douze, M., Massa, F., Sablayrolles, A., and J\u00e9gou, H. (2021). Training data-efficient image transformers & dis-tillation through attention. International Conference on Machine Learning, PMLR. Available online: https:\/\/proceedings.mlr.press\/v139\/touvron21a."},{"key":"ref_38","unstructured":"Peng, Z., Dong, L., Bao, H., Ye, Q., and Wei, F. (2022). BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers. arXiv."},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Heo, B., Yun, S., Han, D., Chun, S., Choe, J., and Oh, S.J. (2021). Rethinking spatial dimensions of vision transformers. Proceedings of the IEEE\/CVF International Conference on Computer Vision, IEEE. Available online: http:\/\/openaccess.thecvf.com\/content\/ICCV2021\/html\/Heo_Rethinking_Spatial_Dimensions_of_Vision_Transformers_ICCV_2021_paper.html.","DOI":"10.1109\/ICCV48922.2021.01172"},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Chen, C.F.R., Fan, Q., and Panda, R. (2021). Crossvit: Cross-attention multi-scale vision transformer for image classification. Proceedings of the IEEE\/CVF International Conference on Computer Vision, IEEE. Available online: http:\/\/openaccess.thecvf.com\/content\/ICCV2021\/html\/Chen_CrossViT_Cross-Attention_Multi-Scale_Vision_Transformer_for_Image_Classification_ICCV_2021_paper.html.","DOI":"10.1109\/ICCV48922.2021.00041"},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"2252","DOI":"10.52202\/075280-0106","article-title":"Patch n\u2019 pack: NaViT, a vision transformer for any aspect ratio and resolution","volume":"36","author":"Dehghani","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."}],"container-title":["International Journal of Molecular Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1422-0067\/27\/19\/8929\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,10]],"date-time":"2026-10-10T01:47:12Z","timestamp":1791596832000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1422-0067\/27\/19\/8929"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,8]]},"references-count":41,"journal-issue":{"issue":"19","published-online":{"date-parts":[[2026,10]]}},"alternative-id":["ijms27198929"],"URL":"https:\/\/doi.org\/10.3390\/ijms27198929","relation":{},"ISSN":["1422-0067"],"issn-type":[{"value":"1422-0067","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,8]]}}}