High-confidence pseudo-labels for domain adaptation in COVID-19 detection (preprint)

This paper outlines our submission for the 4th COV19D competition as part of the `Domain adaptation, Explainability, Fairness in AI for Medical Image Analysis' (DEF-AI-MIA) workshop at the Computer Vision and Pattern Recognition Conference (CVPR). The competition consists of two challenges. The first is to train a classifier to detect the presence of COVID-19 from over one thousand CT scans from the COV19-CT-DB database. The second challenge is to perform domain adaptation by taking the dataset from Challenge 1 and adding a small number of scans (some annotated and other not) for a different distribution. We preprocessed the CT scans to segment the lungs, and output volumes with the lungs individually and together. We then trained 3D ResNet and Swin Transformer models on these inputs. We annotated the unlabeled CT scans using an ensemble of these models and chose the high-confidence predictions as pseudo-labels for fine-tuning. This resulted in a best cross-validation mean F1 score of 93.39\% for Challenge 1 and a mean F1 score of 92.15 for Challenge 2.

Phytest: Quality Control for Phylogenetic Analyses (preprint)

Motivation: The ability to automatically conduct quality control checks on phylogenetic analyses is becoming more important with the increase of genetic sequencing and use of real-time pipelines e.g. in the SARS-CoV-2 era. Implementations of ”nowcasting” or real-time phylogenetic analyses require automated testing to make sure that problems in the data are caught automatically within analysis pipelines and in a timely manner. Here we present Phytest (version 1.0) a tool for automating quality control checks on sequence, trees and metadata during phylogenetic analyses. Results: Phytest is a phylogenetic analysis testing program that easily integrates into existing phylogenetic pipelines. We demonstrate the utility of Phytest with real-world examples. Availability: Phytest source code available on GitHub (https://github.com/phytest-devs/phytest) and can be installed via PyPI with the command ‘pip install phytest‘. Extensive documentation can be found at https://phytest-devs.github.io/phytest/. Contact: wytamma.wirth@unimelb.edu.au