Servas Adolph
About me
I am a PhD researcher in Future Convergence Technology / Big Data Engineering at Soonchunhyang University, South Korea. My work focuses on AI for healthcare, with an emphasis on building intelligent systems that can support clinical decision-making and improve access to reliable medical insight.
My core research area is domain adaptation. I develop AI models that remain reliable when transferred across hospitals, imaging devices, clinical environments, and patient populations where data distributions may differ. I also design multimodal systems that combine medical images, clinical notes, and other patient data to support a more complete understanding of health conditions.
Beyond research, I build applied AI systems using large language models. I use Retrieval-Augmented Generation (RAG) to make AI responses more accurate, traceable, and grounded in trusted information rather than unsupported generation. I have also applied this thinking beyond medicine through Matokeo Yangu, a bilingual platform that helps Tanzanian students check exam results and receive AI-guided advice on university and career pathways.
I am from Tanzania, and that background shapes the problems I care about. My goal is to build AI health systems that are not only accurate, but also practical, trustworthy, and useful in real-world settings, especially in places like Tanzania and other parts of Africa where access to specialists can be limited.
I hold an MSc in Big Data Engineering (2023) from Soonchunhyang University and a BSc in Computer Engineering & Information Technology from the United African University of Tanzania. Full education and research experience details are on the Education & Experience page.
Research Focus
Domain Adaptation in Medical AI
I build models that keep working well when moved from one hospital to another, even when the images, machines, or patient data look different. This helps AI tools stay accurate in real hospitals, not just in one lab's dataset.
Multimodal Report Generation for Blood Smear Analysis
I build AI systems that read microscopy images from blood smear tests and generate clinical reports automatically, combining image data with attention mechanisms and language models. This helps reduce manual reporting work and supports faster diagnosis.
AI Systems with LLMs and RAG
Beyond research, I build practical AI applications using Retrieval-Augmented Generation (RAG), so answers stay grounded in real data. I used this approach in Matokeo Yangu, a platform that helps Tanzanian students check exam results and receive AI-guided academic advice.
Blood Cell Classification & Domain Adaptation
I work on classifying white blood cells and other blood cells across hospitals with different equipment. This includes a Weight Module that prevents the model from learning misleading patterns when source and target hospitals do not fully match, and reinforcement learning-based style transfer to bridge visual differences between hospitals.
Recent News
Completed the FLOuRISH Academic Entrepreneurship Global Program 2026 in Japan at Tokyo University of Agriculture and Technology and received the Interdisciplinary Innovation Award.
Our paper "Pancreas Segmentation Using a Two-Stage Pipeline of Faster R-CNN and TransUNet" was accepted by Applied Sciences.
Our paper "WBC YOLO-ViT: 2-Way 2-Stage White Blood Cell Detection and Classification with a Combination of YOLOv5 and Vision Transformer" was accepted by Computers in Biology and Medicine.
Our paper "Diffusion-based Wasserstein Generative Adversarial Network for Blood Cell Image Augmentation" was accepted by Engineering Applications of Artificial Intelligence.
Our paper "Adapting YOLO-ViT for Differential Diagnosis of Myelodysplastic Syndromes and Normal Blood Cell" was presented at the Proceedings of the Korea Society of Computer and Information Conference.
Started my PhD in Future Convergence Technology / Big Data Engineering at Soonchunhyang University, South Korea.
Presented "White Blood Cell Detection and Classification using YOLOv5 with Hybrid ResNet50-VGG16-SVM" at the 6th International Conference on ICT for Smart Health & Home (ICT4sHealth & Home), Kota Kinabalu, Malaysia.