Artificial intelligence for imaging-based COVID-19 detection: Systematic review comparing added value of AI versus human readers

PURPOSE

A growing number of studies have examined whether Artificial Intelligence (AI) systems can support imaging-based diagnosis of COVID-19-caused pneumonia, including both gains in diagnostic performance and speed. However, what is currently missing is a combined appreciation of studies comparing human readers and AI.

METHODS

We followed PRISMA-DTA guidelines for our systematic review, searching EMBASE, PUBMED and Scopus databases. To gain insights into the potential value of AI methods, we focused on studies comparing the performance of human readers versus AI models or versus AI-supported human readings.

RESULTS

Our search identified 1270 studies, of which 12 fulfilled specific selection criteria. Concerning diagnostic performance, in testing datasets reported sensitivity was 42-100% (human readers, n = 9 studies), 60-95% (AI systems, n = 10) and 81-98% (AI-supported readers, n = 3), whilst reported specificity was 26-100% (human readers, n = 8), 61-96% (AI systems, n = 10) and 78-99% (AI-supported readings, n = 2). One study highlighted the potential of AI-supported readings for the assessment of lung lesion burden changes, whilst two studies indicated potential time savings for detection with AI.

CONCLUSIONS

Our review indicates that AI systems or AI-supported human readings show less performance variability (interquartile range) in general, and may support the differentiation of COVID-19 pneumonia from other forms of pneumonia when used in high-prevalence and symptomatic populations. However, inconsistencies related to study design, reporting of data, areas of risk of bias, as well as limitations of statistical analyses complicate clear conclusions. We therefore support efforts for developing critical elements of study design when assessing the value of AI for diagnostic imaging.

Lymphocytes are a major source of circulating soluble dipeptidyl peptidase 4

Dipeptidyl peptidase 4 (DPP4, CD26) is a serine protease that is expressed constitutively by many haematopoietic and non-haematopoietic tissues. It exists as a membrane-associated protein, as well as in an active, soluble form (herein called sDPP4), present at high concentrations in bodily fluids. Despite the proposed use of sDPP4 as a biomarker for multiple diseases, its cellular sources are not well defined. Here, we report that individuals with congenital lymphocyte immunodeficiency had markedly lower serum concentrations of sDPP4, which were restored upon successful treatment and restoration of lymphocyte haematopoiesis. Using irradiated lymphopenic mice and wild-type to Dpp4-/- reciprocal bone marrow chimeric animals, we found that haematopoietic cells were a major source of circulating sDPP4. Furthermore, activation of human and mouse T lymphocytes resulted in increased sDPP4, providing a mechanistic link between immune system activation and sDPP4 concentration. Finally, we observed that acute viral infection induced a transient increase in sDPP4, which correlated with the expansion of antigen-specific CD8+ T cell responses. Our study demonstrates that sDPP4 concentrations are determined by the frequency and activation state of lymphocyte populations. Insights from these studies will support the use of sDPP4 concentration as a biomarker for inflammatory and infectious diseases.

Autoimmune dysregulation and purine metabolism in adenosine deaminase deficiency

Genetic defects in the adenosine deaminase (ADA) gene are among the most common causes for severe combined immunodeficiency (SCID). ADA-SCID patients suffer from lymphopenia, severely impaired cellular and humoral immunity, failure to thrive, and recurrent infections. Currently available therapeutic options for this otherwise fatal disorder include bone marrow transplantation (BMT), enzyme replacement therapy with bovine ADA (PEG-ADA), or hematopoietic stem cell gene therapy (HSC-GT). Although varying degrees of immune reconstitution can be achieved by these treatments, breakdown of tolerance is a major concern in ADA-SCID. Immune dysregulation such as autoimmune hypothyroidism, diabetes mellitus, hemolytic anemia, and immune thrombocytopenia are frequently observed in milder forms of the disease. However, several reports document similar complications also in patients on long-term PEG-ADA and after BMT or GT treatment. A skewed repertoire and decreased immune functions have been implicated in autoimmunity observed in certain B-cell and/or T-cell immunodeficiencies, but it remains unclear to what extent specific mechanisms of tolerance are affected in ADA deficiency. Herein we provide an overview about ADA-SCID and the autoimmune manifestations reported in these patients before and after treatment. We also assess the value of the ADA-deficient mouse model as a useful tool to study both immune and metabolic disease mechanisms. With focus on regulatory T- and B-cells we discuss the lymphocyte subpopulations particularly prone to contribute to the loss of self-tolerance and onset of autoimmunity in ADA deficiency. Moreover we address which aspects of immune dysregulation are specifically related to alterations in purine metabolism caused by the lack of ADA and the subsequent accumulation of metabolites with immunomodulatory properties.

Lentiviral-mediated gene therapy leads to improvement of B-cell functionality in a murine model of Wiskott-Aldrich syndrome

BACKGROUND

Wiskott-Aldrich syndrome (WAS) is an X-linked primary immunodeficiency characterized by thrombocytopenia, eczema, infections, autoimmunity, and lymphomas. Transplantation of hematopoietic stem cells from HLA-identical donors is curative, but it is not available to all patients. We have developed a gene therapy (GT) approach for WAS by using a lentiviral vector encoding for human WAS promoter/cDNA (w1.6W) and demonstrated its preclinical efficacy and safety.

OBJECTIVE

To evaluate B-cell reconstitution and correction of B-cell phenotype in GT-treated mice.

METHODS

We transplanted Was(-/-) mice sublethally irradiated (700 rads) with lineage marker-depleted bone marrow wild-type cells, Was(-/-) cells untransduced or transduced with the w1.6W lentiviral vector and analyzed B-cell reconstitution in bone marrow, spleen, and peritoneum.

RESULTS

Here we show that WAS protein(+) B cells were present in central and peripheral B-cell compartments from GT-treated mice and displayed the strongest selective advantage in the splenic marginal zone and peritoneal B1 cell subsets. After GT, splenic architecture was improved and B-cell functions were restored, as demonstrated by the improved antibody response to pneumococcal antigens and the reduction of serum IgG autoantibodies.

CONCLUSION

WAS GT leads to improvement of B-cell functions, even in the presence of a mixed chimerism, further validating the clinical application of the w1.6W lentiviral vector.