{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T20:26:07Z","timestamp":1787343967934,"version":"build-2736575974"},"reference-count":43,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Online J Public Health Inform"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Skin neglected tropical diseases (NTDs) are the most prevalent diseases worldwide, affecting people living in resource-limited areas with low health care services and trained professionals. While machine learning (ML)\u2013based diagnostic tools can be used for initial clinical assessment and patient screening, especially in resource-limited areas (including in Ethiopia), little effort has been made in this area.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This pilot study develops a foundational ML model for the diagnosis of skin NTDs using patient metadata to analyze the feasibility of ML-based models for skin NTDs by identifying and experimentally evaluating 8 ML models.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n For this study, we acquired a tabular skin NTD diagnostic dataset collected from a specific affected district in the southwest of Ethiopia. We used the data in 3 different structures, which include using the initial dataset (IDS) that contains huge null values, using a final dataset (FDS) created through preliminary preprocessing, and a third dataset created by applying feature engineering (FEFDS). Selecting 8 ML models, we trained the models in 4 major experimental settings: baseline training, handling structural missing values, handling severe class imbalance through conditional class weighting, and a hybrid approach based on robust dual cross-validation (CV) consisting of an outer repeated stratified\n k<\/jats:italic>\n -fold and nested CV methods. We used the macro and class-specific metrics (such as precision, recall, and\n F<\/jats:italic>\n 1<\/jats:sub>\n -score), including balanced accuracy, due to the severe class imbalance. Feature importance scores are also used for evaluating overall model performance.\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n After the final training applying the hybrid approach, 4 models scored a perfect test score (1.0) across all the metrics and all experiments except na\u00efve Bayes and multilayer perceptron (MLP), similarly scoring 0.997 balanced accuracy, 0.97 recall, and 0.985\n F<\/jats:italic>\n 1<\/jats:sub>\n -score. In the same experiment, the nested CV loop revealed a slight performance drop for light gradient boosting machine (LightGBM) and extreme gradient boosting (XGBoost), though both models similarly maintained higher mean scores of balanced accuracy and macro recall of 0.993 (SD 0.013), including the mean macro\n F<\/jats:italic>\n 1<\/jats:sub>\n -score of 0.996 (SD 0.007), scoring declining performance with a mean score of 0.993 (SD 0.013) in balanced accuracy, macro recall, and\n F<\/jats:italic>\n 1<\/jats:sub>\n -score, showing predictive biases. In terms of feature utilization, only CatBoost showed optimal feature ranking, while 4 models showed over-feature utilization (selection bias), with 3 models using small subsets of features (showing feature parsimony).\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n Overall, this study has been highly challenged by data scarcity, class imbalances, limited disease representation, specific geographic representation, and lack of more data modalities. Hence, further studies are suggested to confirm the results on larger datasets having a representative distribution of disease classes and geographic locations.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/84966","type":"journal-article","created":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T19:54:16Z","timestamp":1787342056000},"page":"e84966-e84966","source":"Crossref","is-referenced-by-count":0,"title":["Automating Diagnosis of Skin Neglected Tropical Diseases via Patient Metadata through Machine Learning Model with Adaptive Balancing and Dual Cross-Validation: Retrospective Diagnostic Accuracy Study"],"prefix":"10.2196","volume":"18","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-0163-6900","authenticated-orcid":false,"given":"Yohannes","family":"Minyilu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0622-4841","authenticated-orcid":false,"given":"Mohammed Abebe","family":"Yimer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1823-0301","authenticated-orcid":false,"given":"Million","family":"Meshesha","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,8,21]]},"reference":[{"key":"R1","unstructured":"Elimination of neglected tropical diseases (NTDs) in Ethiopia \u2013 WOREDA level coordination toolkit for the WASH and NTD sectors. Ethiopian Federal Ministry of Health; 2019. URL: https:\/\/www.susana.org\/knowledge-hub\/resources?id=3709 [Accessed 01-08-2026]"},{"key":"R2","unstructured":"Neglected tropical diseases. World Health Organization. Dec 14, 2025. URL: https:\/\/www.who.int\/news-room\/questions-and-answers\/item\/neglected-tropical-diseases [Accessed 01-08-2026]"},{"key":"R3","unstructured":"Ending the neglect to attain the sustainable development goals: a road map for neglected tropical diseases 2021\u20132030. World Health Organization; Jan 28, 2020. URL: https:\/\/www.who.int\/publications\/i\/item\/9789240010352 [Accessed 01-08-2026]"},{"issue":"6.1","key":"R4","doi-asserted-by":"publisher","first-page":"16S","DOI":"10.3855\/jidc.11707","article-title":"Looking for NTDs in the skin; an entry door for offering patient centered holistic care","volume":"14","author":"Abdela","journal-title":"J Infect Dev Ctries"},{"issue":"1","key":"R5","doi-asserted-by":"publisher","first-page":"240","DOI":"10.1186\/1756-3305-5-240","article-title":"The burden of neglected tropical diseases in Ethiopia, and opportunities for integrated control and elimination","volume":"5","author":"Deribe","journal-title":"Parasit Vectors"},{"issue":"12","key":"R6","doi-asserted-by":"publisher","DOI":"10.2196\/22478","article-title":"Mobile health strategies to tackle skin neglected tropical diseases with recommendations from innovative experiences: systematic review","volume":"8","author":"Carrion","journal-title":"JMIR Mhealth Uhealth"},{"key":"R7","unstructured":"The third national neglected tropical diseases strategic plan 2021-2025. Ethiopian Federal Ministry of Health; Nov 2021:1-116. URL: https:\/\/espen.afro.who.int\/sites\/default\/files\/content\/document\/Third%20NTD%20national%20Strategic%20Plan%202021-2025_0.pdf [Accessed 01-08-2026]"},{"issue":"11","key":"R8","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pntd.0012661","article-title":"A pathway for skin NTD diagnostic development","volume":"18","author":"Marks","journal-title":"PLoS Negl Trop Dis"},{"key":"R9","doi-asserted-by":"publisher","DOI":"10.1186\/s12879-025-12324-0","article-title":"Decentralizing care for cutaneous leishmaniasis and other skin diseases to primary health facilities in Southern Ethiopia: what are the needs?","volume":"26","author":"Tadesse","journal-title":"BMC Infect Dis"},{"key":"R10","doi-asserted-by":"publisher","DOI":"10.2196\/46295","article-title":"An mHealth app (eSkinHealth) for detecting and managing skin diseases in resource-limited settings: mixed methods pilot study","volume":"6","author":"Yotsu","journal-title":"JMIR Dermatol"},{"key":"R11","doi-asserted-by":"publisher","DOI":"10.1016\/j.lana.2022.100192","article-title":"Reimagining leprosy elimination with AI analysis of a combination of skin lesion images with demographic and clinical data","volume":"9","author":"Barbieri","journal-title":"Lancet Reg Health Am"},{"key":"R12","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2024.12.039","article-title":"A GAN-enhanced multimodal diagnostic framework utilizing an ensemble of BiLSTM, BiGRU, and RNN models for malaria and dengue detection","volume":"252","author":"Achary","journal-title":"Procedia Comput Sci"},{"issue":"4","key":"R13","doi-asserted-by":"publisher","DOI":"10.3390\/tropicalmed3040120","article-title":"Integrated management of skin NTDs\u2014lessons learned from existing practice and field research","volume":"3","author":"Yotsu","journal-title":"Trop Med Infect Dis"},{"issue":"1","key":"R14","doi-asserted-by":"publisher","DOI":"10.1177\/00368504211029777","article-title":"Machine learning on small size samples: a synthetic knowledge synthesis","volume":"105","author":"Kokol","journal-title":"Sci Prog"},{"issue":"1","key":"R15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-023-01000-z","article-title":"Small data machine learning in materials science","volume":"9","author":"Xu","journal-title":"npj Comput Mater"},{"key":"R16","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2021.11.011","article-title":"Tabular data: Deep learning is not all you need","volume":"81","author":"Shwartz-Ziv","journal-title":"Inf Fusion"},{"key":"R17","unstructured":"Florek P Zagda\u0144ski A . Benchmarking state-of-the-art gradient boosting algorithms for classification. arXiv. Preprint posted online on May 26, 2023. [doi: 10.48550\/arXiv.2305.17094]"},{"key":"R18","doi-asserted-by":"crossref","unstructured":"Wen Z Li Q He B Cui B . Challenges and opportunities of building fast GBDT systems. In: Proceedings of the 30th International Joint Conference on Artificial Intelligence. IJCAI; 2021:4661-4668. [doi: 10.24963\/ijcai.2021\/632]","DOI":"10.24963\/ijcai.2021\/632"},{"key":"R19","doi-asserted-by":"crossref","unstructured":"Wang X Jia W . Optimizing edge AI: a comprehensive survey on data, model, and system strategies. arXiv. Preprint posted online on Jan 4, 2025. [doi: 10.48550\/arXiv.2501.03265]","DOI":"10.32388\/IZOHCH"},{"issue":"2","key":"R20","doi-asserted-by":"publisher","first-page":"152","DOI":"10.3390\/info16020152","article-title":"Early detection of skin diseases across diverse skin tones using hybrid machine learning and deep learning models","volume":"16","author":"Aquil","journal-title":"Information"},{"issue":"1","key":"R21","doi-asserted-by":"publisher","DOI":"10.1016\/j.xjidi.2022.100150","article-title":"Deep learning in dermatology: a systematic review of current approaches, outcomes, and limitations","volume":"3","author":"Jeong","journal-title":"JID Innov"},{"key":"R22","doi-asserted-by":"publisher","DOI":"10.2196\/60231","article-title":"Sample size requirements for popular classification algorithms in tabular clinical data: empirical study","volume":"26","author":"Silvey","journal-title":"J Med Internet Res"},{"key":"R23","doi-asserted-by":"publisher","DOI":"10.1007\/s11334-022-00465-3","article-title":"Product pricing solutions using hybrid machine learning algorithm","author":"Namburu","journal-title":"Innov Syst Softw Eng"},{"key":"R24","doi-asserted-by":"crossref","unstructured":"Pattnayak P Mohanty A Das T Patnaik S . Applying artificial intelligence and deep learning to identify neglected tropical skin disorders. In: 2024 3rd International Conference for Innovation in Technology (INOCON). IEEE; 2024. [doi: 10.1109\/INOCON60754.2024.10511323]","DOI":"10.1109\/INOCON60754.2024.10511323"},{"key":"R25","doi-asserted-by":"publisher","DOI":"10.1016\/j.imu.2022.101078","article-title":"Optimized real-time diagnosis of neglected tropical diseases by automatic recognition of skin lesions","volume":"33","author":"Steyve","journal-title":"Inform Med Unlocked"},{"key":"R26","unstructured":"Recognizing neglected tropical diseases through changes on the skin: a training guide for front-line health workers. World Health Organization; Jun 1, 2018. URL: https:\/\/www.who.int\/publications\/i\/item\/9789241513531 [Accessed 01-08-2026]"},{"key":"R27","doi-asserted-by":"publisher","DOI":"10.1002\/aidi.202500089","article-title":"The challenge of handling structured missingness in integrated data sources","author":"Jackson","journal-title":"Adv Intell Discov"},{"issue":"4","key":"R28","doi-asserted-by":"publisher","first-page":"372","DOI":"10.1177\/1094428114548590","article-title":"Missing data: five practical guidelines","volume":"17","author":"Newman","journal-title":"Organ Res Methods"},{"key":"R29","doi-asserted-by":"publisher","DOI":"10.1016\/j.asoc.2022.109924","article-title":"The choice of scaling technique matters for classification performance","volume":"133","author":"de Amorim","journal-title":"Appl Soft Comput"},{"key":"R30","unstructured":"Metadata routing (\u201csample_weight\u201d). scikit-learn. URL: https:\/\/scikit-learn.org\/stable\/metadata_routing.html [Accessed 01-08-2026]"},{"issue":"8","key":"R31","doi-asserted-by":"publisher","DOI":"10.1136\/bmjopen-2013-002847","article-title":"Comparison of imputation methods for missing laboratory data in medicine","volume":"3","author":"Waljee","journal-title":"BMJ Open"},{"key":"R32","doi-asserted-by":"crossref","unstructured":"Zhuang N Howells J . A computational pipeline for stratifying autoimmune patients using binary antibody data. bioRxiv. Preprint posted online on Dec 19, 2026. [doi: 10.1101\/2025.09.11.670596]","DOI":"10.1101\/2025.09.11.670596"},{"issue":"5","key":"R33","doi-asserted-by":"publisher","first-page":"127","DOI":"10.11648\/j.ajtas.20241305.13","article-title":"Comparative analysis of cross-validation techniques: LOOCV, k-folds cross-validation, and repeated k-folds cross-validation in machine learning models","volume":"13","author":"Lumumba","journal-title":"Am J Theor Appl Stat"},{"issue":"10","key":"R34","doi-asserted-by":"publisher","first-page":"273","DOI":"10.1007\/s10462-024-10884-2","article-title":"Handling imbalanced medical datasets: review of a decade of research","volume":"57","author":"Salmi","journal-title":"Artif Intell Rev"},{"key":"R35","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2024.04.082","article-title":"Exploratory analysis of methods, techniques, and metrics to handle class imbalance problem","volume":"235","author":"Hr","journal-title":"Procedia Comput Sci"},{"issue":"6","key":"R36","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-024-10759-6","article-title":"A survey on imbalanced learning: latest research, applications and future directions","volume":"57","author":"Chen","journal-title":"Artif Intell Rev"},{"issue":"2","key":"R37","doi-asserted-by":"publisher","first-page":"92","DOI":"10.3390\/info14020092","article-title":"Effective handling of missing values in datasets for classification using machine learning methods","volume":"14","author":"Palanivinayagam","journal-title":"Information"},{"key":"R38","doi-asserted-by":"crossref","unstructured":"Grinsztajn L Oyallon E Varoquaux G . Why do tree-based models still outperform deep learning on typical tabular data? In: Proceedings of the 36th International Conference on Neural Information Processing Systems. Curran Associates; 2022. [doi: 10.52202\/068431-0037]","DOI":"10.52202\/068431-0037"},{"key":"R39","doi-asserted-by":"publisher","DOI":"10.7717\/peerj-cs.55","article-title":"Probabilistic programming in Python using PyMC3","volume":"2","author":"Salvatier","journal-title":"PeerJ Comput Sci"},{"issue":"1","key":"R40","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-023-32396-9","article-title":"Recursive computed ABC (cABC) analysis as a precise method for reducing machine learning based feature sets to their minimum informative size","volume":"13","author":"L\u00f6tsch","journal-title":"Sci Rep"},{"issue":"11","key":"R41","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0224365","article-title":"Machine learning algorithm validation with a limited sample size","volume":"14","author":"Vabalas","journal-title":"PLOS ONE"},{"issue":"10","key":"R42","doi-asserted-by":"publisher","first-page":"3093","DOI":"10.1093\/bioinformatics\/btaa046","article-title":"Consensus features nested cross-validation","volume":"36","author":"Parvandeh","journal-title":"Bioinformatics"},{"issue":"4","key":"R43","doi-asserted-by":"publisher","first-page":"395","DOI":"10.3934\/fods.2024021","article-title":"On marginal feature attributions of tree-based models","volume":"6","author":"Filom","journal-title":"Found Data Sci"}],"container-title":["Online Journal of Public Health Informatics"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T19:54:18Z","timestamp":1787342058000},"score":1,"resource":{"primary":{"URL":"https:\/\/ojphi.jmir.org\/2026\/1\/e84966"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,21]]},"references-count":43,"URL":"https:\/\/doi.org\/10.2196\/84966","relation":{},"ISSN":["1947-2579"],"issn-type":[{"value":"1947-2579","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,21]]},"article-number":"v18i6.1e84966"}}