Competitors will be competing to create a disease classification model to predict respiratory disease using audio files and other features. This challenge focuses on using Medical AI and Sound Recognition to build models for identifying conditions like asthma and COPD. Participants will utilize audio files and patient metadata to tackle a critical healthcare AI problem. By solving this, you'll contribute to advancements in machine learning for respiratory disease detection.

In this competition, we challenge you to push the boundaries of machine learning in healthcare. With a small but crucial dataset, the task is to predict medical outcomes with high accuracy. Can you devise innovative solutions to tackle limited data in this high-stakes domain?

### Background
diseases such as asthma, COPD (Chronic Obstructive Pulmonary Disease), and tuberculosis affect millions worldwide and are leading causes of mortality and morbidity, especially in low- and middle-income countries (LMICs).

Impact of Respiratory Diseases:
**Asthma**: Asthma affects an estimated 262 million people globally. In 2019, asthma contributed to 21.6 million disability-adjusted life years (DALYs) and was ranked 34th among the leading causes of disease burden globally. Severe asthma impacts the quality of life for both children and adults, with symptoms including wheezing, breathlessness, and frequent hospitalizations​

**COPD**: As the third leading cause of death worldwide, COPD was responsible for 3.23 million deaths in 2019. COPD often results from long-term exposure to harmful pollutants like tobacco smoke and household air pollution. Nearly 90% of COPD deaths occur in LMICs, highlighting the significant health disparities globally​
WORLD HEALTH ORGANIZATION (WHO)

**Tuberculosis (TB)**: TB remains a significant cause of death globally, with approximately 1.5 million deaths annually. Although primarily affecting the lungs, TB and post-tuberculosis sequelae can lead to chronic respiratory issues like COPD, further increasing the burden on healthcare systems, especially in LMICs.

By participating in this competition, you contribute to developing AI solutions that can improve early diagnosis and disease management in healthcare settings. Early and accurate detection of these diseases using machine learning on audio recordings could revolutionize healthcare, especially in resource-constrained environments.

The information provided in this competition include these sounds, vector embeddings for these sounds created using a self supervised model and information of the person. The names of the diseases and information of people are not revealed for confidentiality and ethical considerations to prevent misuse of diagnosis in case of poor detection. 

**Goal:** To differentiate diseased individuals from healthy and chronic respiratory diseases of individuals using respiratory sounds.