Mallesh Nanditha, Zhao Max, Meintker Lisa, Höllein Alexander, Elsner Franz, Lüling Hannes, Haferlach Torsten, Kern Wolfgang, Westermann Jörg, Brossart Peter, Krause Stefan W, Krawitz Peter M
Institute for Genomic Statistics and Bioinformatics, University Bonn, Bonn, Germany.
Institute of Human Genetics and Medical Genetics, Charité University Hospital, Berlin, Germany.
Patterns (N Y). 2021 Sep 17;2(10):100351. doi: 10.1016/j.patter.2021.100351. eCollection 2021 Oct 8.
Multi-parameter flow cytometry (MFC) is a cornerstone in clinical decision making for leukemia and lymphoma. MFC data analysis requires manual gating of cell populations, which is time-consuming, subjective, and often limited to a two-dimensional space. In recent years, deep learning models have been successfully used to analyze data in high-dimensional space and are highly accurate. However, AI models used for disease classification with MFC data are limited to the panel they were trained on. Thus, a key challenge in deploying AI into routine diagnostics is the robustness and adaptability of such models. This study demonstrates how transfer learning can be applied to boost the performance of models with smaller datasets acquired with different MFC panels. We trained models for four additional datasets by transferring the features learned from our base model. Our workflow increased the model's overall performance and, more prominently, improved the learning rate for small training sizes.
多参数流式细胞术（MFC）是白血病和淋巴瘤临床决策的基石。MFC数据分析需要对细胞群体进行手动设门，这既耗时又主观，而且通常局限于二维空间。近年来，深度学习模型已成功用于分析高维空间中的数据，且准确性很高。然而，用于基于MFC数据进行疾病分类的人工智能模型仅限于其训练所用的检测板。因此，将人工智能应用于常规诊断的一个关键挑战是此类模型的稳健性和适应性。本研究展示了如何应用迁移学习来提高使用不同MFC检测板获取的较小数据集的模型性能。我们通过转移从基础模型学到的特征，为另外四个数据集训练了模型。我们的工作流程提高了模型的整体性能，更显著的是，提高了小训练规模的学习率。
