The objective of this prospective observational study is to rigorously examine the feasibility and efficacy of utilizing latent diffusion models for data augmentation in anti-nuclear antibody (ANA) Hep-2 cell immunofluorescence images. The main question it aims to answer is: Can the application of such models potentially enhance the data quality, increase sample diversity, or improve the accuracy and efficiency of subsequent analytical processes (like disease diagnosis and classification) when utilized with ANA-related images?
A fundamental problem in biomedical research is the low number of observations available, mostly due to a lack of available biosamples, prohibitive costs, or ethical reasons. Augmenting few real observations with generated in silico samples could lead to more robust analysis results and a higher reproducibility rate. Here, The investigators propose to use unsupervised learning with latent diffusion models for the realistic generation of ANA-IIF image data. The investigators hypothesize that the the generation of ANA-IIF image will be realistic if it is hard to differentiate them (fake) from real (true) . To test this hypothesis, the investigators present a Multi-center Visual Turing tests (https://turing.rednoble.net/) in order to evaluate the quality of the generated (fake) images. This experimental setup allows the investigators to validate the overall quality of the generated ANA-IIF images, which can then be used to (1) train cytopathologists for educational purposes, and (2) generate realistic samples to train deep networks with big data.
Study Type
OBSERVATIONAL
Enrollment
300
determining the ANA pattern type with or without referring to the results of AI model output.
The realistic of images synthesized by diffusion models
The investigators conducted a study using the visual Turing test method, measuring through a questionnaire format, and assessed the measurement results using a 5-point Likert Scale. The 5-point Likert Scale assesses participants' opinions on the quality of images through five response options: Real, Much like, Uncertain, Not quite like, Fake. It calculates scores by assigning numbers (e.g., 5 to 1) to these options, summing up scores for each participant. Results are evaluated by analyzing the distribution of scores, including mean scores, and assessing their reliability and validity. Additionally, the investigator calculated a range of parameters utilized for internal model assessment, including: including precision, recall, F1 score, and mean average precision (mAP).
Time frame: Baseline
The impact of the AI model's output on the participants
The investigator evaluated the change in the accuracy rate of participants' interpretations before and after being assisted by AI model, investigators will conduct a comparative analysis. Additionally, the investigator calculate the Kappa coefficient of agreement between human interpretations and the model, and evaluate whether there are differences in accuracy among cytopathologists with varying levels of experience when assisted by AI.
Time frame: Baseline
The time taken of ANA pattern interpretation
The investigator compare the time taken of participant to complete interpretations before and after the AI model's intervention, assessing whether there is a reduction in average interpretation time per case, from X minutes pre-AI assistance to Y minutes post-AI.
Time frame: Baseline
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.