Artificial intelligence utilising corneal confocal microscopy for the diagnosis of peripheral neuropathy in diabetes mellitus and prediabetes

Abstract Aims/hypothesis We aimed to develop an artificial intelligence (AI)-based deep learning algorithm (DLA) applying attribution methods without image segmentation to corneal confocal microscopy images and to accurately classify peripheral neuropathy (or lack of). Methods The AI-based DLA utilised convolutional neural networks with data augmentation to increase the algorithm’s generalisability. The algorithm was trained using a high-end graphics processor for 300 epochs on 329 corneal nerve images and tested on 40 images (1 image/participant). Participants consisted of healthy volunteer (HV) participants (n = 90) and participants with type 1 diabetes (n = 88), type 2 diabetes (n = 141) and prediabetes (n = 50) (defined as impaired fasting glucose, impaired glucose tolerance or a combination of both), and were classified into HV, those without neuropathy (PN−) (n = 149) and those with neuropathy (PN+) (n = 130). For the AI-based DLA, a modified residual neural network called ResNet-50 was developed and used to extract features from images and perform classification. The algorithm was tested on 40 participants (15 HV, 13 PN−, 12 PN+). Attribution methods gradient-weighted class activation mapping (Grad-CAM), Guided Grad-CAM and occlusion sensitivity displayed the areas within the image that had the greatest impact on the decision of the algorithm. Results The results were as follows: HV: recall of 1.0 (95% CI 1.0, 1.0), precision of 0.83 (95% CI 0.65, 1.0), F1-score of 0.91 (95% CI 0.79, 1.0); PN−: recall of 0.85 (95% CI 0.62, 1.0), precision of 0.92 (95% CI 0.73, 1.0), F1-score of 0.88 (95% CI 0.71, 1.0); PN+: recall of 0.83 (95% CI 0.58, 1.0), precision of 1.0 (95% CI 1.0, 1.0), F1-score of 0.91 (95% CI 0.74, 1.0). The features displayed by the attribution methods demonstrated more corneal nerves in HV, a reduction in corneal nerves for PN− and an absence of corneal nerves for PN+ images. Conclusions/interpretation We demonstrate promising results in the rapid classification of peripheral neuropathy using a single corneal image. A large-scale multicentre validation study is required to assess the utility of AI-based DLA in screening and diagnostic programmes for diabetic neuropathy. Graphical abstract

Download Full-text

An artificial intelligence-based deep learning algorithm for the diagnosis of diabetic neuropathy using corneal confocal microscopy: a development and validation study

Diabetologia ◽

10.1007/s00125-019-05023-4 ◽

2019 ◽

Vol 63 (2) ◽

pp. 419-430 ◽

Cited By ~ 10

Author(s):

Bryan M. Williams ◽

Davide Borroni ◽

Rongjun Liu ◽

Yitian Zhao ◽

Jiong Zhang ◽

...

Keyword(s):

Deep Learning ◽

Confocal Microscopy ◽

Diabetic Neuropathy ◽

Nerve Fibre ◽

Learning Algorithm ◽

Fibre Length ◽

Branch Points ◽

Deep Learning Algorithm ◽

Corneal Confocal Microscopy ◽

Corneal Nerve

Abstract Aims/hypothesis Corneal confocal microscopy is a rapid non-invasive ophthalmic imaging technique that identifies peripheral and central neurodegenerative disease. Quantification of corneal sub-basal nerve plexus morphology, however, requires either time-consuming manual annotation or a less-sensitive automated image analysis approach. We aimed to develop and validate an artificial intelligence-based, deep learning algorithm for the quantification of nerve fibre properties relevant to the diagnosis of diabetic neuropathy and to compare it with a validated automated analysis program, ACCMetrics. Methods Our deep learning algorithm, which employs a convolutional neural network with data augmentation, was developed for the automated quantification of the corneal sub-basal nerve plexus for the diagnosis of diabetic neuropathy. The algorithm was trained using a high-end graphics processor unit on 1698 corneal confocal microscopy images; for external validation, it was further tested on 2137 images. The algorithm was developed to identify total nerve fibre length, branch points, tail points, number and length of nerve segments, and fractal numbers. Sensitivity analyses were undertaken to determine the AUC for ACCMetrics and our algorithm for the diagnosis of diabetic neuropathy. Results The intraclass correlation coefficients for our algorithm were superior to those for ACCMetrics for total corneal nerve fibre length (0.933 vs 0.825), mean length per segment (0.656 vs 0.325), number of branch points (0.891 vs 0.570), number of tail points (0.623 vs 0.257), number of nerve segments (0.878 vs 0.504) and fractals (0.927 vs 0.758). In addition, our proposed algorithm achieved an AUC of 0.83, specificity of 0.87 and sensitivity of 0.68 for the classification of participants without (n = 90) and with (n = 132) neuropathy (defined by the Toronto criteria). Conclusions/interpretation These results demonstrated that our deep learning algorithm provides rapid and excellent localisation performance for the quantification of corneal nerve biomarkers. This model has potential for adoption into clinical screening programmes for diabetic neuropathy. Data availability The publicly shared cornea nerve dataset (dataset 1) is available at http://bioimlab.dei.unipd.it/Corneal%20Nerve%20Tortuosity%20Data%20Set.htm and http://bioimlab.dei.unipd.it/Corneal%20Nerve%20Data%20Set.htm.

Download Full-text

Using corneal confocal microscopy to compare Mecobalamin intramuscular injections vs oral tablets in treating diabetic peripheral neuropathy: a RCT

Scientific Reports ◽

10.1038/s41598-021-94284-4 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Yuanjin Zhang ◽

Dongsheng Fan ◽

Yixuan Zhang ◽

Shuo Zhang ◽

Haikun Wang ◽

...

Keyword(s):

Peripheral Neuropathy ◽

Confocal Microscopy ◽

Nerve Fibre ◽

Diabetic Peripheral Neuropathy ◽

Controlled Study ◽

Oral Tablet ◽

Injection Treatment ◽

Intramuscular Injections ◽

Corneal Confocal Microscopy ◽

Corneal Nerve

AbstractThis randomized controlled study used corneal confocal microscopy (CCM) to compare the efficacy of Mecobalamin intramuscular injections vs oral tablets in treating mild to moderate diabetic peripheral neuropathy (DPN) by detecting early nerve fiber repair. Enrolled patients were randomized approximately 1:1 to receive Mecobalamin intramuscular injections (0.5 mg/day, 3 times/week) or Mecobalamin oral tablets (1.5 mg/day) for 8 weeks. Primary outcome was change of inferior whorl length (IWL) from baseline. Secondary outcomes included changes of corneal nerve fibre length (CNFL), corneal nerve fibre density (CNFD), corneal nerve branch density (CNBD) and the Survey of Autonomic Symptoms (SAS). 15 (93.75%) patients in the injection group and 17 (89.47%) patients in the tablet group completed the study. The injection treatment significantly improved patients’ IWL from baseline (21.64 ± 3.00 mm/mm2 vs 17.64 ± 4.83 mm/mm2, P < 0.01) while the tablet treatment didn’t. Additionally, the injection treatment led to significantly improved CNFL, CNBD and SAS from baseline (all P < 0.05) while the tablet treatment did not. No patient experienced any adverse events. In conclusion, CCM is sensitive enough to detect the superior efficacy of 8-week Mecobalamin intramuscular injection treatment for DPN compared to the oral tablet treatment.ClinicalTrials.gov registration number: NCT04372316 (30/04/2020).

Download Full-text

Artificial Intelligence–Based Classification of Diabetic Peripheral Neuropathy From Corneal Confocal Microscopy Images

Diabetes Care ◽

10.2337/dc20-2012 ◽

2021 ◽

pp. dc202012

Author(s):

Tooba Salahouddin ◽

Ioannis N. Petropoulos ◽

Maryam Ferdousi ◽

Georgios Ponirakis ◽

Omar Asghar ◽

...

Keyword(s):

Artificial Intelligence ◽

Peripheral Neuropathy ◽

Confocal Microscopy ◽

Diabetic Peripheral Neuropathy ◽

Corneal Confocal Microscopy ◽

Microscopy Images

Download Full-text

Selective Disinfection Based on Directional Ultraviolet Irradiation and Artificial Intelligence

Electronics ◽

10.3390/electronics10202557 ◽

2021 ◽

Vol 10 (20) ◽

pp. 2557

Author(s):

Ben Zierdt ◽

Taichu Shi ◽

Thomas DeGroat ◽

Sam Furman ◽

Nicholas Papas ◽

...

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Ultraviolet Radiation ◽

Ultraviolet Irradiation ◽

Human Exposure ◽

Large Scale ◽

Learning Algorithm ◽

Ultraviolet Disinfection ◽

Time Efficiency ◽

Deep Learning Algorithm

Ultraviolet disinfection has been proven to be effective for surface sanitation. Traditional ultraviolet disinfection systems generate omnidirectional radiation, which introduces safety concerns regarding human exposure. Large scale disinfection must be performed without humans present, which limits the time efficiency of disinfection. We propose and experimentally demonstrate a targeted ultraviolet disinfection system using a combination of robotics, lasers, and deep learning. The system uses a laser-galvo and a camera mounted on a two-axis gimbal running a custom deep learning algorithm. This allows ultraviolet radiation to be applied to any surface in the room where it is mounted, and the algorithm ensures that the laser targets the desired surfaces avoids others such as humans. Both the laser-galvo and the deep learning algorithm were tested for targeted disinfection.

Download Full-text

T13. CORNEAL CONFOCAL MICROSCOPY DETECTS NEURAL CHANGES IN SCHIZOPHRENIA

Schizophrenia Bulletin ◽

10.1093/schbul/sbaa029.573 ◽

2020 ◽

Vol 46 (Supplement_1) ◽

pp. S235-S236

Author(s):

Peter Woodruff ◽

Georgios Ponirakis ◽

Reem Ghandi ◽

Amani Hussein ◽

Anjushri Bhagat ◽

...

Keyword(s):

Peripheral Neuropathy ◽

Confocal Microscopy ◽

Negative Symptoms ◽

High Density Lipoproteins ◽

Healthy Controls ◽

Sudomotor Function ◽

Corneal Confocal Microscopy ◽

Corneal Nerve ◽

Function Testing ◽

Positive And Negative Symptoms

Abstract Background A combination of neurodevelopmental and degenerative neural changes are likely to underpin positive and negative symptoms in schizophrenia. However, there are currently no validated biomarkers to accurately quantify the extent of neural changes in schizophrenia. Corneal confocal microscopy (CCM) is a non-invasive ophthalmic imaging technique that has been used to demonstrate in vivo corneal nerve fiber abnormalities in a range of peripheral neuropathies and central neurodegenerative disorders including Parkinson’s disease, multiple sclerosis and dementia. We wished to test the hypothesis that corneal nerve abnormalities occur in patients with schizophrenia, particularly those with negative symptoms and cognitive impairment. Methods Patients with DSM-V schizophrenia without other causes of peripheral neuropathy other than metabolic syndrome underwent assessment of clinical ratings (Positive and Negative Symptoms Scale (PANSS), cognitive function (Montreal Cognitive Assessment (MOCA) and Corneal confocal microscopy (CCM), vibration perception threshold (VPT) and sudomotor function testing. Healthy controls underwent the same assessments apart from PANSS. Results 55 subjects without (n=38) and with schizophrenia (n=17) with comparable mean age (35.7±8.5 vs 35.6±12.2, P=0.96) were studied. Patients with schizophrenia had significantly higher body weight (93.9±25.5 vs 77.1±10.1, P=0.02) and lower Low Density Lipoproteins (2.6±1.0 vs 3.4±0.7, P=0.02) compared with healthy controls. The proportion of gender, systolic and diastolic blood pressure, HbA1c, cholesterol, triglyceride and High Density Lipoproteins were comparable between the two groups. Patients with schizophrenia had significantly lower corneal nerve fibre density (CNFD, fibers/mm2) (35.6±6.5 vs 23.5±7.8, p<0.0001), branch density (CNBD, branches/mm2) (98.1±30.6 vs 34.4±26.9, p<0.0001), and fibre length (CNFL, mm/mm2) (24.2±3.9 vs 14.3±4.7, p<0.0001) compared with healthy controls but no difference in peripheral neuropathy assessed by VPT and sudomotor function testing. The area under the Receiver Operating Characteristic Curve (95% CI) of CNFD, CNBD, CNFL to distinguish patients with schizophrenia from healthy controls were 87.0% (76.8–98.2%), 93.2% (84.2–102.3%), 93.2% (84.4–102.1%), respectively. Discussion These preliminary results:

Download Full-text

Application of TLMS Deep Learning Algorithm in Artificial Intelligence Field

Proceedings of the International Symposium on Big Data and Artificial Intelligence - ISBDAI '18 ◽

10.1145/3305275.3305284 ◽

2018 ◽

Author(s):

ZhangYan ◽

ZhangHeng

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Learning Algorithm ◽

Deep Learning Algorithm

Download Full-text

Research on Efficient Deep Learning Algorithm Based on ShuffleGhost in the Field of Virtual Reality

Wireless Communications and Mobile Computing ◽

10.1155/2021/1382781 ◽

2021 ◽

Vol 2021 ◽

pp. 1-11

Author(s):

Bangtong Huang ◽

Hongquan Zhang ◽

Zihong Chen ◽

Lingling Li ◽

Lihua Shi

Keyword(s):

Virtual Reality ◽

Deep Learning ◽

Large Scale ◽

Learning Algorithm ◽

Feature Maps ◽

Embedded Devices ◽

Feature Map ◽

Deep Learning Algorithm ◽

Proper Design ◽

The Cost

Deep learning algorithms are facing the limitation in virtual reality application due to the cost of memory, computation, and real-time computation problem. Models with rigorous performance might suffer from enormous parameters and large-scale structure, and it would be hard to replant them onto embedded devices. In this paper, with the inspiration of GhostNet, we proposed an efficient structure ShuffleGhost to make use of the redundancy in feature maps to alleviate the cost of computations, as well as tackling some drawbacks of GhostNet. Since GhostNet suffers from high computation of convolution in Ghost module and shortcut, the restriction of downsampling would make it more difficult to apply Ghost module and Ghost bottleneck to other backbone. This paper proposes three new kinds of ShuffleGhost structure to tackle the drawbacks of GhostNet. The ShuffleGhost module and ShuffleGhost bottlenecks are utilized by the shuffle layer and group convolution from ShuffleNet, and they are designed to redistribute the feature maps concatenated from Ghost Feature Map and Primary Feature Map. Besides, they eliminate the gap of them and extract the features. Then, SENet layer is adopted to reduce the computation cost of group convolution, as well as evaluating the importance of the feature maps which concatenated from Ghost Feature Maps and Primary Feature Maps and giving proper weights for the feature maps. This paper conducted some experiments and proved that the ShuffleGhostV3 has smaller trainable parameters and FLOPs with the ensurance of accuracy. And with proper design, it could be more efficient in both GPU and CPU side.

Download Full-text