| Autor: |
Mohammad A. U. Khan, Nighat Mir, Fahad Hameed Ahmad |
| Rok vydání: |
2020 |
| Předmět: |
|
| DOI: |
10.1016/b978-0-12-819043-2.00014-9 |
| Popis: |
Healthcare informatics deals with incorporation of computers and information technology into clinical workflow and to improve the quality of diagnosis in context of massive screening process. In the process computerized models are developed to measure the quality, safety, and efficiency of healthcare professionals. One particular innovation related to health informatics reported here is concerned with computer-aided chest radiography analysis for lung cancer public screening. Early detection of lung cancer is the most promising strategy to enhance a potential patient’s chance of survival. Chest X-ray provides an important diagnostic examination but they are complex in the sense that relevant diagnostic structures (soft tissues) are superimposed by irrelevant bony objects (ribs). Computer-aided detection (CAD) schemes are employed as a front end to further radiologist assessment of nodules in a given chest X-ray image. A major challenge in the current CAD scheme is the detection of nodules that are obstructed by ribs. Ribs are found to contribute large variance as compared to that of the soft tissue and this characteristic is exploited in suppressing ribs. It is proposed here if a data driven but sensitive to variance filtering technique is utilized like principal component analysis, the ribs can be identified and then suppressed to an acceptable level, greatly improving the chances of nodule detection hidden under a rib. The experiments are being conducted which provides validity of the proposed method. The rib suppression scheme can be embedded in an automated chest X-ray screening module that will greatly enhance its sensitivity in detecting small nodules. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
