In the captivating journey of our data science workshop, we embarked on the exploration of Chronic Kidney Disease classification and prediction. Our quest began with a thorough dive into data exploration, where we meticulously delved into the dataset's intricacies to unearth hidden patterns and insights. We analyzed the distribution of categorized features, unraveling the nuances that underlie chronic kidney disease. Guided by the principles of machine learning, we embarked on the quest to build predictive models. With the aid of grid search, we fine-tuned our machine learning algorithms, optimizing their hyperparameters for peak performance. Each model, whether K-Nearest Neighbors, Decision Trees, Random Forests, Gradient Boosting, Naive Bayes, Extreme Gradient Boosting, Light Gradient Boosting, or Multi-Layer Perceptron, was meticulously trained and tested, paving the way for robust predictions. The voyage into the realm of deep learning took us further, as we harnessed the power of Artificial Neural Networks (ANNs). By constructing intricate architectures, we designed ANNs to discern intricate patterns from the data. Leveraging the prowess of TensorFlow, we artfully crafted layers, each contributing to the ANN's comprehension of the underlying dynamics. This marked our initial foray into the world of deep learning. Our expedition, however, did not conclude with ANNs. We ventured deeper into the abyss of deep learning, uncovering the potential of Long Short-Term Memory (LSTM) networks. These networks, attuned to sequential data, unraveled temporal dependencies within the dataset, fortifying our predictive capabilities. Diving even further, we encountered Self-Organizing Maps (SOMs) and Restricted Boltzmann Machines (RBMs). These innovative models, rooted in unsupervised learning, unmasked underlying structures in the dataset. As our understanding of the data deepened, so did our repertoire of tools for prediction. Autoencoders, our final frontier in deep learning, emerged as our champions in dimensionality reduction and feature learning. These unsupervised neural networks transformed complex data into compact, meaningful representations, guiding our predictive models with newfound efficiency. To furnish a granular understanding of model behavior, we employed the classification report, which delineated precision, recall, and F1-Score for each class, providing a comprehensive snapshot of the model's predictive capacity across diverse categories. The confusion matrix emerged as a tangible visualization, detailing the interplay between true positives, true negatives, false positives, and false negatives. We also harnessed ROC and precision-recall curves to illuminate the dynamic interplay between true positive rate and false positive rate, vital when tackling imbalanced datasets. For regression tasks, MSE and its counterpart RMSE quantified the average squared differences between predictions and actual values, facilitating an insightful assessment of model fit. Further enhancing our toolkit, the R-squared (R2) score unveiled the extent to which the model explained variance in the dependent variable, offering a valuable gauge of overall performance. Collectively, this ensemble of metrics enabled us to make astute model decisions, optimize hyperparameters, and gauge the models' fitness for accurate disease prognosis in a clinical context. Amidst this whirlwind of data exploration and model construction, our GUI using PyQt emerged as a beacon of user-friendly interaction. Through its intuitive interface, users navigated seamlessly between model selection, training, and prediction. Our GUI encapsulated the intricacies of our journey, bridging the gap between data science and user experience. In the end, our odyssey illuminated the intricate landscape of Chronic Kidney Disease classification and prediction. We harnessed the power of both machine learning and deep learning, uncovering hidden insights and propelling our predictive capabilities to new heights. Our journey transcended the realms of data, algorithms, and interfaces, leaving an indelible mark on the crossroads of science and innovation.
Workshop 1: Heart Failure Analysis and Prediction Using Scikit-Learn, Keras, and TensorFlow with Python GUI Cardiovascular diseases (CVDs) are the number 1 cause of death globally taking an estimated 17.9 million lives each year, which accounts for 31% of all deaths worldwide. Heart failure is a common event caused by CVDs and this dataset contains 12 features that can be used to predict mortality by heart failure. People with cardiovascular disease or who are at high cardiovascular risk (due to the presence of one or more risk factors such as hypertension, diabetes, hyperlipidaemia or already established disease) need early detection and management wherein a machine learning models can be of great help. Dataset used in this project is from Davide Chicco, Giuseppe Jurman. Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone. BMC Medical Informatics and Decision Making 20, 16 (2020). Attribute information in the dataset are as follows: age: Age; anaemia: Decrease of red blood cells or hemoglobin (boolean); creatinine_phosphokinase: Level of the CPK enzyme in the blood (mcg/L); diabetes: If the patient has diabetes (boolean); ejection_fraction: Percentage of blood leaving the heart at each contraction (percentage); high_blood_pressure: If the patient has hypertension (boolean); platelets: Platelets in the blood (kiloplatelets/mL); serum_creatinine: Level of serum creatinine in the blood (mg/dL); serum_sodium: Level of serum sodium in the blood (mEq/L); sex: Woman or man (binary); smoking: If the patient smokes or not (boolean); time: Follow-up period (days); and DEATH_EVENT: If the patient deceased during the follow-up period (boolean). The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performace of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 2: Cervical Cancer Classification and Prediction Using Machine Learning and Deep Learning with Python GUI About 11,000 new cases of invasive cervical cancer are diagnosed each year in the U.S. However, the number of new cervical cancer cases has been declining steadily over the past decades. Although it is the most preventable type of cancer, each year cervical cancer kills about 4,000 women in the U.S. and about 300,000 women worldwide. Numerous studies report that high poverty levels are linked with low screening rates. In addition, lack of health insurance, limited transportation, and language difficulties hinder a poor woman’s access to screening services. Human papilloma virus (HPV) is the main risk factor for cervical cancer. In adults, the most important risk factor for HPV is sexual activity with an infected person. Women most at risk for cervical cancer are those with a history of multiple sexual partners, sexual intercourse at age 17 years or younger, or both. A woman who has never been sexually active has a very low risk for developing cervical cancer. Sexual activity with multiple partners increases the likelihood of many other sexually transmitted infections (chlamydia, gonorrhea, syphilis). Studies have found an association between chlamydia and cervical cancer risk, including the possibility that chlamydia may prolong HPV infection. Therefore, early detection of cervical cancer using machine and deep learning models can be of great help. The dataset used in this project is obtained from UCI Repository and kindly acknowledged. This file contains a List of Risk Factors for Cervical Cancer leading to a Biopsy Examination. The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performace of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 3: Chronic Kidney Disease Classification and Prediction Using Machine Learning and Deep Learning with Python GUI Chronic kidney disease is the longstanding disease of the kidneys leading to renal failure. The kidneys filter waste and excess fluid from the blood. As kidneys fail, waste builds up. Symptoms develop slowly and aren't specific to the disease. Some people have no symptoms at all and are diagnosed by a lab test. Medication helps manage symptoms. In later stages, filtering the blood with a machine (dialysis) or a transplant may be required The dataset used in this project was taken over a 2-month period in India with 25 features (eg, red blood cell count, white blood cell count, etc). The target is the 'classification', which is either 'ckd' or 'notckd' - ckd=chronic kidney disease. It contains measures of 24 features for 400 people. Quite a lot of features for just 400 samples. There are 14 categorical features, while 10 are numerical. The dataset needs cleaning: in that it has NaNs and the numeric features need to be forced to floats. Attribute Information: Age(numerical) age in years; Blood Pressure(numerical) bp in mm/Hg; Specific Gravity(categorical) sg - (1.005,1.010,1.015,1.020,1.025); Albumin(categorical) al - (0,1,2,3,4,5); Sugar(categorical) su - (0,1,2,3,4,5); Red Blood Cells(categorical) rbc - (normal,abnormal); Pus Cell (categorical) pc - (normal,abnormal); Pus Cell clumps(categorical) pcc - (present, notpresent); Bacteria(categorical) ba - (present,notpresent); Blood Glucose Random(numerical) bgr in mgs/dl; Blood Urea(numerical) bu in mgs/dl; Serum Creatinine(numerical) sc in mgs/dl; Sodium(numerical) sod in mEq/L; Potassium(numerical) pot in mEq/L; Hemoglobin(numerical) hemo in gms; Packed Cell Volume(numerical); White Blood Cell Count(numerical) wc in cells/cumm; Red Blood Cell Count(numerical) rc in millions/cmm; Hypertension(categorical) htn - (yes,no); Diabetes Mellitus(categorical) dm - (yes,no); Coronary Artery Disease(categorical) cad - (yes,no); Appetite(categorical) appet - (good,poor); Pedal Edema(categorical) pe - (yes,no); Anemia(categorical) ane - (yes,no); and Class (categorical) class - (ckd,notckd). The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performace of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 4: Lung Cancer Classification and Prediction Using Machine Learning and Deep Learning with Python GUI The effectiveness of cancer prediction system helps the people to know their cancer risk with low cost and it also helps the people to take the appropriate decision based on their cancer risk status. The data is collected from the website online lung cancer prediction system. Total number of attributes in the dataset is 16, while number of instances is 309. Following are attribute information of dataset: Gender: M(male), F(female); Age: Age of the patient; Smoking: YES=2 , NO=1; Yellow fingers: YES=2 , NO=1; Anxiety: YES=2 , NO=1; Peer_pressure: YES=2 , NO=1; Chronic Disease: YES=2 , NO=1; Fatigue: YES=2 , NO=1; Allergy: YES=2 , NO=1; Wheezing: YES=2 , NO=1; Alcohol: YES=2 , NO=1; Coughing: YES=2 , NO=1; Shortness of Breath: YES=2 , NO=1; Swallowing Difficulty: YES=2 , NO=1; Chest pain: YES=2 , NO=1; and Lung Cancer: YES , NO. The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performace of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 5: Alzheimer’s Disease Classification and Prediction Using Machine Learning and Deep Learning with Python GUI Alzheimer's is a type of dementia that causes problems with memory, thinking and behavior. Symptoms usually develop slowly and get worse over time, becoming severe enough to interfere with daily tasks. Alzheimer's is not a normal part of aging. The greatest known risk factor is increasing age, and the majority of people with Alzheimer's are 65 and older. But Alzheimer's is not just a disease of old age. Approximately 200,000 Americans under the age of 65 have younger-onset Alzheimer’s disease (also known as early-onset Alzheimer’s). The dataset consists of a longitudinal MRI data of 374 subjects aged 60 to 96. Each subject was scanned at least once. Everyone is right-handed. 206 of the subjects were grouped as 'Nondemented' throughout the study. 107 of the subjects were grouped as 'Demented' at the time of their initial visits and remained so throughout the study. 14 subjects were grouped as 'Nondemented' at the time of their initial visit and were subsequently characterized as 'Demented' at a later visit. These fall under the 'Converted' category. Following are some important features in the dataset: EDUC:Years of Education; SES: Socioeconomic Status; MMSE: Mini Mental State Examination; CDR: Clinical Dementia Rating; eTIV: Estimated Total Intracranial Volume; nWBV: Normalize Whole Brain Volume; and ASF: Atlas Scaling Factor. The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performance of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 6: Parkinson Classification and Prediction Using Machine Learning and Deep Learning with Python GUI The dataset was created by Max Little of the University of Oxford, in collaboration with the National Centre for Voice and Speech, Denver, Colorado, who recorded the speech signals. The original study published the feature extraction methods for general voice disorders. This dataset is composed of a range of biomedical voice measurements from 31 people, 23 with Parkinson's disease (PD). Each column in the table is a particular voice measure, and each row corresponds one of 195 voice recording from these individuals ("name" column). The main aim of the data is to discriminate healthy people from those with PD, according to "status" column which is set to 0 for healthy and 1 for PD. The data is in ASCII CSV format. The rows of the CSV file contain an instance corresponding to one voice recording. There are around six recordings per patient, the name of the patient is identified in the first column. Attribute information of this dataset are as follows: name - ASCII subject name and recording number; MDVP:Fo(Hz) - Average vocal fundamental frequency; MDVP:Fhi(Hz) - Maximum vocal fundamental frequency; MDVP:Flo(Hz) - Minimum vocal fundamental frequency; MDVP:Jitter(%); MDVP:Jitter(Abs); MDVP:RAP; MDVP:PPQ; Jitter:DDP – Several measures of variation in fundamental frequency; MDVP:Shimmer; MDVP:Shimmer(dB); Shimmer:APQ3; Shimmer:APQ5; MDVP:APQ; Shimmer:DDA - Several measures of variation in amplitude; NHR; HNR - Two measures of ratio of noise to tonal components in the voice; status - Health status of the subject (one) - Parkinson's, (zero) – healthy; RPDE,D2 - Two nonlinear dynamical complexity measures; DFA - Signal fractal scaling exponent; and spread1,spread2,PPE - Three nonlinear measures of fundamental frequency variation. The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performance of the model, scalability of the model, training loss, and training accuracy. WORKSHOP 7: Liver Disease Classification and Prediction Using Machine Learning and Deep Learning with Python GUI Patients with Liver disease have been continuously increasing because of excessive consumption of alcohol, inhale of harmful gases, intake of contaminated food, pickles and drugs. This dataset was used to evaluate prediction algorithms in an effort to reduce burden on doctors. This dataset contains 416 liver patient records and 167 non liver patient records collected from North East of Andhra Pradesh, India. The "Dataset" column is a class label used to divide groups into liver patient (liver disease) or not (no disease). This data set contains 441 male patient records and 142 female patient records. Any patient whose age exceeded 89 is listed as being of age "90". Columns in the dataset: Age of the patient; Gender of the patient; Total Bilirubin; Direct Bilirubin; Alkaline Phosphotase; Alamine Aminotransferase; Aspartate Aminotransferase; Total Protiens; Albumin; Albumin and Globulin Ratio; and Dataset: field used to split the data into two sets (patient with liver disease, or no disease). The models used in this project are K-Nearest Neighbor, Random Forest, Naive Bayes, Logistic Regression, Decision Tree, Support Vector Machine, Adaboost, LGBM classifier, Gradient Boosting, XGB classifier, MLP classifier, and CNN 1D. Finally, you will develop a GUI using PyQt5 to plot boundary decision, ROC, distribution of features, feature importance, cross validation score, and predicted values versus true values, confusion matrix, learning curve, performance of the model, scalability of the model, training loss, and training accuracy.
Build a solid foundation in surgical AI with this engaging, comprehensive guide for AI novices Machine learning, neural networks, and computer vision in surgical education, practice, and research will soon be de rigueur. Written for surgeons without a background in math or computer science, Artificial Intelligence in Surgery provides everything you need to evaluate new technologies and make the right decisions about bringing AI into your practice. Comprehensive and easy to understand, this first-of-its-kind resource illustrates the use of AI in surgery through real-life examples. It covers the issues most relevant to your practice, including: Neural Networks and Deep Learning Natural Language Processing Computer Vision Surgical Education and Simulation Preoperative Risk Stratification Intraoperative Video Analysis OR Black Box and Tracking of Intraoperative Events Artificial Intelligence and Robotic Surgery Natural Language Processing for Clinical Documentation Leveraging Artificial Intelligence in the EMR Ethical Implications of Artificial Intelligence in Surgery Artificial Intelligence and Health Policy Assessing Strengths and Weaknesses of Artificial Intelligence Research Finally, the appendix includes a detailed glossary of terms and important learning resources and techniques―all of which helps you interpret claims made by studies or companies using AI.
This book provides a collection of comprehensive research articles on data analytics and applications of wearable devices in healthcare. This Special Issue presents 28 research studies from 137 authors representing 37 institutions from 19 countries. To facilitate the understanding of the research articles, we have organized the book to show various aspects covered in this field, such as eHealth, technology-integrated research, prediction models, rehabilitation studies, prototype systems, community health studies, ergonomics design systems, technology acceptance model evaluation studies, telemonitoring systems, warning systems, application of sensors in sports studies, clinical systems, feasibility studies, geographical location based systems, tracking systems, observational studies, risk assessment studies, human activity recognition systems, impact measurement systems, and a systematic review. We would like to take this opportunity to invite high quality research articles for our next Special Issue entitled “Digital Health and Smart Sensors for Better Management of Cancer and Chronic Diseases” as a part of Sensors journal.
This book presents an overview of how machine learning and data mining techniques are used for tracking and preventing diseases. It covers several aspects such as stress level identification of a person from his/her speech, automatic diagnosis of disease from X-ray images, intelligent diagnosis of Glaucoma from clinical eye examination data, prediction of protein-coding genes from big genome data, disease detection through microscopic analysis of blood cells, information retrieval from electronic medical record using named entity recognition approaches, and prediction of drug-target interactions. The book is suitable for computer scientists having a bachelor degree in computer science. The book is an ideal resource as a reference book for teaching a graduate course on AI for Medicine or AI for Health care. Researchers working in the multidisciplinary areas use this book to discover the current developments. Besides its use in academia, this book provides enough details about the state-of-the-art algorithms addressing various biomedical domains, so that it could be used by industry practitioners who want to implement AI techniques to analyze the diseases. Medical institutions use this book as reference material and give tutorials to medical experts on how the advanced AI and ML techniques contribute to the diagnosis and prediction of the diseases.
In this project, Data Science Workshop focused on Liver Disease Classification and Prediction, we embarked on a comprehensive journey through various stages of data analysis, model development, and performance evaluation. The workshop aimed to utilize Python and its associated libraries to create a Graphical User Interface (GUI) that facilitates the classification and prediction of liver disease cases. Our exploration began with a thorough examination of the dataset. This entailed importing necessary libraries such as NumPy, Pandas, and Matplotlib for data manipulation, visualization, and preprocessing. The dataset, representing liver-related attributes, was read and its dimensions were checked to ensure data integrity. To gain a preliminary understanding, the dataset's initial rows and column information were displayed. We identified key features such as 'Age', 'Gender', and various biochemical attributes relevant to liver health. The dataset's structure, including data types and non-null counts, was inspected to identify any potential data quality issues. We detected that the 'Albumin_and_Globulin_Ratio' feature had a few missing values, which were subsequently filled with the median value. Our exploration extended to visualizing categorical distributions. Pie charts provided insights into the proportions of healthy and unhealthy liver cases among different gender categories. Stacked bar plots further delved into the connections between 'Total_Bilirubin' categories and the prevalence of liver disease, fostering a deeper understanding of these relationships. Transitioning to predictive modeling, we embarked on constructing machine learning models. Our arsenal included a range of algorithms such as Logistic Regression, Support Vector Machines, K-Nearest Neighbors, Decision Trees, Random Forests, Gradient Boosting, Extreme Gradient Boosting, Light Gradient Boosting. The data was split into training and testing sets, and each model underwent rigorous evaluation using metrics like accuracy, precision, recall, F1-score, and ROC-AUC. Hyperparameter tuning played a pivotal role in model enhancement. We leveraged grid search and cross-validation techniques to identify the best combination of hyperparameters, optimizing model performance. Our focus shifted towards assessing the significance of each feature, using techniques such as feature importance from tree-based models. The workshop didn't halt at machine learning; it delved into deep learning as well. We implemented an Artificial Neural Network (ANN) using the Keras library. This powerful model demonstrated its ability to capture complex relationships within the data. With distinct layers, activation functions, and dropout layers to prevent overfitting, the ANN achieved impressive results in liver disease prediction. Our journey culminated with a comprehensive analysis of model performance. The metrics chosen for evaluation included accuracy, precision, recall, F1-score, and confusion matrix visualizations. These metrics provided a comprehensive view of the model's capability to correctly classify both healthy and unhealthy liver cases. In summary, the Data Science Workshop on Liver Disease Classification and Prediction was a holistic exploration into data preprocessing, feature categorization, machine learning, and deep learning techniques. The culmination of these efforts resulted in the creation of a Python GUI that empowers users to input patient attributes and receive predictions regarding liver health. Through this workshop, participants gained a well-rounded understanding of data science techniques and their application in the field of healthcare.
This book presents a detailed review of the state of the art in deep learning approaches for semantic object detection and segmentation in medical image computing, and large-scale radiology database mining. A particular focus is placed on the application of convolutional neural networks, with the theory supported by practical examples. Features: highlights how the use of deep neural networks can address new questions and protocols, as well as improve upon existing challenges in medical image computing; discusses the insightful research experience of Dr. Ronald M. Summers; presents a comprehensive review of the latest research and literature; describes a range of different methods that make use of deep learning for object or landmark detection tasks in 2D and 3D medical imaging; examines a varied selection of techniques for semantic segmentation using deep learning principles in medical imaging; introduces a novel approach to interleaved text and image deep mining on a large-scale radiology image database.
This book discusses new cognitive informatics tools, algorithms and methods that mimic the mechanisms of the human brain which lead to an impending revolution in understating a large amount of data generated by various smart applications. The book is a collection of peer-reviewed best selected research papers presented at the International Conference on Data Intelligence and Cognitive Informatics (ICDICI 2020), organized by SCAD College of Engineering and Technology, Tirunelveli, India, during 8–9 July 2020. The book includes novel work in data intelligence domain which combines with the increasing efforts of artificial intelligence, machine learning, deep learning and cognitive science to study and develop a deeper understanding of the information processing systems.
This book is about making machine learning models and their decisions interpretable. After exploring the concepts of interpretability, you will learn about simple, interpretable models such as decision trees, decision rules and linear regression. Later chapters focus on general model-agnostic methods for interpreting black box models like feature importance and accumulated local effects and explaining individual predictions with Shapley values and LIME. All interpretation methods are explained in depth and discussed critically. How do they work under the hood? What are their strengths and weaknesses? How can their outputs be interpreted? This book will enable you to select and correctly apply the interpretation method that is most suitable for your machine learning project.
This unique book introduces a variety of techniques designed to represent, enhance and empower multi-disciplinary and multi-institutional machine learning research in healthcare informatics. Providing a unique compendium of current and emerging machine learning paradigms for healthcare informatics, it reflects the diversity, complexity, and the depth and breadth of this multi-disciplinary area. Further, it describes techniques for applying machine learning within organizations and explains how to evaluate the efficacy, suitability, and efficiency of such applications. Featuring illustrative case studies, including how chronic disease is being redefined through patient-led data learning, the book offers a guided tour of machine learning algorithms, architecture design, and applications of learning in healthcare challenges.
