Basics and Background

1. Machine Learning Algorithms

This section provides a comprehensive overview of the most commonly used machine learning algorithms in disease diagnosis.

1.1 Decision Trees

Decision tree (DT) algorithms follow the rule of partitioning. In a DT model, attributes may have various values called classification trees; leaves represent different classes, while branches reflect the combination of features that lead to these class labels. DT, on the other hand, may take continuous variables called regression trees. C4.5 and EC4.5 are two of the well-known and most widely used DT algorithms.

1.2 Support vector machines

For classification and regression-related challenges, support vector machines (SVMs) are a popular machine learning method. Support vector machines were introduced by Vapnik in the late 20th century. In addition to disease diagnosis, support vector machines have been used in a variety of other disciplines, including facial expression recognition, protein folding, distant homolog discovery, speech recognition, and text classification. For unlabelled data, supervised ML algorithms cannot perform. Using hyperplane discovery for clustering between data, support vector machines can classify unlabelled data. However, the output of a support vector machine is not non-linearly differentiable. To overcome these problems, the selection of the appropriate kernel and parameters are two key factors for the application of SVMs in data analysis.

1.3 k - Nearest Neighbour

(KNN) classification is a non-parametric classification technique invented by Evelyn Fix and Joseph Hodges in 1951. kNN is suitable for both classification and regression analysis. kNN classification results in class affiliation. A voting mechanism is used to classify items. The Euclidean distance technique is used to determine the distance between two data samples. The projection value for regression analysis is the average of the KNN values.

Click here to get the effect of herbal Cistanches

1.4 Naïve Bayes

The naïve Bayesian (NB) classifier is a Bayesian-based probabilistic classifier. Based on a given record or data point, it predicts the probability of membership in each class. The most probable class is the one with the highest probability. the NB classifier is used to predict probabilities, not predictions.

1.5 Logistic regression

Logistic regression (LR) is a machine-learning method used to solve classification problems. LR models have a probabilistic framework with projection values ranging from 0 to 1. Examples of LR-based ML include spam identification, online fraudulent transaction detection, and malignancy detection. The surrogate function, often referred to as the sigmoid function, is used by LR. sigmoid functions transform every real number between 0 and 1.

1.6 AdaBoost algorithm

Yoav Freund and Robert Schapire developed Adaptive Boosting, commonly known as AdaBoost. AdaBoost is a classifier that combines multiple weak classifiers into a single classifier. AdaBoost works by giving more weight to samples that are harder to classify and giving those that are already well classified AdaBoost works by giving more weight to samples that are harder to classify and less weight to those that are already well classified. It can be used for classification as well as regression analysis.

the benefits of Cistanche tubulosa

2. Deep Learning Overview

Deep learning (DL) is a subfield of machine learning (ML) that uses multiple levels to extract higher and lower-level information (i.e. images, numerical values, classification values) from the input. Most contemporary deep learning models are built on artificial neural networks (ANNs), in particular convolutional neural networks (CNNs), which can be integrated with other deep learning models, including generative models, deep belief networks, and Boltzmann machines. Deep learning can be divided into three types: supervised, semi-supervised, and unsupervised. Deep neural networks (DNNs), reinforcement learning, and recurrent neural networks (RNNs) are some of the most prominent DL architectures (RNNs).

Each layer in deep learning learns different data attributes while learning to transform its input data to subsequent layers. For example, in an image recognition application, the original input may be a matrix of pixels and the first layer may detect the edges of the image. The second layer, on the other hand, would construct and encode the nose and eyes, and the third layer might recognize faces by combining all the information gathered from the first two layers.

In the medical field, DL holds great promise. Radiology and pathology are two prominent medical fields that have used deep learning extensively in disease diagnosis for many years. In addition, gathering valuable information from molecular states and determining disease progression or treatment susceptibility are practical uses of DL, which are often recognized by human studies.

Cistanche supplement

The value of the herb Cistanche extract to the kidney:

Medical studies have found that the active ingredients contained in Cistanche extract can promote human cell regeneration and metabolism, enhance immune regulation ability, and have obvious anti-cancer, antiviral, and anti-aging effects. In recent years, the treatment of kidney diseases with Cistanche has achieved good results. Cistanche is a supplement for both men and women and can treat both male impotence and female infertility. Moreover, tonify the kidney without hurting the yin, and long-term use generally does not cause symptoms such as heat and dry mouth.

According to Dr. Xu Jing, Department of Nephrology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, patients with the following diseases should pay special attention:

1. Diabetes： Diabetes will aggravate my glucose load, whether it is type 1 or type 2 diabetes, diabetic nephropathy can occur, and the longer the course of the disease, the greater the risk. Once the end-stage renal disease progresses, it becomes more difficult to control.

2. High blood： pressure High blood pressure and I am nicknamed "hard brother". Not only does high blood pressure put high pressure on me and damage my health, but when I am dysfunctional, it also causes high blood pressure secondarily. The two of us "colluded" together, it was even trickier!

3. Hyperuricemia： With lifestyle changes, the incidence of gout (hyperuricemia) gradually increases. Urate crystals are deposited in the kidneys, accelerating the deterioration of kidney function; The deterioration of renal function will reduce the ability of the kidneys to excrete uric acid, aggravate hyperuricemia again, and form a vicious circle.

4. Bad lifestyle： Long-term sedentary and lack of exercise lead to obesity, overweight, or habitual staying up late, drinking and smoking, etc., will make me work for a long time and cannot rest, thereby increasing my burden and causing damage.

Cistanche tubulosa extract

6. Activate sweating

Whenever we sweat, we lighten the load on our kidneys because we strengthen another way of removing fluid from the body. In this sense, increasing sweating can be a good "therapy" to take care of kidney function. Moderate or high-intensity physical activity, or by intervals. Dry and wet sauna baths. Eating hot spices, such as ginger or cayenne pepper.

7. Regular use of cleansers

Once or twice a year, especially during seasonal changes, it is a good idea to do some cleansing or purification to promote the elimination of toxins and improve liver and kidney function. This includes improving dietary habits, mainly doing kidney cleansing, which you can try:

a: Adding garlic and onion broth to your diet more often.

b: Replace industrial drinks with green pear, celery, and apple juice.

c: Use natural infusions of dandelion, burdock, and horsetail to stimulate the elimination of fluid.

8. Apply local heat

Local heat using a hot water bottle or heating pad can provide relief in case of lower back discomfort in the kidney area. On the other hand, avoiding extreme cold in the trunk area, especially cold winds in winter is essential to prevent (or relieve, as the case may be) discomfort in this area, regardless of the presence of kidney disease.

9. Avoid alcohol and overdose

Tips to take care of the pancreas: do not drink alcohol. Of course, avoiding alcohol is vital to taking care of kidney health. Various studies have shown that alcohol can deteriorate kidney function to a great extent and that the greater the amount of alcohol consumed, the greater the risk of developing various health problems. On the other hand, it is important to always remember that everything in excess is bad for your health, not just alcohol or other drugs.

REFERENCES:

1. Brijain, M.; Patel, R.; Kaushik, M.; Rana, K. A survey on decision tree algorithm for classification. Int. J. Eng. Dev. Res. 2014. Available

2. Walse, R.S.; Kurundkar, G.D.; Khamitkar, S.D.; Muley, A.A.; Bhalchandra, P.U.; Lokhande, S.N. Effective Use of Naïve Bayes, Decision Tree, and Random Forest Techniques for Analysis of Chronic Kidney Disease. In Proceedings of the International Conference on Information and Communication Technology for Intelligent Systems, Ahmedabad, India, 15–16 May 2020; Springer: Berlin/Heidelberg, Germany, 2020; pp. 237–245.

3. Rajendran, K.; Jayabalan, M.; Thiruchelvam, V. Predicting breast cancer via supervised machine learning methods on class imbalanced data. Int. J. Adv. Comput. Sci. Appl. 2020, 11, 54–63.

4. Tsao, H.Y.; Chan, P.Y.; Su, E.C.Y. Predicting diabetic retinopathy and identifying interpretable biomedical features using machine learning algorithms. BMC Bioinform. 2018, 19, 111–121.

5. . Nurrohman, A.; Abdullah, S.; Murfifi, H. Parkinson’s disease subtype classification: Application of decision tree, logistic regression, and logit leaf model. In AIP Conference Proceedings; AIP Publishing LLC: Melville, NY, USA, 2020; Volume 2242, p. 030015.

6. . Drucker, H.; Wu, D.; Vapnik, V.N. Support vector machines for spam categorization. IEEE Trans. Neural Netw. 1999, 10, 1048–1054.

7. Fix, E.; Hodges, J.L. Discriminatory analysis. Nonparametric discrimination: Consistency properties. Int. Stat. Rev. Int. De Stat. 1989, 57, 238–247.

8. Wright, R.E. Logistic regression. In Reading and Understanding Multivariate Statistics; American Psychological Association: Washington, DC, USA, 1995.

9. Schapire, R.E. Explaining AdaBoost. In Empirical Inference; Springer: Berlin/Heidelberg, Germany, 2013; pp. 37–52.

10. Goodfellow, I.; Bengio, Y.; Courville, A. Deep Learning; MIT Press: Cambridge, MA, USA, 2016.

11. Hayashi, Y. The right direction needed to develop white-box deep learning in radiology, pathology, and ophthalmology: A short review. Front. Robot. AI 2019, 6, 24.

12. Akkus, Z.; Galimzianova, A.; Hoogi, A.; Rubin, D.L.; Erickson, B.J. Deep learning for brain MRI segmentation: State of the art and future directions. J. Digit. Imaging 2017, 30, 449–459.

13. Ahsan, M.M.; E Alam, T.; Trafalis, T.; Huebner, P. Deep MLP-CNN model using mixed-data to distinguish between COVID-19 and Non-COVID-19 patients. Symmetry 2020, 12, 1526.

14. Houssein, E.H.; Emam, M.M.; Ali, A.A.; Suganthan, P.N. Deep and machine learning techniques for medical imaging-based breast cancer: A comprehensive review. Expert Syst. Appl. 2021, 167, 114161.