[This article belongs to Volume - 54, Issue - 02]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-28-10-2022-373

Title : EARLY PNEUMONIA DETECTION: A NEW AUTOMATED APPROACH WITH IMPROVED DIAGNOSTIC PERFORMANCE USING MACHINE LEARNING
D.Sreenivasa Rao1, Dr.S.Anu H Nair2, Dr.Thota Venkat Narayana Rao3,Dr.K.P. Sanal Kumar4

Abstract :

An efficient method of Pneumonia screening is the Low Dose Computed Tomography (LDCT) imaging modality but there are certain perils involved in the LDCT screening method. One such risk factor is the occurrence of FP in the tests, which leads to avoidable conditions such as surgery and complications. There are also chances of over diagnosis. The repeated screening for patients may increase the chances of developing secondary Pneumonia in a patient who are not affected by Pneumonia. Pneumonia detected at early stages can be treated more effectively with surgery. The survival rate of the patients diagnosed through LDCT screening test at an early stage was considerable compared to the patients diagnosed through chest x-ray screening test. When the Pneumonia is localized the five year survival rate is 56%. LDCT screening makes an early detection of Pneumonia which decreases the mortality rate by 14% to 20% among the high risk population. Diagnosis of the Pneumonia patient not only relies on the LDCT images but also on the report of the radiologist. In order to ameliorate the diagnosis of the Pneumonia by the radiologist there is a need to design a CADx for lung nodule classification. This CADx is an aid to the radiologist providing a second opinion on the conclusions made by the radiologist regarding the status of the patient‟s condition. In a nutshell developing a CADx for lung nodule classification will prove to be an effective tool for the radiologist to conclude their diagnosis with increased accuracy.