Raw Hematological Dataset for Anemia Analysis and Classification

Description

This dataset comprises 1004 raw hematological records collected from patients at Medical Centre Hospital, Chattogram, for use in anemia analysis and classification. Each row represents one anonymized patient sample, and the dataset contains nine variables: Gender, Age, HGB (Hemoglobin), RBC (Red Blood Cell Count), PCV/HCT (Packed Cell Volume/Hematocrit), MCV (Mean Corpuscular Volume), MCH (Mean Corpuscular Hemoglobin), MCHC (Mean Corpuscular Hemoglobin Concentration), and Decision_Class as the target label for anemia classification. The data are presented in their original, unprocessed form to preserve transparency and allow researchers to perform their own preprocessing, feature engineering, statistical analysis, and machine learning modeling. Because the dataset is openly structured and clinically relevant, it can support applications in anemia detection, supervised machine learning, deep learning development, biomedical data analysis, and educational research. All records are de-identified and contain no personally identifiable information, supporting ethical use and open-data dissemination. The dataset was validated by Hasanul Alam Shaikat, a biochemist, who reviewed and verified the clinical relevance and consistency of the hematological measurements and class annotations before release.

Steps to reproduce

Data Collection: The dataset was collected from Medical Centre Hospital, Chattogram. It contains raw hematological data from patients, including key variables for anemia analysis and classification, such as Hemoglobin (HGB), Red Blood Cell Count (RBC), Packed Cell Volume (PCV), Mean Corpuscular Volume (MCV), Mean Corpuscular Hemoglobin (MCH), Mean Corpuscular Hemoglobin Concentration (MCHC), and a Decision_Class label indicating anemia classification. The data is raw and unprocessed, meaning it hasn't undergone any transformations, normalization, or imputation of missing values. Researchers are encouraged to apply their own preprocessing steps based on their analysis needs. Data Availability: The dataset is provided in Excel format (e.g., Anemia_Dataset.SHAIKAT (1).xlsx). The dataset includes anonymized data with no personally identifiable information. Researchers can directly use the provided raw file for analysis. It is available for download and use in its original form, facilitating transparency and custom preprocessing. Software Requirements (optional): To load and analyze the dataset, you can use software like Microsoft Excel, or you can use a data science environment like Python (using libraries such as pandas) or R for more advanced analysis or machine learning tasks. Example Python code to load the raw dataset: import pandas as pd dataset = pd.read_excel('Anemia_Dataset.SHAIKAT (1).xlsx') Analysis (Optional, if relevant): Since the dataset is raw, researchers may choose to preprocess the data (e.g., handling missing values, normalization) before applying machine learning models for classification tasks. Popular algorithms like Decision Trees, Random Forest, SVM, and others can be used for predicting Decision_Class (Anemia vs. Normal) based on the provided hematological features.

Institutions

• Daffodil International University

  Dhaka Division, Dhaka

Categories

Licence