Prostate cancer (PCa) is a severe type of cancer and causes major deaths among men due to its poor diagnostic system. The images obtained from patients with carcinoma consist of complex and necessary features that cannot be extracted readily by traditional diagnostic techniques. This research employed deep learning long short-term memory $(LSTM)$ and Residual Net $(ResNet-101)$ , independent of hand-crafted features, and is fine-tuned. The results were compared with hand-crafted features such as texture, morphology, and gray level co-occurrence matrix $(GLCM)$ using non-deep learning classifiers such as support vector machine $(SVM)$ Gaussian Kernel, k-nearest neighbor-Cosine $(KNN-Cosine)$ , kernel naive Bayes, decision tree $(DT)$ and RUSBoost tree. This study reduces the features of carcinoma images, employed machine learning and deep learning approaches. For validation of training and testing data, a jack-knife ten-fold cross-validation method was used. The performance was measured using a confusion matrix such as sensitivity, specificity, positive predictive value $(PPV)$ , negative predictive value $(NPV)$ , accuracy $(AC)$ , Mathews Correlation Coefficient ( $MCC$ ), and area under the curve $(AUC)$ . The most remarkable performance was obtained using non-deep learning methods with $GLCM$ features using KNN-Cosine with sensitivity (98.00%), specificity (99.25%), PPV (98.99%), NPV (99.11%), accuracy (99.07%), and AUC (0.998). The LSTM deep learning method yields performance with sensitivity (98.33%), specificity (100%), PPV (100%), NPV (99.26%), accuracy (99.48%), MCC (0.9879) and AUC (0.9999), where using Deep learning method $ResNet-101$ , we obtained (100%) Accuracy and AUC (1) for Kernel Naive Bayes, SVM Gaussian and RUSBoost Tree. The results show that $ResNet-101$ deep learning outperformed than non-deep learning methods and $LSTM$ . Thus, the deep learning method $ResNet-101$ could be used as a better predictor for the detection of prostate cancer.

Prostate is a second leading causes of cancer deaths among men. Early detection of cancer can effectively reduce the rate of mortality caused by Prostate cancer. Due to high and multiresolution of MRIs from prostate cancer require a proper diagnostic

The incidence of congestive heart failure (CHF) is approximately 10 per 1000 for Americans over the age of 65 years. The dynamics of CHF are highly complex, nonlinear, and temporal dynamics. Based on these characteristics, we extracted multimodal features from congestive heart failure (CHF) and normal sinus rhythm (NSR) signals. We performed the synthetic minority over-sampling technique (SMOTE) to increase the number of CHF subjects to balance our train data. The classification between these subjects with original data and SMOTE data was performed using machine learning classifiers such as classification and regression tree (CART), support vector machine linear (SVM-L), Naïve Bayes, neural network, and ensemble classifiers such as random forest (RF), XG boost, averaged neural network (AVNNET). With the original data, the highest performance was obtained using SVM-L with accuracy (94.28%), sensitivity (84.61%), specificity (100%), p-value (0.0002), AUC (0.9605) with 95% CI: 0.9006-1.00. By applying the SMOET, the highest performance was obtained with SVM-L with accuracy (97.14%), sensitivity (92.30%), specificity (100%), p-value (7.99e-06), AUC (0.9650) with 95% CI: 0.8945–1.00. The results reveal that proposed approach with SMOTE improved the detection performance which can be very effective and computationally efficient tool for automatic detection of congestive heart failure patients.

Lung Cancer is one of the leading causes of cancer-related deaths worldwide with minimal survival rate due to poor diagnostic system at the advanced cancer stage. In the past, researchers developed various tools in image processing to detect the Lung cancer of types as non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) which are based on few features extracting methods. In this research, we extracted multimodal features such as texture, morphological, entropy based, scale invariant Fourier transform (SIFT), Ellipse Fourier Descriptors (EFDs) by considering multiple aspects and shapes morphologies. We then applied robust machine learning classification methods such as Naïve Bayes, Decision Tree and Support Vector Machine (SVM) with its kernels such as Gaussian, Radial Base Function (RBF) and Polynomial. Jack-knife 10-fold cross validation was applied for training/ validation of data. The performance was evaluated in terms of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), total accuracy (TA), false positive rate (FPR) and area under the receiving curve (AUC). The highest detection accuracy was obtained with (TA=100%) with entropy, SIFT and texture features using Naïve Bayes, texture features using SVM Polynomial. Moreover, the highest separation was obtained using entropy, morphological, SIFT and texture features with (AUC=1.00) using Naïve Bayes classifier and texture features using Decision tree and SVM polynomial kernel.

Cardiac arrhythmias are disturbances in the rhythm of the heart manifested by irregularity or by abnormally fast rates (‘tachycardia’) or abnormally slow rates (‘bradycardias’). In the past researchers extracted different features extracting strategies to detect the arrhythmia. Since, signals acquired from subjects suffered with arrhythmia are multivariate, highly nonlinear, nonstationary, time variant and very complex. Traditional features extraction approaches could not capture full dynamics to detect these signals. In this study, we extracted hybrid features based on refined sample entropy (SampEn) and Fuzzy entropy (FuzEn) computed on mean, standard deviation and variance in combination with traditional features. The machine learning support vector machines with its kernels, decision tree, KNN and ensemble methods were employed to detect the arrhythmia. The 10-fold cross-validation was used for training/validation of data. The Support Vector Machine and its kernels, Fine KNN, and Ensemble Boosted Tree give the highest performance with an overall accuracy of 100%, AUC of 1.00. An accuracy of 95.5% and 100% was obtained using SVM coarse Gaussian, 97% and 100% using Fine KNN, 95.5% and 97% using Decision tree with traditional and hybrid features respectively. The results revealed that SVM coarse Gaussian, Coarse KNN, Decision tree methods outer performed using hybrid features than the traditional feature extracting strategy.