scholarly journals Application of Artificial Neural Networks for Prognostic Modeling in Lung Cancer after Combining Radiomic and Clinical Features

2019 ◽  
Vol 05 (02) ◽  
pp. 050-055
Author(s):  
Kundan S. Chufal ◽  
Irfan Ahmad ◽  
Anjali K. Pahuja ◽  
Alexis A. Miller ◽  
Rajpal Singh ◽  
...  

Abstract Objective This study was aimed to investigate machine learning (ML) and artificial neural networks (ANNs) in the prognostic modeling of lung cancer, utilizing high-dimensional data. Materials and Methods A computed tomography (CT) dataset of inoperable nonsmall cell lung carcinoma (NSCLC) patients with embedded tumor segmentation and survival status, comprising 422 patients, was selected. Radiomic data extraction was performed on Computation Environment for Radiation Research (CERR). The survival probability was first determined based on clinical features only and then unsupervised ML methods. Supervised ANN modeling was performed by direct and hybrid modeling which were subsequently compared. Statistical significance was set at <0.05. Results Survival analyses based on clinical features alone were not significant, except for gender. ML clustering performed on unselected radiomic and clinical data demonstrated a significant difference in survival (two-step cluster, median overall survival [ mOS]: 30.3 vs. 17.2 m; p = 0.03; K-means cluster, mOS: 21.1 vs. 7.3 m; p < 0.001). Direct ANN modeling yielded a better overall model accuracy utilizing multilayer perceptron (MLP) than radial basis function (RBF; 79.2 vs. 61.4%, respectively). Hybrid modeling with MLP (after feature selection with ML) resulted in an overall model accuracy of 80%. There was no difference in model accuracy after direct and hybrid modeling (p = 0.164). Conclusion Our preliminary study supports the application of ANN in predicting outcomes based on radiomic and clinical data.

Atmosphere ◽  
2018 ◽  
Vol 9 (2) ◽  
pp. 77 ◽  
Author(s):  
Henrique do Nascimento Camelo ◽  
Paulo Sérgio Lucio ◽  
João Verçosa Leal Junior ◽  
Daniel von Glehn dos Santos ◽  
Paulo Cesar Marques de Carvalho

2017 ◽  
Vol 234 ◽  
pp. 13-18
Author(s):  
Rafaela Beatriz Pintor Torrecilha ◽  
Yuri Tani Utsunomiya ◽  
Luís Fábio da Silva Batista ◽  
Anelise Maria Bosco ◽  
Cáris Maroni Nunes ◽  
...  

2020 ◽  
Vol 17 (4) ◽  
pp. 1069-1078
Author(s):  
Cengiz Akarçeşme ◽  
Hasan Aka ◽  
Semih Özden ◽  
Zait Burak Aktuğ

This study was conducted to estimate the Olympic ranking of the games played in the qualifying groups by the countries that were qualified for the 2016 Rio Olympics in volleyball branch by analyzing with the developed artificial neural networks (ANN) and linear equation model. In the study, the difficulty level of all games (n=324) that total 22 teams played in the qualifying for the 2016 Rio Olympics in volleyball branch (11 female and 11 male volleyball teams)  and International Volleyball Federation (FIVB) ranking score was evaluated separately. Feedforward network structure having two hidden layers in the modeling with ASS developed for 9 different input variables was preferred in the study. In addition, linear modeling method, which provides an easier calculation than artificial neural networks, was performed by “regress” instruction in MATLAB. In the female group, the percentage mean error value of the models was calculated as 18.86 by ANN model, and as 4.53 by linear model. In male groups, it was calculated as 19,34 by ANN model, and as 0,74 by linear model. According to the modeling results obtained in the study, both female and male volleyball teams’ results were modeled with a higher accuracy by linear model. As a result, team rankings of the volleyball branch in the women's group in the 2016 Rio Olympic Games was estimated with an accuracy over 98% separately by ANN modeling regression results and linear modeling regression results. In men’s volleyball games, it was estimated with an accuracy over 98% by ANN modeling regression results, and with an accuracy over 99% by linear modeling regression results. It can be stated that the difficulty level of the games that countries participating in Olympics in volleyball branch played in the qualifying groups and FIVB ranking scores are among the variables that have a significant effect on determining the Olympic ranking. ​Extended English summary is in the end of Full Text PDF (TURKISH) file.   Özet   Bu çalışma 2016 Rio Olimpiyatlarına voleybol branşından katılma hakkı kazanan ülkelerin eleme gruplarındaki maçlarının, geliştirilen yapay sinir ağları (YSA) ve doğrusal eşitlik modeli ile analiz edilerek olimpiyat sıralamasının tahmin edilmesi amacı ile yapılmıştır. Çalışmada 2016 Rio Olimpiyatlarına voleybol branşından katılan (11 kadın ve 11 erkek voleybol takımı) toplam 22 takımın grup elemelerinde oynadığı tüm maçlar (n=324) zorluk derecesi ve Uluslararası Voleybol Federasyonu (FIVB) sıralama puanı göz önüne alınarak değerlendirilmiştir. Çalışmada dokuz farklı giriş değişkenine göre geliştirilen YSA modeli  ile modellemede iki gizli katmana sahip ileri yayılımlı ağ yapısı tercih edilmiştir. Ayrıca çalışmada YSA’na göre çok daha basit bir hesaplama sağlayan doğrusal modelleme yöntemi de, MATLAB’de bulunan “regress” komutu ile gerçekleştirilmiştir. Kadınlar grubunda; test verilerine bakıldığında modellerin yüzde ortalama hata değeri, YSA modelinde 18.86, doğrusal modelde 4.53 olarak; erkekler grubunda ise YSA modelinde 19.34, doğrusal modelde 0.74 olarak hesaplanmıştır. Çalışmada elde edilen modelleme sonuçlarına göre; hem kadın hem de erkek voleybol takımlarının sonuçları doğrusal model ile daha yüksek doğrulukla modellenmiştir. Sonuç olarak, kadınlar kategorisinde 2016 Rio Olimpiyat Oyunlarında voleybol branşının takım sıralaması, YSA modelleme regresyon sonuçları ve doğrusal modelleme regresyon sonuçları ile ayrı ayrı %98’in üstünde doğrulukla tahmin edilmiştir. Erkek voleybol maçlarında ise YSA modelleme regresyon sonuçları %98’in üstünde, doğrusal modelleme regresyon sonuçları ise %99’un üstünde doğrulukla tahmin edilmiştir. Voleybolda Olimpiyatlara katılan ülkelerin eleme gruplarında oynadıkları maçların zorluk derecesi ve FIVB sıralama puanlarının Olimpiyat sıralamasının belirlenmesine önemli etkisi olan değişkenlerden olduğu söylenebilir.


Sign in / Sign up

Export Citation Format

Share Document