The potential of convolutional neural network in the evaluation of tumor-stroma ratio from colorectal cancer histopathological images
Authors
Date
2022Copyright
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Tässä Pro gradu-työssä tutkitaan konvoluutioneuroverkkojen käyttömahdollisuuksia histopatologisista kuvista tehtävässä kasvain-strooma suhdeluvun arvioinnissa. Tarkoituksena on selvittää, mikä on siirto-opettamisen vaikutus, kun opettamisessa käytetään kohdealuespesifistä dataa. Mallin ennustamaa kasvain-strooma suhdelukua verrataan patologin visuaalisesti tekemään arvioon. Tutkimuksesta selvisi, että kohdealuespesifisen datan käyttö esiopetuksessa lisää konvoluutioneuroverkkomallin tarkkuutta. Myös korrelaatiota ennustetun ja visuaalisen arvion välillä
oli havaittavissa. Tulevaisuudessa olisi hyvä tutkia kasvain-strooma-suhdeluvun yhteyttä muihin kliinispatologisiin tekijöihin ja potilaan elinaikaan. In this Master’s Thesis, the ability of convolutional neural networks in the evaluation of tumor-stroma ratio from histopathological images, is studied. The goal is to find
out, whether pre-training with domain-specific data brings more accuracy to the convolutional neural network model. Tumor-stroma ratio is predicted with the trained model and the predicted values are compared with visual tumor-stroma estimations made by pathologist. When domain-specific data was used in the pre-training of the convolutional neural network, a slight improvement in the validation accuracy of the model was observed. Correlation between the predicted and visual values was also found. Further analysis is needed to study what is the connection of these computationally predicted values to other clinicopathological factors and overall survival of the patient.
Keywords
Metadata
Show full item recordCollections
- Pro gradu -tutkielmat [24521]
Related items
Showing items with similar title or keywords.
-
Causality-Aware Convolutional Neural Networks for Advanced Image Classification and Generation
Terziyan, Vagan; Vitko, Oleksandra (Elsevier, 2023)Smart manufacturing uses emerging deep learning models, and particularly Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs), for different industrial diagnostics tasks, e.g., classification, ... -
Transformers for breast cancer classification
Lindroos, Jari (2022)Rintasyöpä on maailmanlaajuisesti naisten yleisin syöpä, sen varhainen havaitseminen voi merkittävästi vähentää siihen liittyvää kuolleisuutta. Histopatologista analyysiä tarvitaan kasvainten laadun määrittämiseksi ... -
Hyper-flexible Convolutional Neural Networks based on Generalized Lehmer and Power Means
Terziyan, Vagan; Malyk, Diana; Golovianko, Mariia; Branytskyi, Vladyslav (Elsevier, 2022)Convolutional Neural Network is one of the famous members of the deep learning family of neural network architectures, which is used for many purposes, including image classification. In spite of the wide adoption, such ... -
Robustness, Stability, and Fidelity of Explanations for a Deep Skin Cancer Classification Model
Saarela, Mirka; Geogieva, Lilia (MDPI AG, 2022)Skin cancer is one of the most prevalent of all cancers. Because of its being widespread and externally observable, there is a potential that machine learning models integrated into artificial intelligence systems will ... -
CCTVCV : Computer Vision model/dataset supporting CCTV forensics and privacy applications
Turtiainen, Hannu; Costin, Andrei; Hämäläinen, Timo; Lahtinen, Tuomo; Sintonen, Lauri (IEEE, 2022)The increased, widespread, unwarranted, and unaccountable use of Closed-Circuit TeleVision (CCTV) cameras globally has raised concerns about privacy risks for the last several decades. Recent technological advances implemented ...