Aim:

To evaluate the application effect of the teaching platform iLN in the teaching of lupus nephritis, aiming to provide practical experience and evidence-based support for the implementation of generative AI technology in professional medical education.

Method:

Based on the DeepSeek large language model, the investigators developed "iLN" - an interactive teaching platform for lupus nephritis. This platform integrates the latest authoritative textbooks, clinical guidelines, pathological atlases, and real clinical cases, and includes seven core modules, such as "case interaction", "knowledge teaching", and "AI question answering". The investigators conducted a randomized controlled trial (RCT) on 50 fourth-year undergraduate students from Nanjing University School of Medicine to evaluate the effectiveness of the iLN platform compared to traditional teaching methods. The teaching effect was evaluated through objective test scores and questionnaires about student satisfaction and platform usability.

Artificial Intelligence (AI) technology is reshaping the landscape of medical education with unprecedented depth and breadth. From early rule-based teaching expert systems to the current generation of Generative AI (GAI) systems like ChatGPT and DeepSeek, the continuous evolution of technology has not only greatly enriched teaching methods but also fundamentally changed the ways knowledge is acquired, disseminated, and evaluated. Internationally, Generative AI (GAI) has been widely applied in creating virtual cases, constructing clinical simulation scenarios, assisting medical writing, and providing personalized learning plans for medical students. However, there are almost no dedicated large language model (GAI) teaching applications specifically for lupus nephritis (LN) worldwide. As the most common severe organ complication of systemic lupus erythematosus (SLE), LN affects approximately 50% to 70% of SLE patients with kidney involvement. Its diagnosis and treatment involve multiple complex processes, including the identification of clinical manifestations, the interpretation of immune markers, the assessment of renal pathological types, the quantification of disease activity, and the formulation of personalized treatment plans. All of these place extremely high demands on the knowledge integration and clinical reasoning skills of medical students. The traditional LN teaching model mainly relies on theoretical classroom lectures and is supplemented by static pathological image presentations. Some scholars have introduced problem-based learning (PBL) teaching models and integrated ideological and political education, but there are still significant deficiencies: (1) Knowledge update lag - In the field of LN, new diagnostic standards (such as the 2019 EULAR/ACR systemic lupus erythematosus classification criteria), new pathological classifications (such as the 2018 ISN/RPS revised version), and new drug treatment regimens (such as belimumab) are constantly emerging, while the textbook revision cycle is relatively long, making it difficult for students to obtain the latest knowledge in a timely manner; (2) Lack of personalized teaching - Facing students with varying levels of basic knowledge, teachers find it difficult to meet the needs of different learning paces; students with weak foundations have difficulty understanding complex pathological mechanisms, while those with solid foundations may find the content too simplistic; (3) Insufficient clinical reasoning training - Traditional classrooms lack real clinical scenario simulations, and students listen passively rather than actively exploring, making it difficult for them to establish a complete clinical reasoning chain from "symptoms - signs - laboratory tests - pathology - diagnosis - treatment". In this context, building a GAI-assisted teaching platform specifically designed for LN teaching, fully leveraging the personalized interaction, dynamic content generation, and intelligent assessment capabilities of AI, is of great significance for teaching reform and innovation.

In this study, the investigators constructed an iLN interactive teaching platform for lupus nephritis based on the DeepSeek large language model and by integrating the latest authoritative textbooks, guidelines, and clinical pathology materials. Through a randomized controlled teaching trial, the investigators systematically evaluated the application effect of this platform in LN teaching, aiming to provide practical experience and evidence-based support for the implementation of GAI technology in specialized medical education.

Inclusion Criteria:

• As a student majoring in clinical medicine
• Volunteering to participate in the study

Exclusion Criteria:

• Inability to understand, read, or communicate in Chinese
• Failure to participate in any stage of the study (e.g. in-class education session, or post-test)
• Requesting withdrawal from the study

Individual participant data will not be shared