Multi-cancer classification; an analysis of neural network complexity

Background: Cancer identification is generally framed as binary classification, normally discrimination of a control group from a single cancer group. However, such models lack any cancer-specific information, as they are only trained on one cancer type. The models fail to account for competing cancer risks. For example, an ostensibly healthy individual may have any number of different cancer types, and a tumor may originate from one of several primary sites. Pan-cancer evaluation requires a model trained on multiple cancer types, and controls, simultaneously, so that a physician can be directed to the correct area of the body for further testing. Methods: We introduce novel neural network models to address multi-cancer classification problems across several data types commonly applied in cancer prediction, including circulating miRNA expression, protein, and mRNA. In particular, we present an analysis of neural network depth and complexity, and investigate how this relates to classification performance. Comparisons of our models with state-of-the-art neural networks from the literature are also presented. Results: Our analysis evidences that shallow, feed-forward neural net architectures offer greater performance when compared to more complex deep feed-forward, Convolutional Neural Network (CNN), and Graph CNN (GCNN) architectures considered in the literature. Conclusion: The results show that multiple cancers and controls can be classified accurately using the proposed models, across a range of expression technologies in cancer prediction. Impact: This study addresses the important problem of pan-cancer classification, which is often overlooked in the literature. The promising results highlight the urgency for further research.

Ovarian Cancer Prediction Using PCA, K-PCA, ICA and Random Forest

Ovarian cancer, which is the most common in women and occurs mostly in the post-menopausal period, develops with the uncontrolled proliferation of the cells in the ovaries and the formation of tumors. Early diagnosis is very difficult and in most cases, it is a type of cancer that is in advanced stages when first diagnosed. While it tends to be treated successfully in the early stages where it is confined to the ovary, it is more difficult to treat in the advanced stages and is often fatal. For this reason, it has been focused on studies that predict whether people have ovarian cancer. In our study, we designed a RF-based ovarian cancer prediction model using a data set consisting of 49 features including blood routine tests, general chemistry tests and tumor marker data of 349 real patients. Since the data set containing too many dimensions will increase the time and resources that need to be spent, we reduced the dimension of the data with PCA, K-PCA and ICA methods and examined its effect on the result and time saving. The best result was obtained with a score of 0.895 F1 by using the new smaller-sized data obtained by the PCA method, in which the dimension was reduced from 49 to 6, in the RF method, and the training of the model took 18.191 seconds. This result was both better as a success and more economical in terms of time spent during model training compared to the prediction made over larger data with 49 features, where no dimension reduction method was used. The study has shown that in predictions made with machine learning models over large-scale medical data, dimension reduction methods will provide advantages in terms of time and resources by improving the prediction results.