Presentation Musli Yanto (Automatic Detection of Ground Glass Opacity (GGO) Objects in CT Scan Images Tuberculosis).pptx

Presentation Musli Yanto (Automatic Detection of Ground Glass Opacity (GGO) Objects in CT Scan Images Tuberculosis
Presentation Musli Yanto (Automatic Detection of Ground Glass Opacity (GGO) Objects in CT Scan Images Tuberculosis

1. Presentation Result Research Automatic Detection of Ground Glass Opacity (GGO) Objects in CT Scan Images Tuberculosis

2. List of contents Introduction Result and Discussion Research Methods Conclusion

3. Introduction Tuberculosis is a fairly common disease in the medical world ( Banapuram et al., 2024). TB is often caused by bacteria that enter the lungs (Sharma et al., 2024). TB is also considered an infectious disease that can affect other parts of the body (Vats et al., 2024). This disease also causes significant damage, resulting in suboptimal lung function (Hattori et al., 2024). The persistence of GGO also follows a similar path to tuberculosis (Herath et al., 2022). GGO can be identified using medical imaging, which shows a dense patch in the lung area (Herskovitz et al., 2022). The role of digital images has provided a significant boost in the efficiency of object detection processes, such as GGO (Wu et al., 2021). The role of digital images in the medical world has also significantly contributed to maximizing the diagnostic process (Y. Li et al., 2021).

4. State of the Art The implementation of GGO object detection cases using segmentation techniques also significant results, with accuracy values of 80.69%, specificity of 99.69%, and sensitivity of 83.54% ( Enshaei et al., 2022). Previous research has also confirmed that detection models using segmentation techniques yield a detection precision of 90.23% (Faruk, 2021). The application of segmentation techniques, also focused on TB object detection, has been confirmed to result an accuracy of 97% based on Deep Learning performance (Wang & Liang, 2024). Subsequent research also reported that the application of segmentation techniques using Region of Interest (ROI) to automatic detection results with 75% accuracy (Ejaz et al., 2024). The same study also noted that automatic detection using CNN-LSTM-based DL implementation also for result to improved detection results compared to previous models ( Kotei & Thirunavukarasu , 2024).

5. Research Framework Input Image CT-Scan Image Preprocessing Image Segmentation Object Segmentation Performance Measurement

6. Novelty Research Interpolation Linear Morphology Adaptive (ILMA) Algorithm The ILMA algorithm's role can be supported by the performance of the eAGT function in generating threshold parameter values for threshold-based segmentation and the linear interpolation method in improving the segmentation results. The ILMA algorithm's performance in the automatic detection process will be carried out using a dataset of CT scan images from TB.

7. Image Dataset A total of 134 CT scan images were used to test the performance of the ILMA algorithm in carrying out the automatic detection process. The 134 images in the divided into 117 training sets and 17 testing sets

8. Preprocessing Result The preprocessing output, which is used as the initial step in the detection process. The filtering image results are the preprocessing results used in object detection using the segmentation technique developed in the ILMA algorithm. Grayscale Transformation Image Adjustment Filtering Result

9. Detection Ilma Algorithm Result the results ILMA algorithm in detecting GGO objects. The ILMA algorithm's performance can to provide precise and accurate detection results in GGO objects. Image Input Detection Result

10. Validation of ILMA Algorithm Detection Results Performance of the ILMA algorithm in automatically detecting GGO objects. Based on the measurement results, it can be proven that the ILMA algorithm provides quite significant detection results with an accuracy rate of 97.76%, Specificity of 99.15%, and Sensitivity of 93.33%. Segmentation Result Detected Not detected Positive 117 1 Negative 14 2 Total 131 3 Accuracy 97.76% Specificity 99.15% Sensitivity 93.33%

11. Review of Previous Research These results demonstrate that the automated detection process based on the ILMA algorithm is capable of delivering precise and accurate detection results. The ILMA algorithm's performance has also provided novel object segmentation techniques for the development of medical technology for diagnostic No Previous Research Results Results of work performed 1 Detection process presents a 97% accuracy rate with artificial intelligence-based Automatic Detection (Wang & Liang, 2024) The performance of the ILMA Algorithm in the detection process provides an accuracy rate of 97.76%, Specificity of 99.15%, and Sensitivity of 93.33% 2 The Region of Interest (ROI) based segmentation method combined with the ANN method is able to provide a detection output of 75% (Ejaz et al., 2024) 3 Development of segmentation techniques in automatic detection provides an output accuracy level of 95% (Duong et al., 2023).

12. Conclusion the ILMA algorithm has been proven to provide a detection accuracy of 97.76%. This result is based on the performance of the Extended Adaptive Global Threshold ( eAGT ) function and the Linear Interpolation method in optimizing segmentation performance. The segmentation results presented based on previous testing can see the detection object sufficiently to provide a precise and accurate description of the detection results.

13. Reference C. Banapuram , A. C. Naik, M. K. Vanteru , V. S. Kumar, and K. K. Vaigandla , “A Comprehensive Survey of Machine Learning in Healthcare: Predicting Heart and Liver Disease, Tuberculosis Detection in Chest X-Ray Images,” SSRG Int. J. Electron. Commun. Eng., vol. 11, no. 5, pp. 155–169, 2024, https://doi.org/10.14445/23488549/IJECE-V11I5P116 . V. Sharma, S. K. Gupta, and K. K. Shukla, “Deep learning models for tuberculosis detection and infected region visualization in chest X-ray images,” Intell . Med., vol. 4, no. 2, pp. 104–113, 2024, https://doi.org/10.1016/j.imed.2023.06.001 . D. Li, C. Deng, S. Wang, Y. Li, Y. Zhang, and H. Chen, “Ten-year follow-up results of pure ground-glass opacity–featured lung adenocarcinomas after surgery,” Ann. Thorac . Surg., vol. 116, no. 2, pp. 230–237, 2023, https://doi.org/10.1016/j.athoracsur.2023.01.014 H. Herath, G. Karunasena , and B. Madhusanka , “Early detection of COVID-19 pneumonia based on ground-glass opacity (GGO) features of computerized tomography (CT) angiography,” in 5G IoT and Edge Computing for Smart Healthcare, Elsevier, 2022, pp. 257–277, https://doi.org/10.1016/B978-0-323-90548-0.00013-9 H. Qureshi, Z. Shah, M. A. Z. Raja, M. Y. Alshahrani, W. A. Khan, and M. Shoaib, “Machine learning investigation of tuberculosis with medicine immunity impact,” Diagn . Microbiol . Infect. Dis., vol. 110, no. 3, p. 116472, 2024.

14. Thank You