Prediction of clinically actionable genetic alterations from colorectal cancer histopathology images using deep learning

Sporadic Colorectal Cancer (sCRC) is the third leading cause of cancer death in the Western world, and the sCRC patients presenting with synchronic metastasis have the poorest prognosis. Genetic alterations accumulated in sCRC tumor cells translate into mutated proteins and/or abnormal protein expression levels, which contribute to the development of sCRC. Then, the tumor-associated proteins (TAAs) might induce the production of auto-antibodies (aAb) via humoral immune response. Here, Nucleic Acid Programmable Protein Arrays (NAPPArray) are employed to identify aAb in plasma samples from a set of 50 sCRC patients compared to seven healthy donors. Our goal was to establish a systematic workflow based on NAPPArray to define differential aAb profiles between healthy individuals and sCRC patients as well as between non-metastatic (n = 38) and metastatic (n = 12) sCRC, in order to gain insight into the role of the humoral immune system in controlling the development and progression of sCRC. Our results showed aAb profile based on 141 TAA including TAAs associated with biological cellular processes altered in genesis and progress of sCRC (e.g., FSCN1, VTI2 and RPS28) that discriminated healthy donors vs. sCRC patients. In addition, the potential capacity of discrimination (between non-metastatic vs. metastatic sCRC) of 7 TAAs (USP5, ML4, MARCKSL1, CKMT1B, HMOX2, VTI2, TP53) have been analyzed individually in an independent cohort of sCRC patients, where two of them (VTI2 and TP53) were validated (AUC ~75%). In turn, these findings provided novel insights into the immunome of sCRC, in combination with transcriptomics profiles and protein antigenicity characterizations, wich might lead to the identification of novel sCRC biomarkers that might be of clinical utility for early diagnosis of the tumor. These results explore the immunomic analysis as potent source for biomarkers with diagnostic and prognostic value in CRC. Additional prospective studies in larger series of patients are required to confirm the clinical utility of these novel sCRC immunomic biomarkers.

Download Full-text

A Design of Polygenic Risk Model with Deep Learning for Colorectal Cancer in Multiethnic Indonesians

Procedia Computer Science ◽

10.1016/j.procs.2021.01.049 ◽

2021 ◽

Vol 179 ◽

pp. 632-639

Author(s):

Steven Amadeus ◽

Tjeng Wawan Cenggoro ◽

Arif Budiarto ◽

Bens Pardamean

Keyword(s):

Colorectal Cancer ◽

Deep Learning ◽

Risk Model ◽

Polygenic Risk

Download Full-text

Two Ensemble-CNN Approaches for Colorectal Cancer Tissue Type Classification

Journal of Imaging ◽

10.3390/jimaging7030051 ◽

2021 ◽

Vol 7 (3) ◽

pp. 51

Author(s):

Emanuela Paladini ◽

Edoardo Vantaggiato ◽

Fares Bougourzi ◽

Cosimo Distante ◽

Abdenour Hadid ◽

...

Keyword(s):

Colorectal Cancer ◽

Deep Learning ◽

Digital Pathology ◽

Texture Features ◽

Tissue Type ◽

Cancer Tissue ◽

Learning Methods ◽

Feature Based ◽

Type Classification ◽

Whole Slide Images

In recent years, automatic tissue phenotyping has attracted increasing interest in the Digital Pathology (DP) field. For Colorectal Cancer (CRC), tissue phenotyping can diagnose the cancer and differentiate between different cancer grades. The development of Whole Slide Images (WSIs) has provided the required data for creating automatic tissue phenotyping systems. In this paper, we study different hand-crafted feature-based and deep learning methods using two popular multi-classes CRC-tissue-type databases: Kather-CRC-2016 and CRC-TP. For the hand-crafted features, we use two texture descriptors (LPQ and BSIF) and their combination. In addition, two classifiers are used (SVM and NN) to classify the texture features into distinct CRC tissue types. For the deep learning methods, we evaluate four Convolutional Neural Network (CNN) architectures (ResNet-101, ResNeXt-50, Inception-v3, and DenseNet-161). Moreover, we propose two Ensemble CNN approaches: Mean-Ensemble-CNN and NN-Ensemble-CNN. The experimental results show that the proposed approaches outperformed the hand-crafted feature-based methods, CNN architectures and the state-of-the-art methods in both databases.

Download Full-text

Stacking and Chaining of Normalisation Methods in Deep Learning-based Classification of Colorectal Cancer using Gut Microbiome Data

IEEE Access ◽

10.1109/access.2021.3094529 ◽

2021 ◽

pp. 1-1

Author(s):

Mwenge Mulenga ◽

Sameem Abdul Kareem ◽

Aznul Qalid Md Sabri ◽

Manjeevan Seera

Keyword(s):

Colorectal Cancer ◽

Deep Learning ◽

Gut Microbiome ◽

Microbiome Data

Download Full-text