Evaluating progress in automatic chest X-ray radiology report generation.

Artificial intelligence (AI) models for automatic generation of narrative radiology reports from images have the potential to enhance efficiency and reduce the workload of radiologists. However, evaluating the correctness of these reports requires metrics that can capture clinically pertinent differences. In this study, we investigate the alignment between automated metrics and radiologists' scoring of errors in report generation. We address the limitations of existing metrics by proposing new metrics, RadGraph F1 and RadCliQ, which demonstrate stronger correlation with radiologists' evaluations. In addition, we analyze the failure modes of the metrics to understand their limitations and provide guidance for metric selection and interpretation. This study establishes RadGraph F1 and RadCliQ as meaningful metrics for guiding future research in radiology report generation.

GaitForeMer: Self-Supervised Pre-Training of Transformers via Human Motion Forecasting for Few-Shot Gait Impairment Severity Estimation.

Parkinson's disease (PD) is a neurological disorder that has a variety of observable motor-related symptoms such as slow movement, tremor, muscular rigidity, and impaired posture. PD is typically diagnosed by evaluating the severity of motor impairments according to scoring systems such as the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS). Automated severity prediction using video recordings of individuals provides a promising route for non-intrusive monitoring of motor impairments. However, the limited size of PD gait data hinders model ability and clinical potential. Because of this clinical data scarcity and inspired by the recent advances in self-supervised large-scale language models like GPT-3, we use human motion forecasting as an effective self-supervised pre-training task for the estimation of motor impairment severity. We introduce GaitForeMer, Gait Forecasting and impairment estimation transforMer, which is first pre-trained on public datasets to forecast gait movements and then applied to clinical data to predict MDS-UPDRS gait impairment severity. Our method outperforms previous approaches that rely solely on clinical data by a large margin, achieving an F1 score of 0.76, precision of 0.79, and recall of 0.75. Using GaitForeMer, we show how public human movement data repositories can assist clinical use cases through learning universal motion representations. The code is available at https://github.com/markendo/GaitForeMer.

An automated system for screening retroviral expression constructs in microplate format.

Retroviruses are useful for genetics studies to deliver genes that express proteins, peptides, and RNAs. Several steps, including DNA preparation, transfection, packaging, transduction, and assay, are required to execute screens using retroviral constructs. Unlike screens with purified components, whole-cell assays using retroviral constructs need a large number of steps with microplate manipulations. The nature of these steps, especially the involvement of cultured mammalian cells, limits the throughput of such screens. To improve the efficiency of genetic experiments with retroviral expression vectors, an automated system for retroviral screening in microplates was devised and tested. The system, called Somata, provides high throughputs and robust, reproducible performance.

A high-throughput, homogeneous microplate assay for agents that kill mammalian tissue culture cells.

Screens for cytostasis/cytoxicity have considerable value for the discovery of therapeutic agents and the investigation of the biology of apoptosis. For instance, genetic screens for proteins, protein fragments, peptides, RNAs, or chemicals that kill tissue culture cells may aid in identifying new cancer therapeutic targets. A microplate assay for cell death is needed to achieve throughputs sufficient to sift through thousands of agents from expression or chemical libraries. The authors describe a homogeneous assay for cell death in tissue culture cells compatible with 96- or 384-well plates. In combination with a previously described system for retroviral packaging and transduction, nearly 6000 expression library clones could be screened per week in a 96-well plate format. The screening system may also prove useful for chemical screens.

Genetic selection for modulators of the MAP kinase and beta-catenin growth-control pathways in mammalian cells.

Transdominant genetic selections can yield protein fragment and peptide modulators of specific biochemical pathways. In yeast, such screens have been highly successful in targeting the MAP (mitogen-activated protein) kinase growth-control pathway. We performed a similar type of selection aimed at recovery of modulators of the mammalian MAP kinase cascade. Two pathway activators were identified, fragments of the TrkB and Raf-1 kinases. In a second selection directed at the beta-catenin growth-control pathway, three different clones encoding cadherin fragments were recovered. In neither selection were peptide inhibitors observed. We conclude that some transdominant selections in mammalian cells can readily yield high-penetrance protein fragments, but may be less amenable to isolation of peptide inhibitors.