UMAP Based Anomaly Detection for Minimal Residual Disease Quantification within Acute Myeloid Leukemia

Abstract

Leukemia is the most frequent malignancy in children and adolescents, with acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) as the most common subtypes. Minimal residual disease (MRD) measured by flow cytometry (FCM) has proven to be a strong prognostic factor in ALL as well as in AML. Machine learning techniques have been emerging in the field of automated MRD quantification with the objective of superseding subjective and time-consuming manual analysis of FCM-MRD data. In contrast to ALL, where supervised multi-class classification methods have been successfully deployed for MRD detection, AML poses new challenges: AML is rarer (with fewer available training data) than ALL and much more heterogeneous in its immunophenotypic appearance, where one-class classification (anomaly detection) methods seem more suitable. In this work, a new semi-supervised approach based on the UMAP algorithm for MRD detection utilizing only labels of blast free FCM samples is presented. The method is tested on a newly gathered set of AML FCM samples and results are compared to state-of-the-art methods. We reach a median F1-score of 0.794, while providing a transparent classification pipeline with explainable results that facilitates inter-disciplinary work between medical and technical experts. This work shows that despite several issues yet to overcome, the merits of automated MRD quantification can be fully exploited also in AML.

Identifying Novel Subtypes of Functional Gastrointestinal Disorder by Analyzing Nonlinear Structure in Integrative Biopsychosocial Questionnaire Data

“…UMAP is a recently proposed nonlinear graph-based dimensionality reduction method, which is known for improving clustering accuracy by seeking to accurately represent local structure and to better incorporate global structure [39]. There have been many successful cases of UMAP applications in medical fields [46][47][48]. Meanwhile, there have been concerns regarding the reliability of the results due to the unsupervised nature of the procedure.…”

Fraud Detection in Credit Card Transactions Using HDBSCAN, UMAP and SMOTE Methods

“…Various under-sampling methods have been proposed, and the most commonly used are random under-sampling and Tomek link (T-link). The Synthetic Minority Oversampling Technique (SMOTE) is an improved oversampling technique developed by [15]. SMOTE is based on k's nearest neighbor to produce new synthetic sampling in feature spaces based on a certain percentage for minority classes.…”

Preterm infants show an atypical processing of the mother’s voice

“…To extract the topographical organization of oscillatory responses to maternal speech in the FW or BW conditions in preterm and full-term infants, we employed a data-driven topographical clustering method described in the literature as effective for high-dimensional data (Allaoui, Kherfi, & Cheriet, 2020; Weijler et al, 2022; Yang et al, 2021). First, dimension reduction was performed using Uniform Manifold Approximation and Projection through the dedicated python package (UMAP) (McInnes, Healy, & Melville, 2018) with electrodes as cases and, as columns, the mean power of the electrode for each frequency band in 100ms timebins starting at voice onset (20 neighbors).…”