Details

Autor(en) / Beteiligte

• Huidrom, Ratishchandra
• Chanu, Yambem Jina
• Singh, Khumanthem Manglem

Titel

Neuro-evolutional based computer aided detection system on computed tomography for the early detection of lung cancer

Ist Teil von

• Multimedia tools and applications, 2022-09, Vol.81 (22), p.32661-32673

Ort / Verlag

New York: Springer US

Erscheinungsjahr

2022

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• Lung cancer is one of the highest deadly disease which can be treated effectively in its early stage. Computer aided detection (CADe) can detect pulmonary nodules of lung cancer accurately and faster than manual detection. This paper presents a new CADe system using neuro-evolutional approach. The proposed method is focused on machine learning algorithm which is a crucial area of the system. The CADe system extracts lung regions from computed tomography images and detects pulmonary nodules within the lung regions. False positive reduction is performed by using a new neuro-evolutionary approach which consists of a feed-forward neural network and a combination of cuckoo search algorithm and particle swarm optimization. The performance of the proposed method is further improved by using regularized discriminant features and achieves 95.8% sensitivity, 95.3% specificity and 95.5% accuracy.