Patricia E. Velez-Varela

Glioblastoma, IDH-wildtype (GBM), is the most aggressive and complex brain tumour classified by the World Health Organization (WHO), characterised by high mortality rates and diagnostic limitations inherent to invasive conventional procedures. Early detection is essential for improving patient outcomes, underscoring the need for non-invasive diagnostic tools. This study presents a convolutional neural network (CNN) specifically optimised for GBM detection from T1-weighted magnetic resonance imaging (MRI), with systematic evaluations of layer depth, activation functions, and hyperparameters. The model was trained on the RSNA-MICCAI data set and externally validated on the Erasmus Glioma Database (EGD), which includes gliomas of various grades and preserves cranial structures, unlike the skull-stripped RSNA-MICCAI images. This morphological discrepancy demonstrates the generalisation capacity of the model across anatomical and acquisition differences, achieving an F1-score of 0.88. Furthermore, statistical tests, such as Shapiro–Wilk, Mann–Whitney U, and Chi-square, confirmed the absence of demographic bias in model predictions, based on p-values, confidence intervals, and statistical power analyses supporting its demographic fairness. The proposed model achieved an area under the curve–receiver operating characteristic (AUC-ROC) of 0.63 on the RSNA-MICCAI test set, surpassing all prior results submitted to the BraTS 2021 challenge, and establishing a reliable and generalisable approach for non-invasive GBM detection.

Background: Glioblastoma multiforme (GBM) is an aggressive brain tumor with a poor prognosis. Traditional diagnosis relies on invasive biopsies, which pose surgical risks. Advances in artificial intelligence (AI) and machine learning (ML) have improved non-invasive GBM diagnosis using magnetic resonance imaging (MRI), offering potential advantages in accuracy and efficiency. Objective: This review aims to identify the methodologies and technologies employed in AI-based GBM diagnostics. It further evaluates the performance of AI models using standard metrics, highlighting both their strengths and limitations. Methodology: In accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines, a systematic review was conducted across major academic databases. A total of 104 articles were identified in the initial search, and 15 studies were selected for final analysis after applying inclusion and exclusion criteria. Outcomes: The included studies indicated that the signal T1-weighted imaging (T1WI) is the most frequently used MRI modality in AI-based GBM diagnostics. Multimodal approaches integrating T1WI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) have demonstrated improved classification performance. Additionally, AI models have shown potential in surpassing conventional diagnostic methods, enabling automated tumor classification and enhancing prognostic predictions.

Introduction and importance: BIA-ALCL is a non-Hodgkin lymphoma occurring primarily in women with textured breast implants, typically presenting as late seroma. Diagnosis involves ultrasound-guided fine-needle aspiration or core needle biopsy, followed by cytologic and immunohistochemical evaluation. Positive results show CD30 cell expression and lack ALK expression. Treatment includes removing breast implants and the periprosthetic capsule. If the lymphoma has spread, en bloc capsulectomy, immunotherapy, and chemotherapy are required. Reconstruction can be done with smooth implants or autologous tissue. Presentation of case: We present here the case of a woman with a 12-year history of microtextured breast implants, without any cancer family background, who presented with peri-implant effusion in the left breast, which tested positive for BIA-ALCL at stage IA. The patient underwent breast implant removal with total capsulectomy and posterior successful autologous tissue reconstruction, demonstrating that outcomes can be satisfactory when properly treated. Clinical discussion: Paying attention to signs in women with breast implants, beyond imaging tests, can aid in the early diagnosis of BIA-ALCL and ensure a not aggressive treatment. This approach allows the reconstruction with autologous tissue without the need of further implants. Conclusion: BIA-ALCL is a rare disease, further studies about this lymphoma can help with early diagnosis and potential prevention. Highlights:

Cancer is one of the leading causes of mortality worldwide. Breast cancer is the most frequent cancer in women, and in recent years it has become a serious public health problem in Colombia. The development of large-scale omic techniques allows simultaneous analysis of all active genes in tumor cells versus normal cells, providing new ways to discover the drivers of malignant transformations. Whole exome sequencing (WES) was obtained to provide a deep view of the mutational genomic profile in a set of cancer samples from Southwest Colombian women. WES was performed on 52 tumor samples from patients diagnosed with invasive breast cancer, which in most cases (33/52) were ductal luminal breast carcinomas (IDC-LM-BRCA). Global variant call was calculated, and six different algorithms were applied to filter out false positives and identify pathogenic variants. To compare and expand the somatic tumor variants found in the Colombian cohort, exome mutations and genome-wide expression alterations were detected in a larger set of tumor samples of the same breast cancer subtype from TCGA (that included DNA-seq and RNA-seq data). Genes with significant changes in both the mutational and expression profiles were identified, providing a set of genes and mutations associated with the etiology of ductal luminal breast cancer. This set included 19 single mutations identified as tumor driver mutations in 17 genes. Some of the genes (ATM, ERBB3, ESR1, TP53) are well-known cancer genes, while others (CBLB, PRPF8) presented driver mutations that had not been reported before. In the case of the CBLB gene, several mutations were identified in TCGA IDC-LM-BRCA samples associated with overexpression of this gene and repression of tumor suppressive activity of TGF-β pathway.

Repetitive DNA sequences (Repeats) are significant regions in the human genome that have a specific genomic distribution, structure, and several binding sites for genome architecture and function. In consequence, the possible configurations of Repeats in specific and dynamic regions like the gene promoters could define footprints for molecular mechanisms, pathways, and cell function beyond their density in the genome. Here we explored the distribution of Repeats in the upstream promoter region of the human coding genes with the aim to identify specific configurations, clusters and functional meaning of those elements. Our method includes structural descriptions, hierarchical clustering, pathway association, and functional enrichment analysis. We report here several configurations of Repeats in the upstream promoter region (UPR), which define 2729 patterns for the 80% of the human coding genes. There are 47 types of Repeats in these configurations, where the most frequent were Alu, Low_complexity, MIR, Simple_repeat, LINE/L2, LINE/L1, hAT-Charlie, and ERV1. The distribution, length, and the high frequency of Repeats in the UPR defines several patterns and clusters, where the minimum frequency of configuration among Repeats was higher than 0.7. We found those clusters associated with cellular pathways and ontologies; thus, it was plausible to determine groups of Repeats to specific functional insights, for example, pathways for Genetic Information Processing or Metabolism shows particular groups of Repeats with specific configurations. Based on these findings, we propose that specific configurations of repetitive elements describe frequent patterns in the upstream promoter for sets of human coding genes, which those correlated to specific and essential cell pathways and functions.

Several studies have shown that genomes can be studied via a multifractal formalism. Recently, we used a multifractal approach to study the genetic information content of the Caenorhabditis elegans genome. Here we investigate the possibility that the human genome shows a similar behavior to that observed in the nematode.