Involution Receptive Field Network for COVID-19 Diagnosis | Chapter 3 | New Trends in Physical Science Research Vol. 6

 Since its outbreak, COVID-19, a brand-new viral disease that causes severe pneumonia-related respiratory failure, has been sweeping the globe. The Community-Acquired Pneumonia (CAP), Normal, and COVID-19 Computed Tomography (CT) scan pictures are identified using the Involution Receptive Field Network from the Large COVID-19 CT scan slice dataset. N-pair contrastive loss is included during the network's training in order to improve the embedding representation for CT scan slices in the latent dimension. In order to enhance feature map extraction, the lightweight Involution Receptive Field Network-Medium (InRFNet-M) is presented. It is both channel-neutral and spatially specific. The assessment findings for the InRFNet-M model show a high level of validation accuracy (99 percent). The suggested InRFNet-M: Involution Receptive Field Network-Medium has shown effective classification with good accuracy and recall scores.

Author(s) Details:

M. Dhruv,
CKM VIGIL Pvt Ltd, Hyderabad, Telangana, India.

R. Sai Chandra Teja,
CKM VIGIL Pvt Ltd, Hyderabad, Telangana, India.

R. Sri Devi,
Department of Anesthesiology, Sri Venkateswara Institute of Medical Sciences SVIMS Tirupati, Chittor District, Andhra Pradesh, India.

S. Nagesh Kumar,
Department of Surgical Oncology, Sri Venkateswara Institute of Medical Sciences SVIMS Tirupati, Chittor District, Andhra Pradesh, India.

Please see the link here: https://stm.bookpi.org/NTPSR-V6/article/view/7209

Statistical Measures Performance for Preprocessing and Locate ROI Using Segmentation in Mammograms | Chapter 9 | Innovations in Medicine and Medical Research Vol. 3

Breast cancer is a cancer that develops from breast tissue and it is considered one of the deadly diseases for women. X-ray mammography is the most successful method for the early phase detection of breast diseases. Micro-calcifications in mammogram have been targeted as earliest symptom of breast cancer and their detection is very important to improve its prediction. For micro calcifications, the interpretation of their presence is very tricky because they are very small, typically in the range of 0.1 mm- 1.0 mm and average is about 0.3 mm. Micro-calcification is a tiny bit, which may not be clearly visible because of low contrast in the mammogram image. To and locate the region of interest using segmentation technique. This paper presents statistical measures of different preprocessing to lighten the low contrast images and selecting filter for accurate detection of microcalicifications with proper segmentation technique. 

Author(s)  Details

Neha N. Ganvir
Department of Electronics Engineering, Sinhgad Institute of Technology and Science (SITS), Narhe, Sinhgad Technical Education Society (STES), Pune, India.

Dr. D. M. Yadav
Pune University, India.

View Book :- http://bp.bookpi.org/index.php/bpi/catalog/book/178