NONE 20260224114002804 wseas/wseas content-registration-form+ja@crossref.org content-registration-form WSEAS TRANSACTIONS ON BIOLOGY AND BIOMEDICINE 1109-9518 2224-2902 Divin Kayeye Kabeya Faculté Polytechnique, Université de Kinshasa (UNIKIN), Kinshasa, DEMOCRATIC REPUBLIC OF THE CONGO Witesyavwirwa Vianney Kambale Faculty of Information and Communication Technology, Tshwane University of Technology, Pretoria, SOUTH AFRICA Kutabuna Kubakisa Christian Faculté Polytechnique, Université de Kinshasa (UNIKIN), Kinshasa, DEMOCRATIC REPUBLIC OF THE CONGO Isaac Lukusa Kayembe Faculté Polytechnique, Université de Kinshasa (UNIKIN), Kinshasa, DEMOCRATIC REPUBLIC OF THE CONGO Mahmoud Hamed Institute for Smart Systems Technologies, Universität Klagenfurt, AUSTRIA Kyandoghere Kyamakya Faculté Polytechnique, Université de Kinshasa (UNIKIN), Kinshasa, DEMOCRATIC REPUBLIC OF THE CONGO Poor digital infrastructure and a lack of local language resources result in Sub-Saharan Africa having limited access to specialized medical information. Swahili, spoken by over 100 million people, is crucial for the democratization of digital healthcare. The lack of an extensive Swahili medical corpus complicates the development of credible and culturally tailored language models. Large Language Models (LLMs) promise to be a boon for medical chatbots, providing accessible, context-aware, and language-appropriate health information. However, fine-tuning techniques have constraints, including diminished factual robustness and a larger risk of medical hallucinations. The Swahili-speaking medical chatbot Lueji, initially based on a fine-tuned model, is extended with a Retrieval-Augmented Generation (RAG) architecture to address these limitations. A FAISS-based semantic retriever utilizing a Swahili-translated Huatuo-26M Chinese medical corpus and the UlizaLlama model, tailored for African languages, is employed in the proposed system. This hybrid methodology improves factual dependability and contextual accuracy while retaining generative fine-tuning-achieved verbal fluency. The BLEU, ROUGE, and GLEU metrics were used to quantify lexical consistency, structural similarity, and text quality. The fine-tuned model outperforms the RAG-based version on typical text similarity measures (BLEU-1 = 0.2217, ROUGE-1 = 0.3168, GLEU = 0.1077 vs. 0.1912, 0.2916, 0.0918). A qualitative investigation reveals that the RAG architecture significantly reduces hallucinations and enhances clinical factuality. Low-resource environments present a fundamental trade-off between linguistic fluency and factual accuracy. The study gives fresh empirical insights into the balance between fine-tuning and retrieval augmentation for low-resource medical LLMs. It lays the groundwork for reliable, hybrid, and culturally inclusive African medical chatbots. 02 24 2026 02 24 2026 102 10 https://creativecommons.org/licenses/by/4.0/deed.en_US 10.37394/23208.2026.23.10 https://wseas.com/journals/bab/2026/a205108-006(2026).pdf B. Petrucci et al., “The global otolaryngology-head and neck surgery workforce,” JAMA Otolaryngology–Head & Neck Surgery, vol. 149, no. 10, pp. 904–911, 2023. DOI: 10.1001/jamaoto.2023.2339. [Online]. Available: https://doi.org/10. 1001/jamaoto.2023.2339. M. M. Willie, “Examining the scarcity of oncology healthcare providers in cancer management: A case study of the eastern cape province, south africa,” Open Health, vol. 6, no. 1, p. 20 250 058, 2025, (Accessed: May 02, 2025). DOI: 10 . 1515 / ohe - 2025 - 0058. [Online]. Available: https : //doi.org/10.1515/ohe-2025-0058. V. Vanderpuye et al., “Cancer care workforce in africa: Perspectives from a global survey,” Infectious Agents and Cancer, vol. 14, no. 1, p. 11, 2019. DOI: 10 . 1186 / s13027 - 019 - 0227-8. [Online]. Available: https://doi. org/10.1186/s13027-019-0227-8. B. S. Sylla and C. P. Wild, “A million africans a year dying from cancer by 2030: What can cancer research and control offer to the continent?” International Journal of Cancer, vol. 130, no. 2, pp. 245–250, 2012. DOI: 10 . 1002 / ijc . 26333. [Online]. Available: https://doi.org/10.1002/ijc.26333. Kumar and S. Joshi, “Applications of AI in healthcare sector for enhancement of medical decision making and quality of service,” in Proceedings of the 2022 International Conference on Decision Aid Sciences and Applications (DASA), Chiang Mai, Thailand: IEEE, 2022, pp. 37–41. DOI: 10.1109/DASA54658.2022.9765041. [Online]. Available: https://doi.org/10.1109/DASA54658.2022.976 5041. J. C. L. Chow and K. Li, “Large language models in medical chatbots: Opportunities, challenges, and the need to address AI risks,” Information, vol. 16, no. 7, p. 549, 2025. DOI: 10 . 3390 / info16070549. [Online]. Available: https : / / doi . org / 10 . 3390 / info16070549. H. Al Shamsi, A. G. Almutairi, S. Al Mashrafi, and T. Al Kalbani, “Implications of language barriers for healthcare: A systematic review,” Oman Medical Journal, vol. 35, no. 2, e122, 2020. DOI: 10.5001/omj.2020.40. [Online]. Available: https : / / doi . org / 10 . 5001 / omj.2020.40. A. E. Babalola, V. Johnson, A. Oromakinde, et al., “The role of local languages in effective health service delivery,” Discover Public Health, vol. 22, no. 1, p. 59, 2025. DOI: 10. 1186 / s12982 - 025 - 00429 - 5. [Online]. Available: https : / / doi . org / 10 . 1186 / s12982-025-00429-5. D. K. Kabeya, W. V. Kambale, J.-G. M. Mboma, V. N. Bendo, S. K. Kasereka, and K. Kyamakya, “Designing a swahili-speaking medical chatbot for oncology, dermatology, and otorhinolaryngology care in low-resource settings,” in Proceedings of the 2025 Conference on Information Communications Technology and Society (ICTAS), Durban, South Africa, 2025, pp. 1–6. DOI: 10.1109/ ICTAS64866 . 2025 . 11155337. [Online]. Available: https : / / doi . org / 10 . 1109 / ICTAS64866.2025.11155337. K. Singhal et al., “Toward expert-level medical question answering with large language models,” Nature Medicine, vol. 31, no. 3, pp. 943–950, Mar. 2025. DOI: 10 . 1038 / s41591- 024- 03423- 7. [Online]. Available: https://doi.org/10.1038/s41591-024- 03423-7. Y. Li, Z. Li, K. Zhang, R. Dan, S. Jiang, and Y. Zhang, “ChatDoctor: A medical chat model fine-tuned on a large language model Meta-AI (LLaMA) using medical domain knowledge,” Cureus, vol. 15, no. 6, e40704, 2023. DOI: 10.48550/arXiv.2303.14070 [Online]. Available: https://doi.org/10.48550/arXiv.2303.14070. C. Wu, W. Lin, X. Zhang, Y. Zhang, W. Xie, and Y. Wang, “Pmc-llama: Toward building open-source language models for medicine,” Journal of the American Medical Informatics Association, vol. 31, no. 9, pp. 1833–1843, Apr. 2024, ISSN: 1527-974X. DOI: 10 . 1093 / jamia / ocae045. eprint: https : / / academic . oup . com / jamia / article - pdf/31/9/1833/58868261/ocae045.pdf. [Online]. Available: https://doi.org/10. 1093/jamia/ocae045. H. Xiong et al., DoctorGLM: Fine-tuning your chinese doctor is not a herculean task, arXiv preprint, arXiv:2304.01097 (Accessed: December 04, 2025), 2023. [Online]. Available: https://arxiv.org/ abs/2304.01097. Azam, Z. Naz, and M. U. G. Khan, “Cancerbot: A retrieval-augmented generation based cancer chatbot using large language models,” in Proceedings of the 2024 18th International Conference on Open Source Systems and Technologies (ICOSST), Lahore, Pakistan, 2024, pp. 1–6. DOI: 10 . 1109 / ICOSST64562 . 2024 . 10871155. [Online]. Available: https://doi.org/10.1109/ICOSST64562.2024.1 0871155. C. Li et al., “Development of a meta-question enhanced retrieval-augmented generation model and its application in dermatology,” in Proceedings of the 2024 17th International Conference on Advanced Computer Theory and Engineering (ICACTE), Xiamen, China: IEEE, 2024, pp. 281–285. DOI: 10 . 1109 / ICACTE62428 . 2024 . 10871274. [Online]. Available: https : / / doi . org / 10 . 1109 / ICACTE62428.2024.10871274. S. P. Bande, V. N, and P. K. K, “Augmenting medical diagnostics with AI: A dual approach using RAG-based chatbots and NanoGPT models,” in Proceedings of the 2025 6th International Conference on Recent Advances in Information Technology (RAIT), Dhanbad, India: IEEE, 2025, pp. 1–6. DOI: 10.1109/ RAIT65068 . 2025 . 11088917. [Online]. Available: https : / / doi . org / 10 . 1109 / RAIT65068.2025.11088917. H. Zhang et al., “HuatuoGPT, towards taming language models to be a doctor,” in Findings of the Association for Computational Linguistics: EMNLP 2023, Singapore: Association for Computational Linguistics, 2023, pp. 10 859–10 885. DOI: 10 . 18653 / v1/2023.findings- emnlp.725. [Online]. Available: https://doi.org/10.18653/ v1/2023.findings-emnlp.725. M. Alshammary, M. N. Uddin, and L. Khan, “RFPG: Question-answering from low-resource language (Arabic) texts using factually aware RAG,” in Proceedings of the 2024 IEEE 10th International Conference on Collaboration and Internet Computing (CIC), Cincinnati, OH, USA: IEEE, 2024, pp. 107–116. DOI: 10 . 1109 / CIC62241 . 2024.00023. [Online]. Available: https:// doi.org/10.1109/CIC62241.2024.00023. H. Hosseini, M. S. Zare, A. H. Mohammadi, A. Kazemi, Z. Zojaji, and M. A. Nematbakhsh, “PersianRAG: A retrieval-augmented generation system for the Persian language,” in Proceedings of the 2024 15th International Conference on Information and Knowledge Technology (IKT), Isfahan, Iran: IEEE, 2024, pp. 272–278. DOI: 10 . 1109 / IKT65497 . 2024.10892726. [Online]. Available: https: / / doi . org / 10 . 1109 / IKT65497 . 2024 . 10892726. S. M. M. R. J. Senanayaka, A. W. A. D. N. D. Abeysekara, and M. G. N. N. Premadasa, “SingRAG: A translation-augmented framework for code-mixed Singlish processing,” in Proceedings of the 2024 9th International Conference on Information Technology Research (ICITR), Colombo, Sri Lanka: IEEE, 2024, pp. 1–6. DOI: 10.1109/ICITR64794. 2024.10857714. [Online]. Available: https: //doi.org/10.1109/ICITR64794.2024. 10857714. J. Singh and R. Thakur, Quantum-RAG and PunGPT2: Advancing low-resource language generation and retrieval for the Punjabi language, arXiv preprint, arXiv:2508.01918 (Accessed: December 04, 2025), 2025. [Online]. Available: https://arxiv.org/ abs/2508.01918. B. Bogale, T. Tegegne, S. Teferra, et al., RAG based QA for low-resource languages, Research Square preprint, (Accessed: December 04, 2025), 2024. DOI: 10.21203/ rs.3.rs-5360450/v1. [Online]. Available: https://doi.org/10.21203/rs.3.rs5360450/v1. G. Martin, M. E. Mswahili, Y.-S. Jeong, and J. Woo, “SwahBERT: Language model of Swahili,” in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Seattle, United States: Association for Computational Linguistics, 2022, pp. 303–313. DOI: 10 . 18653 / v1 / 2022 . naacl - main . 23. [Online]. Available: https://doi.org/10. 18653/v1/2022.naacl-main.23. S. A. Sani, S. H. Muhammad, and D. Jarvis, Investigating the impact of language-adaptive fine-tuning on sentiment analysis in the Hausa language using AfriBERTa, arXiv preprint, arXiv:2501.11023 (Accessed: December 04, 2025), 2025. [Online]. Available: https : / / arxiv.org/abs/2501.11023. B. W. Wanjawa, L. D. A. Wanzare, F. Indede, O. Mconyango, L. Muchemi, and E. Ombui, “KenSwQuAD: A question answering dataset for swahili low-resource language,” ACM Transactions on Asian and Low-Resource Language Information Processing, vol. 22, no. 4, pp. 1–20, 2023. DOI: 10 . 1145 / 3578553. [Online]. Available: https://doi. org/10.1145/3578553. F. P. Dossou et al., “Afrolm: A self-active learning-based multilingual pretrained language model for 23 african languages,” in Proceedings of the 29th International Conference on Computational Linguistics, (Accessed: December 04, 2025), Gyeongju, Republic of Korea: International Committee on Computational Linguistics, 2022, pp. 5235– 5269. [Online]. Available: https://doi.org/10.18653/v1/2022.sustainlp1.11. Adelani et al., “MasakhaNER: Named entity recognition for african languages,” Transactions of the Association for Computational Linguistics, vol. 9, pp. 1116– 1131, 2021. DOI: 10.1162/tacl_a_00416. [Online]. Available: https://doi.org/10.1162/tacl_a_00416. O. Ogundepo et al., “AfriQA: Cross-lingual open-retrieval question answering for african languages,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada: Association for Computational Linguistics, 2023, pp. 14 445–14 465. DOI: 10.18653/v1/2023.findings-emnlp.997. [Online]. Available: https://doi.org/10.18653/v1/2023.findingsemnlp.997. J. Li et al., Huatuo-26m, a large-scale chinese medical qa dataset, 2023. arXiv: 2305.01526 [cs.CL]. [Online]. Available: https : / / arxiv.org/abs/2305.01526. S. O. Ayodele and S. K. Aremu, “The cost of setting up an ENT endoscopic practice in lower middle-income countries of sub-saharan africa,” Journal of the West African College of Surgeons, vol. 12, no. 2, pp. 104–108, 2022. DOI: 10.4103/jwas.jwas_57_22. [Online]. Available: https : / / doi . org / 10 . 4103 / jwas.jwas_57_22. K. Khoza-Shangase, “Occupational noise regulation and hearing conservation in african LMICs: A narrative policy and implementation review,” Environmental Disease, vol. 8, no. 3, pp. 69–77, 2023. DOI: 10.4103/ed.ed_12_25. [Online]. Available: https://doi.org/10.4103/ed.ed_12_25. S. K. Kiprono, J. W. Muchunu, and J. E. Masenga, “Skin diseases in pediatric patients attending a tertiary dermatology hospital in northern tanzania: A cross-sectional study,” BMC Dermatology, vol. 15, no. 1, p. 16, 2015. DOI: 10 . 1186 / s12895 - 015 - 0035 - 9. [Online]. Available: https://doi.org/10. 1186/s12895-015-0035-9. W. Ngwa et al., “Cancer in sub-saharan africa: A lancet oncology commission,” The Lancet Oncology, vol. 23, no. 6, e251–e312, 2022. DOI: 10.1016/S1470- 2045(21)00720- 8. [Online]. Available: https://doi.org/10. 1016/S1470-2045(21)00720-8. W. Li, L. Yu, M. Wu, J. Liu, M. Hao, and Y. Li, “Doctorgpt: A large language model with chinese medical question-answering capabilities,” in Proceedings of the 2023 International Conference on High Performance Big Data and Intelligent Systems (HDIS), Shenzhen, China, 2023, pp. 186–193. DOI: 10.1109/HDIS60872.2023.10499472. [Online]. Available: https://doi.org/10. 1109/HDIS60872.2023.10499472. J.-H. Jung, D. Kim, K.-B. Lee, and Y. Lee, “Performance evaluation of large language model chatbots for radiation therapy education,” Information, vol. 14, no. 7, p. 397, 2023. DOI: 10.3390/info16070521. [Online]. Available: https://doi.org/10.3390/info16070521.