Detection of Laryngeal Cancer using audio processing.
- 1. SANJIVANI UNIVERSITY, KOPARGAON DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING SEMINAR-1 A Project Presentation on “Detection of Laryngeal Cancer Using Audio Processing” Group Member : 1. Adhav Apurva Kailas 2. Dange Rucha Sachin 3. Taji Vedika Sachin 4. Wagh Aishwarya Rajendra Project Investigator: Prof. Chanda Pathak Project Guide: Prof. Tanay Ghosh
- 2. • Introduction • Problem Statement • Objectives • Research Paper • Proposed System • Scope • Future Work • Primary Search • Secondary Search • Implications of Early Diagnosis • Pros & Cons • Correlation between Voice Changes & Laryngeal Cancer • Conclusion CONTENT
- 3. INTRODUCTION • Laryngeal cancer is developed in the tissues of the larynx (voice box), located in the throat. • The project seeks to develop a sophisticated system capable of analyzing audio recordings of a patient's voice. • Using audio processing, laryngeal cancer can be detected by analyzing voice changes such as hoarseness, pitch alteration, or vocal fatigue.
- 4. PROBLEM STATEMENT • The project aims to develop an innovative system for early detection of laryngeal cancer by analyzing voice recordings using audio processing and machine learning
- 5. v OBJECTIVES • Detect early laryngeal cancer from voice changes • Analyse vocal abnormalities • Extract relevant audio features. • Use machine learning for classification. • Provide a non-invasive screening tool.
- 6. LITERATURE SURVEY 1. “Effective Glottal Instant Detection and Electroglottographic Parameter Extraction for Automated Voice Pathology Assessment” Author - Pranav s. Deshpande, M.Sabarimalai Manikandan Published date - 2018 March Methodology - • The paper discusses the impact of voice disorders and the limitations of traditional assessment methods. • It introduces electroglottography (EGG) as a non-invasive technique and emphasizes the need for automated systems to improve accuracy in diagnosing voice pathologies.
- 7. LITERATURE SURVEY 2. “Analysis and Detection of Pathological Voice using Glottal Source Features” Author - Sudarsana Reddy Kadiri, Paavo Alku Published date -10 December 2019 Methodology - • Author analyse the major glottal source features that mostly beneficial for voice pathology detection in a more effective manner. • Mel frequency ceptrum coefficients (MFCCs) are extracted from the voice signal. • Glottal inverse filtering method and the zero frequency filtering (ZFF) method captures the variations in glottal waveforms to detect pathological voice.
- 8. LITERATURE SURVEY 3. “Non –Invasive Detection of Potentially Precancerous Lesions of Vocal Fold Based on Glottal Wave Signal and SVM Approaches” Author - Anis Ben AICHA Published date - 12 January 2018 Methodology - • The non invasive process used for experiments that get an accuracy for 92% the source signal is first extracted namely the glottal flow signal from the acoustic speech using the IAFF technique. • SVM module is used to classify and discriminate pre malignant lessons from normal voices.
- 9. LITERATURE SURVEY 4. “Investigation and Evaluation of Glottal Flow Waveform for Voice Pathology Detection” Author - YUANBO WU , CHANGWEL ZHOU , ZIQI FAN , DI WU Published by - January 04, 2021 Methodology - • Author detects voice by evaluating the classification result using features extracted from raw speech utterances add corresponding glottal flow waveforms. • The system provides 88.52% accuracy using random forest for pathological voice detection.
- 10. PROPOSED SYSTEM
- 11. SCOPE • Early Detection: Identifying cancer at an early stage through vocal analysis. • Non-Invasive Screening: Providing a simple, non-invasive alternative for diagnosis. • Cost-Effective Tool: Reducing healthcare costs with affordable screening methods. • Clinical Use: Potential for integration into standard medical practices with further validation.
- 12. FUTURE WORK • It involves expanding the dataset to include more patients and improving model optimization. • Real-time processing capabilities will be developed for point-of-care applications, and large- scale clinical trials will be conducted to validate accuracy and reliability. • The project will pursue regulatory approval for clinical use and integrate with telemedicine platforms.
- 13. IMPLICATIONS OF EARLY DIAGNOSIS • Increased Survival Rates: Higher survival rates due to early detection of cancer. • Voice Preservation: Early diagnosis helps to preserve vocal functions, reducing the risk of permanent voice damage or loss. • Cost-Effectiveness: Implementing audio processing technologies for early diagnosis can reduce healthcare costs by minimizing the need for extensive diagnostic procedures
- 14. PROS & CONS 1. Improves treatment success 2. Enhances survival rates 3. Reduces healthcare costs 4. Increases patient awareness 1. Potential for false positives 2. Requires access to advanced technology 3. Can be resource-intensive 4. Depends on patient compliance Pros : Cons :
- 15. Correlation Between Voice Changes & Laryngeal Cancer • Voice Quality Indicators : Changes in voice quality, such as hoarseness or breathness, can serve as early warning signs of laryngeal cancer, highlighting the vocal changes for timely diagnosis and intervention. • Acoustic Analysis Techniques : Advanced audio processing methods, including spectral analysis and machine learning, can quantitatively assess voice changes. • Clinical Implications : It can lead to improved screening protocols, enabling healthcare professionals to identify at-risk patients and initiate early diagnostic procedures.
- 16. CONCLUSION In conclusion, using audio processing for the detection of laryngeal cancer offers a promising, non-invasive, and cost-effective method for early diagnosis. By analysing vocal changes, it enables timely intervention, increases survival rates, and preserves vocal functions, making it a valuable tool for enhancing patient outcomes.
- 17. 1. Anis Ben Aicha "Noninvasive Detection of Potentially Precancerous Lesions of Vocal Fold Based on Glottal Wave Signal and SVM Approaches “, International Conference on Knowledge Based and Intelligent Information and Engineering Systems, KES2018, 3-5 September 2018, Belgrade, Serbia. 2. Pranav S. Deshpande; M.Sabarimalai Manikandan ,"Effective Glottal Instant Detection and Electroglottographic Parameter Extraction for Automated Voice Pathology Assessment", IEEE Journal of Biomedical and Health Informatics ( Volume: 22 , Issue: 2, March 2018. 3. Korutla Sudhir Sai, Polasi Phani Kumar, "Glottal Analysis Using Speech signals “International Journal for Research in Applied Science & Engineering Technology (IJRASET),2017. 4. Sudarsana Reddy Kadiri; Paavo Alku ,“Analysis and Detection of Pathological Voice Using Glottal Source Features”, IEEE Journal of Selected Topics in Signal Processing ( Volume: 14, Issue: 2, Feb. 2020). REFERENCE
- 18. THANK YOU ! v