Brain MRI‐based Wilson disease tissue classification: an optimised deep transfer learning approach. Issue 25 (4th November 2020)

• Record Type: Journal Article
• Authors: Saba, L.; Agarwal, M.; Sanagala, S.S.; Gupta, S.K.; Sinha, G.R.; Johri, A.M.; Khanna, N.N.; Mavrogeni, S.; Laird, J.R.; Pareek, G.; Miner, M.; Sfikakis, P.P.; Protogerou, A.; Viswanathan, V.; Kitas, G.D.; Suri, J.S.
• Is Part Of: Electronics letters. Volume 56:Issue 25(2020)
• Page Start: 1395
• Page End: 1398
• Publication Date: 2020-11-04
• Subjects: learning (artificial intelligence) -- liver -- brain -- diseases -- image classification -- biological tissues -- medical image processing -- biomedical MRI; Visual Geometric Group‐19 -- machine learning paradigm -- VGG‐19 -- MobileNet -- brain MRI‐based Wilson disease tissue classification -- optimised deep transfer learning approach -- liver -- white matter hyperintensity -- MRI scans -- brain -- computer‐aided design‐based automated classification strategy -- four‐fold augmentation -- AUC pairs -- ML‐based soft classifier -- random forest; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2020.2102
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : Wilson's disease (WD) is caused by the excessive accumulation of copper in the brain and liver, leading to death if not diagnosed early. WD shows its prevalence as white matter hyperintensity (WMH) in MRI scans. It is challenging and tedious to classify WD against controls when comparing visually, primarily due to subtle differences in WMH. This Letter presents a computer‐aided design‐based automated classification strategy that uses optimised transfer learning (TL) utilising two novel paradigms known as (i) MobileNet and (ii) the Visual Geometric Group‐19 (VGG‐19). Further, the authors benchmark TL systems against a machine learning (ML) paradigm. Using four‐fold augmentation, VGG‐19 is superior to MobileNet demonstrating accuracy and area under the curve (AUC) pairs as 95.46 ± 7.70 %, 0.932 ( p < 0.0001 ) and 86.87 ± 2.23 %, 0.871 ( p < 0.0001 ), respectively. Further, MobileNet and VGG‐19 showed an improvement of 3.4 and 13.5%, respectively, when benchmarked against the ML‐based soft classifier – Random Forest.