Hong Liang Tey | Nanyang Technological University (original) (raw)

Papers by Hong Liang Tey

Acta dermato-venereologica, 2018

European Journal of Immunology, Mar 23, 2018

Pressure ulcers are a chronic problem for patients or the elderly who require extended periods of... more Pressure ulcers are a chronic problem for patients or the elderly who require extended periods of bed rest. The formation of ulcers is due to repeated cycles of ischemiareperfusion (IR), which initiates an inflammatory response. Advanced ulcers disrupt the skin barrier, resulting in further complications. To date, the immunological aspect of skin IR has been understudied, partly due to the complexity of the skin immune cells. Through a combination of mass cytometry, confocal imaging and intravital multiphoton imaging, this study establishes a workflow for multidimensionality single cell analysis of skin myeloid cell responses in the context of IR injury with high spatiotemporal resolution. The data generated has provided us with previously uncharacterized insights into the distinct cellular behavior of resident dendritic cells (DCs) and recruited neutrophils post IR. Of interest, we observed a drop in DDC numbers in the IR region, which was subsequently replenished 48h post IR. More importantly, in these cells, we observe an attenuated response to repeated injuries, which may have implications in the subsequent wound healing process.

PubMed, Jun 1, 2005

Refeeding oedema in patients with anorexia nervosa is a known but yet under-reported and poorly-u... more Refeeding oedema in patients with anorexia nervosa is a known but yet under-reported and poorly-understood condition. We illustrate this condition in a 19-year-old girl with anorexia nervosa who developed bipedal oedema after she was started on nutritional therapy. It is important to be aware of the differential diagnoses of oedema in such cases, which includes heart failure and previous diuretic abuse. Refeeding oedema generally resolves spontaneously but some individuals may require treatment.

Biomedical Optics Express, Jul 6, 2018

Automatic skin layer segmentation in optical coherence tomography (OCT) images is important for a... more Automatic skin layer segmentation in optical coherence tomography (OCT) images is important for a topographic assessment of skin or skin disease detection. However, existing methods cannot deal with the problem of shadowing in OCT images due to the presence of hair, scales, etc. In this work, we propose a method to segment the topmost layer of the skin (or the skin surface) using 3D graphs with a novel cost function to deal with shadowing in OCT images. 3D graph cuts use context information across B-scans when segmenting the skin surface, which improves the segmentation as compared to segmenting each B-scan separately. The proposed method reduces the segmentation error by more than 20% as compared to the best performing related work. The method has been applied to roughness estimation and shows a high correlation with a manual assessment. Promising results demonstrate the usefulness of the proposed method for skin layer segmentation and roughness estimation in both normal OCT images and OCT images with shadowing.

arXiv (Cornell University), Mar 2, 2017

In this study, a multi-task deep neural network is proposed for skin lesion analysis. The propose... more In this study, a multi-task deep neural network is proposed for skin lesion analysis. The proposed multi-task learning model solves different tasks (e.g., lesion segmentation and two independent binary lesion classifications) at the same time by exploiting commonalities and differences across tasks. This results in improved learning efficiency and potential prediction accuracy for the task-specific models, when compared to training the individual models separately. The proposed multi-task deep learning model is trained and evaluated on the dermoscopic image sets from the International Skin Imaging Collaboration (ISIC) 2017 Challenge "Skin Lesion Analysis towards Melanoma Detection", which consists of 2000 training samples and 150 evaluation samples. The experimental results show that the proposed multi-task deep learning model achieves promising performances on skin lesion segmentation and classification. The average value of Jaccard index for lesion segmentation is 0.724, while the average values of area under the receiver operating characteristic curve (AUC) on two individual lesion classifications are 0.880 and 0.972, respectively.

Journal of Dermatology, Apr 27, 2013

(P = 0.05). Interestingly, different environmental contaminant fungi were isolated, including Asp... more (P = 0.05). Interestingly, different environmental contaminant fungi were isolated, including Aspergillus fumigatus, Fusarium sp. and Scopulariopsis brevicaulis, three of which can cause onychomycosis. Several authors have pointed out that the high recurrence of onychomycosis is due to the maintenance of several predisposing factors and other conditions like physical abnormalities, and short and insufficient therapies. For example, Nishimoto reports that lesions caused by T. rubrum give a mild inflammatory reaction and the majority of patients are left untreated or are treated intermittently. This study demonstrates that dermatophytes can survive in socks, even after they have been washed. According to our findings, at least 10.46% of patients with tinea pedis and dermatophytic onychomycosis are at risk of re-infection by contact with their socks. After this study, we can add to that a percentage of patients could be re-infected because the dermatophytes can survive in the textile of the socks. Alexandro BONIFAZ, Denisse VÁZQUEZ-GONZÁLEZ Marco Antonio HERNÁNDEZ, Javier ARAIZA, Andrés TIRADO-SÁNCHEZ, Rosa Mar ıa PONCE Dermatology Service and Mycology Department, General Hospital of Mexico, Mexico City, Mexico

Dermatology and therapy, Aug 2, 2019

Introduction: Keloids are a prevalent chronic skin disorder with significant psychosocial morbidi... more Introduction: Keloids are a prevalent chronic skin disorder with significant psychosocial morbidity. Intralesional corticosteroid injections are the first-line treatment but are painful and require repeated injections by medical professionals. Dissolving microneedles are a novel method of cutaneous drug delivery that induces minimal/no pain and can be self-administered. The objective of the study was to evaluate the efficacy and safety of triamcinolone-embedded dissolving microneedles in treatment of keloids. Methods: This was a single-blind, intra-individual controlled two-phase clinical trial of 8-week duration each. Two keloids per subject were selected for (1) once-daily 2-min application with microneedles for 4 weeks, followed by no treatment for the next 4 weeks, or (2) nonintervention as control. Primary outcome was change in keloid volume as assessed by a highresolution 3D scanner. Results: There was significant reduction in keloid volume compared with controls after 4 weeks of treatment. This reduction was greater with a higher dosage of triamcinolone used. Conclusions: Once-daily application of dissolving triamcinolone-embedded microneedles significantly reduced the volume of keloids. The treatment was safe, can be self-administered and can serve as an alternative for patients unsuitable for conventional treatments.

Journal of Materials Chemistry B, 2017

Skin Research and Technology, Jan 29, 2021

Dear Editors, Scleroderma is the most common form of fibrotic disease involving the skin, and the... more Dear Editors, Scleroderma is the most common form of fibrotic disease involving the skin, and the subtype that only affects the skin is known as morphea. The conventional investigation for cutaneous fibrotic diseases is skin biopsy, an invasive ex vivo procedure. Typically, another biopsy of adjacent normal skin is required for comparison, in order to visualize the excessive dermal collagen deposition. Optical coherence tomography has been shown to be useful in fibrotic skin disorders, in obtaining in vivo cross-sectional visualization.1-3 The high-definition optical coherence tomography (HD-OCT) (Skintell®) developed more recently allows imaging with an enhanced resolution of 3μm in both axial and lateral planes and has an imaging depth of 570 μm.4 We developed a software platform that performs automated 3D skin layer segmentation in HD-OCT skin images5 and subsequent quantitative measurement of epidermal thickness and dermal heterogeneity across the entire skin layers in 3D. Using the HD-OCT handheld probe, patients with fibrotic skin diseases can be imaged at the bedside within 3 seconds and have their progress monitored without skin biopsies using this non-invasive imaging platform. A graph-based approach was used to segment the different skin layers by considering the voxels in the volumetric HD-OCT scan as vertices of a graph. Weights were assigned between the vertices using a cost function based on gradients in the axial direction. Smoothness constraints were also included between voxels in both axial and lateral dimensions to reduce the sensitivity of the approach to image artefacts such as shadowing. The skin surface and dermal-epidermal junctions (DEJ) were detected and used to segment the volumetric epidermal layer for subsequent analysis. Epidermal thickness was calculated as the mean thickness of the detected epidermis. To quantize the variations in dermal signal intensity, we calculated the dermal heterogeneity, defined as the standard deviation of voxels in the dermis up to 60μm below the DEJ, where a lower value represents reduced intensity variation. Analysis was performed on the HD-OCT images of a 55-year-old Chinese gentleman with biopsy-confirmed generalized morphea. He was successfully treated with bath psoralen with Ultraviolet A therapy, and his skin reverted to near normal clinically after treatment. We analysed the volumetric image scans at 13 pre-treatment and 2 post-treatment sites. Automated segmentation of epidermis and dermis layer was performed, followed by the quantification of epidermal thickness and dermal heterogeneity. There was a trend towards an increase in mean 3D epidermal thickness, from 65.8 ± 1.8 μm to 67.8 ± 0.7 μm (P = .1735) post-treatment (Figure 1). This is consistent with results of a previous study using 2D images in conventional OCT performed on 32 patients with morphea, which showed a small increase in epidermal thickness from affected to unaffected sites.2 There was a trend towards an increase in dermal heterogeneity values, from 41.1 ± 4.8% to 46.9 ± 6.9% (P = .2027) post-treatment (Figure 1). An increase in prominence of adnexal structures and blood vessels with less compactness of collagen is expected in successfully treated skin, which would result in a greater variation in refractive indices in the dermis and a higher dermal heterogeneity value. Optical density at 300 μm (OD300) was validated by Abignano et al as a useful parameter to distinguish between normal and morphea-affected skin.3 It was found that healthy subjects have higher mean OD300 due to the increased prominence of adnexal structures and blood vessels. In our study, we also noted similar findings, with mean OD300 for normal skin at 0.47 ± 0.088 compared to 0.350 ± 0.05 for pre-treatment morphea site. We noted no significant difference between mean OD300 before and after treatment in our patient, at 0.350 ± 0.05 and 0.335 ± 0.021 (P = .0733); this may be due to our small sample size (n = 4). These findings suggest that OD300 could be used as a measure to discriminate between healthy versus fibrotic skin. Using the 3D segmentation function in the software, we were able to remove the epidermal layer in the 3D reconstructed image (Figure 2). The surface of the dermis appears flat at baseline, which could correspond to flattening of the epidermal rete ridges and attenuation of the dermal papillae in morphea. In post-treatment skin, increased unevenness of the dermal surface is observed, which could correspond to the increased undulation of the epidermal rete ridges and dermal papillae.

Dermatology, 2019

Background/Aims: We aim to examine the hypothesis that antihypertensive drugs and statins may be ... more Background/Aims: We aim to examine the hypothesis that antihypertensive drugs and statins may be responsible for the development of idiopathic generalized exfoliative dermatitis (GED) with a case-control study. Methods: All inpatients who were hospitalized under the dermatology service at the Tan Tock Seng Hospital, Singapore, between 1 May 2013 and 31 May 2015, were analysed. Idiopathic GED cases had consistent clinical and histological features but no apparent cause despite comprehensive evaluation. Controls were randomly selected from inpatients with other dermatological conditions in a 1: 1 ratio during the same period. Their relationship was analysed using univariate (χ2 or Fisher exact tests) and multivariate logistic regression analysis. Results: There were 78 cases and 83 controls. Of the 78 cases, 42 patients had a history of treatment with antihypertensive drugs or statins. Cases were not found to be more likely on angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, β-blockers, calcium channel blockers, thiazides or statins compared to controls (OR: 0.81; 95% CI: 0.43–1.51; p = 0.507). Conclusions: There was insufficient evidence to suggest a significant relationship between the chronic use of antihypertensive drugs or statins and idiopathic GED in this study, despite previous evidence reporting this might be so. Further case-control studies with larger sample sizes are needed to evaluate this association.

International Journal of Dermatology, Sep 24, 2018

Hypohidrosis refers to impaired ability to sweat despite appropriate thermal or sympathetic stimu... more Hypohidrosis refers to impaired ability to sweat despite appropriate thermal or sympathetic stimuli. Extensive hypohidrosis is potentially fatal as it may lead to hyperthermia, heat exhaustion, heat stroke, and death. Currently, thermoregulatory sweat tests are used for diagnosis. Due to high logistics and manpower

Journal of Dermatology, Jan 9, 2012

Dear Editor, We would like to thank Dr Dekio for his comments on our earlier article (‘‘Follicula... more Dear Editor, We would like to thank Dr Dekio for his comments on our earlier article (‘‘Follicular spicules associated with Propionibacterium acnes with response to erythromycin’’. J Dermatol 2011; 38: 409– 411). We agree that species-specific confirmation of histological findings, namely in situ hybridization, would be preferable in confirming the identity of the organisms. However, this technique to identify Propionibacterium acnes is not readily available and we believe there is sufficient evidence to associate P. acnes with the follicular spicules. First, the organisms seen histologically on hematoxylin–eosin stain were indeed Gram-positive rods, although we acknowledge that Staphylococcus species, especially when in clumps, can be very difficult to differentiate from bacilli. The rods observed were unlikely to be Corynebacterium species as they tend to colonize intertriginous regions and when pathogenic, cause erythrasma or trichomycosis at these body sites. In our patient, the formation of follicular spicules correspond to the sites typically colonized by P. acnes, namely the neck, upper chest and back. In addition, the pathology involves the follicles, which is the normal habitat of P. acnes. Second, although prolonged skin cultures may be positive for P. acnes, in a study performed in our center, the positive rate of microbiological culture performed on comedones in acne disease was only close to 60% (Dr Audrey Tan, unpubl. data, 2010). In our setting, a positive culture is thereby associated with a high level and pathogenicity of P. acnes. In our patient, the culture was performed from the spicules (and not from skin) and the bacteria grown reflect those seen on histology. Third, we agree that clinical response to treatment with topical and oral erythromycin is not specific for P. acnes. It does, however, corroborate with the histological and microbiological findings, and the overall evidence indicates the association between the follicular spicules and P. acnes.

Advances in Experimental Medicine and Biology, 2020

The skin is the largest organ of our body. Skin disease abnormalities which occur within the skin... more The skin is the largest organ of our body. Skin disease abnormalities which occur within the skin layers are difficult to examine visually and often require biopsies to make a confirmation on a suspected condition. Such invasive methods are not well-accepted by children and women due to the possibility of scarring. Optical coherence tomography (OCT) is a non-invasive technique enabling in vivo examination of sub-surface skin tissue without the need for excision of tissue. However, one of the challenges in OCT imaging is the interpretation and analysis of OCT images. In this review, we discuss the various methodologies in skin layer segmentation and how it could potentially improve the management of skin diseases. We also present a review of works which use advanced machine learning techniques to achieve layers segmentation and detection of skin diseases. Lastly, current challenges in analysis and applications are also discussed.

Dermatologic Clinics, Jul 1, 2018

Chronic pruritus in the geriatric population is defined by pruritus that persists for more than 6... more Chronic pruritus in the geriatric population is defined by pruritus that persists for more than 6 weeks experienced by people 65 year old and above. Aging is associated with pathologic changes is the epidermal barrier, immune system, and the nervous system that predispose the elderly to itch. In our approach to the clinical problem, we dichotomize pruritus based on its pathogenesis, either histaminergic or predominantly nonhistaminergic. Topical treatments are generally safe for the elderly population; systemic treatments are chosen depending on the condition with consideration of comorbid diseases and drug interactions. Numerous new medications are currently undergoing clinical trials and they are anticipated to enter the clinics in the near future.

Journal der Deutschen Dermatologischen Gesellschaft, Aug 27, 2015

At present, beyond clinical assessment, the diagnosis of skin diseases is primarily made histolog... more At present, beyond clinical assessment, the diagnosis of skin diseases is primarily made histologically. However, skin biopsies have many disadvantages, including pain, scarring, risk of infection, and sampling error. With recent advances in skin imaging technology, the clinical use of imaging methods for the practical management of skin diseases has become an option. The in vivo high-definition optical coherence tomography (HD-OCT) has recently been developed and commercialized (Skintell®; Agfa, Belgium). Compared with conventional OCT, it has a higher resolution; compared with reflectance confocal microscopy, it has a shorter time for image acquisition as well as a greater penetration depth and a larger field of view. HD-OCT is promising but much work is still required to develop it from a research tool to a valuable adjunct for the noninvasive diagnosis of skin lesions. Substantial work has been done to identify HD-OCT features in various diseases but interpretation can be time-consuming and tedious. Projects aimed at automating these processes and improving image quality are currently under way.

IEEE Photonics Technology Letters, 2017

Indian Journal of Dermatology, Venereology and Leprology, 2017

A major challenge in the management of adult Henoch-Schönlein purpura is the difficulty in assess... more A major challenge in the management of adult Henoch-Schönlein purpura is the difficulty in assessing the risk of systemic involvement. There is currently a paucity of data in this area. This study sought to determine specific clinical and histopathological features associated with systemic involvement in adult Henoch-Schönlein purpura. We reviewed the records of 99 adult Henoch-Schönlein purpura patients who presented at the National Skin Centre, Singapore, between January 2008 and May 2015. Renal involvement was found in 56 (56.6%) patients, joint involvement in 21 (21.2%) and gastrointestinal involvement in 13 (13.1%). Age &amp;amp;amp;gt; 30 years was an independent predictor of renal involvement with an adjusted odds ratio of 2.97 (95% confidence interval, 1.08-8.16; P = 0.04). Risk factors for significant renal involvement necessitating nephrology referral were further evaluated: the odds were approximately 60% higher for every 10-year increase in age (95% confidence interval, 1.02-2.57; P = 0.04) and patients with cutaneous bullae and/or necrosis had an almost six times higher risk (95% confidence interval, 1.43-25.00; P = 0.01). This study was limited by its retrospective design. We also lacked long-term data to examine how clinical and histopathological characteristics correlated with long-term disease outcomes. Adult Henoch-Schönlein purpura patients older than 30 years have a threefold increased risk of renal involvement. The risk of profound renal disease necessitating nephrology referral rose significantly with age and the presence of cutaneous bullae and/or necrosis.

JAMA Dermatology, Feb 1, 2015

PLOS ONE, Dec 16, 2020

Self-sampling for SARS-CoV-2 would significantly raise testing capacity and reduce healthcare wor... more Self-sampling for SARS-CoV-2 would significantly raise testing capacity and reduce healthcare worker (HCW) exposure to infectious droplets personal, and protective equipment (PPE) use. We conducted a diagnostic accuracy study where subjects with a confirmed diagnosis of COVID-19 (n = 401) and healthy volunteers (n = 100) were asked to self-swab from their oropharynx and mid-turbinate (OPMT), and self-collect saliva. The results of these samples were compared to an OPMT performed by a HCW in the same patient at the same session. In subjects confirmed to have COVID-19, the sensitivities of the HCW-swab, self-swab, saliva, and combined self-swab plus saliva samples were 82.8%, 75.1%, 74.3% and 86.5% respectively. All samples obtained from healthy volunteers were tested negative. Compared to HCW-swab, the sensitivities of a self-swab sample and saliva sample were inferior by 8.7% (95%CI: 2.4% to 15.0%, p = 0.006) and 9.5% (95%CI: 3.1% to 15.8%, p = 0.003) respectively. The combined detection rate of self-swab and saliva had a sensitivity of 2.7% (95%CI: -2.6% to 8.0%, p = 0.321). The sensitivity of both the self-collection methods are higher when the Ct value of the HCW swab is less than 30. The specificity of both the selfswab and saliva testing was 100% (95% CI 96.4% to 100%).

Indian Journal of Dermatology, Venereology and Leprology, Apr 19, 2022