Prostate cancer is the 2nd most commonly occurring male cancer and the 4th most common cancer overall. Early detection and diagnosis are important for clinical treatment. Atomic force microscopy (AFM)-based techniques have been shown to have potential in detecting malignant cancers and artificial intelligence can improve the accuracy of diagnostic and prognostic prediction tests. In this study, the classification of AFM images of prostate cells was performed using machine learning. For early prediction, we used the support vector machine (SVM) to classification prostate cells and compare the classification performance with the remaining four conventional classifiers such as logistic regression (LR), stochastic gradient descent (SGD), K-nearest neighbours (KNN), random forest (RF). Most of the classifiers did well after using the feature selection method (BorutaShap). The results show that the accuracy (ACC) of the features selected using the BorutaShap algorithm combined with the SVM classifier can reach 82.5%. Our current study demonstrates that AFM imaging combined with machine learning can be used to identify prostate cancer cells with an effective classification performance and robustness.
Ovarian cancer is a disease with a high mortality rate in women. The important reasons for high mortality rate of ovarian cancer is the difficulty in early detection. The process of cell carcinogenesis is often accompanied by changes in surface nanostructure of cell membrane. In this study, atomic force microscopy (AFM) was used to obtain the nanostructure features of ovarian cancer cells. IOSE-80 (human ovarian normal cells) and Caov3 (human ovarian cancer cells) cell lines were selected and the morphology of the cell nuclear regions were measured using AFM Quantitative Imaging (QI) mode, which can offer information of hight, adhesion and slope channels. The surface parameters of the cell obtained from the three channels were analyzed. The results showed that there were significant statistical differences in parameters Root-mean-square height (Sq), Skewness (Ssk), Maximum height (Sz) and Arithemetic mean height (Sa) of adhesion channel, Sq, Ssk and Sa of hight channel. These findings indicate that the three channel in AFM imaging can offer different information of the surface nanostructure and the combination of these feature parameters may improve the identification accuracy of cancer. Our study will provide a new idea for the early diagnosis of ovarian cancer based on the nanostructure features of cell surface at the single-cell level.
Artificial intelligence techniques have been deeply involved in the heterogeneous data aspects of biomedical applications. However, the high dimensionality and computational complexity of data can make classification, pattern recognition and data visualization difficult. Choosing appropriate dimensionality reduction techniques can help increase processing speed, reduce the time and effort required to extract valuable information, and ensure high accuracy. In this study, Alzheimer's disease data were taken as an example. Individual cases with missing values were removed, and non-digital data were converted to digital data using Min-Max normalization. Then principal component analysis (PCA) was applied to map the original feature space to 1 dimension and the variance of the validation set was calculated by 5-fold cross-validation to find the appropriate K value. The results showed that when PCA was applied to reduce the data to 1 dimension, the AUC (95% confidence interval) of the KNN classifier reached 0.898 ± 0.014, which was 30.4%higher than the case without PCA. Our current findings suggest that in many busy clinics and hospitals, it is quite worthwhile to use dimensionality reduction methods to save model computing time and to use KNN models to obtain better accuracy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.