Automated identification of mycobacterium bacillus from sputum images for tuberculosis diagnosis
| Autor: | W. R. Sam Emmanuel, K. S. Mithra |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Bacillus (shape)
Tuberculosis Channel (digital image) biology business.industry Computer science 020206 networking & telecommunications Pattern recognition 02 engineering and technology medicine.disease biology.organism_classification Mycobacterium tuberculosis Tuberculosis diagnosis Feature (computer vision) Signal Processing 0202 electrical engineering electronic engineering information engineering medicine Sputum 020201 artificial intelligence & image processing Segmentation Artificial intelligence Electrical and Electronic Engineering medicine.symptom business |
| Zdroj: | Signal, Image and Video Processing. 13:1585-1592 |
| ISSN: | 1863-1711 1863-1703 |
| Popis: | Tuberculosis is a terrible, transferrable disease instigated by infection with Mycobacterium tuberculosis bacillus. A common diagnosis of this infection is the microscopic examination of sputum smears of affected persons. This paper presents an automatic system for detecting the bacillus present in the stained microscopic images of sputum which eases the job of laboratory technician with manual detection. In this work, the Channel Area Thresholding algorithm is proposed for segmentation of the bacterial image. The intensity-based local bacilli features are extracted using location-oriented histogram and speeded up robust feature algorithm. The exact classification of the bacilli objects after segmentation is carried out using deep learning neural networks. Experimental results show that the proposed work had the classification accuracy of about 97.55 on sputum image dataset. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink