scholarly journals Non-invasive scoring of cellular atypia in keratinocyte cancers in 3D LC-OCT images using Deep Learning

2022 ◽  
Vol 12 (1) ◽  
Author(s):  
Sébastien Fischman ◽  
Javiera Pérez-Anker ◽  
Linda Tognetti ◽  
Angelo Di Naro ◽  
Mariano Suppa ◽  
...  
Keyword(s):  
Deep Learning ◽  
Invasive Procedure ◽  
Skin Lesions ◽  
Small Sample ◽  
Three Dimensions ◽  
Line Field ◽  
Biologically Relevant ◽  
Non Invasive ◽  
Cellular Atypia

AbstractDiagnosis based on histopathology for skin cancer detection is today’s gold standard and relies on the presence or absence of biomarkers and cellular atypia. However it suffers drawbacks: it requires a strong expertise and is time-consuming. Moreover the notion of atypia or dysplasia of the visible cells used for diagnosis is very subjective, with poor inter-rater agreement reported in the literature. Lastly, histology requires a biopsy which is an invasive procedure and only captures a small sample of the lesion, which is insufficient in the context of large fields of cancerization. Here we demonstrate that the notion of cellular atypia can be objectively defined and quantified with a non-invasive in-vivo approach in three dimensions (3D). A Deep Learning (DL) algorithm is trained to segment keratinocyte (KC) nuclei from Line-field Confocal Optical Coherence Tomography (LC-OCT) 3D images. Based on these segmentations, a series of quantitative, reproducible and biologically relevant metrics is derived to describe KC nuclei individually. We show that, using those metrics, simple and more complex definitions of atypia can be derived to discriminate between healthy and pathological skins, achieving Area Under the ROC Curve (AUC) scores superior than 0.965, largely outperforming medical experts on the same task with an AUC of 0.766. All together, our approach and findings open the door to a precise quantitative monitoring of skin lesions and treatments, offering a promising non-invasive tool for clinical studies to demonstrate the effects of a treatment and for clinicians to assess the severity of a lesion and follow the evolution of pre-cancerous lesions over time.

scholarly journals A simple method to measure sulfonation in man using paracetamol as probe drug

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Natália Marto ◽  
Judit Morello ◽  
Alexandra M. M. Antunes ◽  
Sofia Azeredo ◽  
Emília C. Monteiro ◽  
...  
Keyword(s):  
High Resolution Mass Spectrometry ◽  
Invasive Procedure ◽  
Urine Samples ◽  
Intraindividual Variability ◽  
Empirical Measure ◽  
Urine Ph ◽  
Simple Method ◽  
Spot Urine ◽  
Non Invasive

AbstractSulfotransferase enzymes (SULT) catalyse sulfoconjugation of drugs, as well as endogenous mediators, gut microbiota metabolites and environmental xenobiotics. To address the limited evidence on sulfonation activity from clinical research, we developed a clinical metabolic phenotyping method using paracetamol as a probe substrate. Our aim was to estimate sulfonation capability of phenolic compounds and study its intraindividual variability in man. A total of 36 healthy adult volunteers (12 men, 12 women and 12 women on oral contraceptives) received paracetamol in a 1 g-tablet formulation on three separate occasions. Paracetamol and its metabolites were measured in plasma and spot urine samples using liquid chromatography-high resolution mass spectrometry. A metabolic ratio (Paracetamol Sulfonation Index—PSI) was used to estimate phenol SULT activity. PSI showed low intraindividual variability, with a good correlation between values in plasma and spot urine samples. Urinary PSI was independent of factors not related to SULT activity, such as urine pH or eGFR. Gender and oral contraceptive intake had no impact on PSI. Our SULT phenotyping method is a simple non-invasive procedure requiring urine spot samples, using the safe and convenient drug paracetamol as a probe substrate, and with low intraindividual coefficient of variation. Although it will not give us mechanistic information, it will provide us an empirical measure of an individual’s sulfonator status. To the best of our knowledge, our method provides the first standardised in vivo empirical measure of an individual’s phenol sulfonation capability and of its intraindividual variability. EUDRA-CT 2016-001395-29, NCT03182595 June 9, 2017.

scholarly journals Antibody-guided in vivo imaging of Aspergillus fumigatus lung infections during antifungal azole treatment

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Sophie Henneberg ◽  
Anja Hasenberg ◽  
Andreas Maurer ◽  
Franziska Neumann ◽  
Lea Bornemann ◽  
...  
Keyword(s):  
Aspergillus Fumigatus ◽  
Ex Vivo ◽  
Invasive Pulmonary Aspergillosis ◽  
Invasive Procedure ◽  
Diagnostic Procedures ◽  
Light Sheet ◽  
Antifungal Drugs ◽  
Non Invasive ◽  
Lung Infections

AbstractInvasive pulmonary aspergillosis (IPA) is a life-threatening lung disease of immunocompromised humans, caused by the opportunistic fungal pathogen Aspergillus fumigatus. Inadequacies in current diagnostic procedures mean that early diagnosis of the disease, critical to patient survival, remains a major clinical challenge, and is leading to the empiric use of antifungal drugs and emergence of azole resistance. A non-invasive procedure that allows both unambiguous detection of IPA and its response to azole treatment is therefore needed. Here, we show that a humanised Aspergillus-specific monoclonal antibody, dual labelled with a radionuclide and fluorophore, can be used in immunoPET/MRI in vivo in a neutropenic mouse model and 3D light sheet fluorescence microscopy ex vivo in the infected mouse lungs to quantify early A. fumigatus lung infections and to monitor the efficacy of azole therapy. Our antibody-guided approach reveals that early drug intervention is critical to prevent complete invasion of the lungs by the fungus, and demonstrates the power of molecular imaging as a non-invasive procedure for tracking IPA in vivo.

scholarly journals Deep Learning Analysis of in Vivo Hyperspectral Images for Automated Intraoperative Nerves Detection

2021 ◽  
Author(s):  
Manuel Barberio ◽  
Toby Collins ◽  
Valentin Bencteux ◽  
Richard Nkusi ◽  
Eric Felli ◽  
...  
Keyword(s):  
Deep Learning ◽  
Cross Validation ◽  
State Of The Art ◽  
Invasive Technique ◽  
Midline Incision ◽  
Average Sensitivity ◽  
Non Invasive ◽  
First Time ◽  
Learning Analysis

Abstract Nerves are difficult to recognize during surgery and inadvertent injuries may occur, bringing catastrophic consequences for the patient. Hyperspectral imaging (HSI) is a non-invasive technique combining photography with spectroscopy, allowing biological tissue property quantification. We show for the first time that HSI combined with deep learning allows nerves and other tissue types to be automatically recognized in-vivo at the pixel level. An animal model is used comprising eight anesthetized pigs with a neck midline incision, exposing several structures (nerve, artery, vein, muscle, fat, skin). State-of-the-art machine learning models have been trained to recognize these tissue types in HSI data. The best model is a Convolutional Neural Network (CNN), achieving an overall average sensitivity of 0.91 and specificity of 0.99, validated with leave-one-patient-out cross-validation. For the nerve, the CNN achieves an average sensitivity of 0.76 and specificity of 1.0. In conclusion, HSI combined with a CNN model is suitable for in vivo nerve recognition.

scholarly journals Frontier in hair loss and trichoscopy: A review

2016 ◽  
Vol 1 (2) ◽  
Author(s):  
Ebtisam Elghblawi
Keyword(s):  
Hair Loss ◽  
Skin Lesions ◽  
Hair Shaft ◽  
Hair Follicles ◽  
Invasive Technique ◽  
Non Invasive ◽  
Exclamation Mark ◽  
Indispensable Tool ◽  
Hair Diseases

<p>Skin surfaces have always been examined using dermoscopy, a familiar tool which is useful to magnify and examine skin especially in cases of pigmented skin lesions. However, to examine the hair and scalp, a practical tool called trichoscopy has surfaced recently and has proven to be handy and functional in diagnosing most hair-related diseases. It is also referred to as dermoscopy of the hair and the scalp. It can aid in assessing active diseases in the scalp and hair, such as yellow dots, dystrophic hairs, cadaverized black dots, white dots, and exclamation mark hairs – all of which denote specific criteria for hair diseases. Trichoscopy is a very newly developed non-invasive technique for hair image analysis. It permits non-invasive visualization of hair shafts at higher intensification (about ×70 and ×100) and enables measurement of hair shaft width without the need for removing hair for diagnostic reasons. Moreover, it helps <em>in vivo</em> visualization of the epidermal portion of hair follicles and perifollicular epidermis (orifices). Consequently, it is valuable as it permits the inspection of structures that are otherwise not seen by the naked eye. Trichoscopy is the new frontier for the diagnosis of hair and scalp disease. Nowadays, a trichoscope is considered a must for dermatologists and it is a hot topic in the treatment of hair diseases. There is pooled evidence that the utilization of trichoscopy in the clinical setting for evaluating hair disorders can improve its diagnostic capability beyond simple clinical scrutiny. Trichoscopy can identify both hair shaft and hair opening abnormalities without the need for hair sampling, as well as distinguish between different scalp and hair diseases. Furthermore, it can give easy and quick evaluation of the hair with a follow-up to determine progress and prognosis of the disease with photos. It can also aid in some genetic hair shaft dystrophies such as trichorrhexis nodosa, trichorrhexis invaginata, monilethrix, pili annulati, and pili torti. The limitation of trichoscopy is that it needs prior knowledge to apply it effectively in order to mandate an efficient use by correctly interpreting the findings and their significance. In cases where there are unsettled discrepancies, a histopathological investigation is needed. The interest in trichoscopy has vastly increased and has become an indispensable tool in evaluating patients with hair loss. The aim of this review is to supplement existing knowledge on trichoscopy with recent readings of different scalp and hair conditions that are commonly encountered in clinical settings.</p>

scholarly journals Deep Learning Analysis of In Vivo Hyperspectral Images for Automated Intraoperative Nerve Detection

Diagnostics ◽  
2021 ◽  
Vol 11 (8) ◽  
pp. 1508
Author(s):  
Manuel Barberio ◽  
Toby Collins ◽  
Valentin Bencteux ◽  
Richard Nkusi ◽  
Eric Felli ◽  
...  
Keyword(s):  
Deep Learning ◽  
State Of The Art ◽  
Hyperspectral Images ◽  
Invasive Technique ◽  
Average Sensitivity ◽  
Non Invasive ◽  
First Time ◽  
Learning Analysis

Nerves are critical structures that may be difficult to recognize during surgery. Inadvertent nerve injuries can have catastrophic consequences for the patient and lead to life-long pain and a reduced quality of life. Hyperspectral imaging (HSI) is a non-invasive technique combining photography with spectroscopy, allowing non-invasive intraoperative biological tissue property quantification. We show, for the first time, that HSI combined with deep learning allows nerves and other tissue types to be automatically recognized in in vivo hyperspectral images. An animal model was used, and eight anesthetized pigs underwent neck midline incisions, exposing several structures (nerve, artery, vein, muscle, fat, skin). State-of-the-art machine learning models were trained to recognize these tissue types in HSI data. The best model was a convolutional neural network (CNN), achieving an overall average sensitivity of 0.91 and a specificity of 1.0, validated with leave-one-patient-out cross-validation. For the nerve, the CNN achieved an average sensitivity of 0.76 and a specificity of 0.99. In conclusion, HSI combined with a CNN model is suitable for in vivo nerve recognition.

scholarly journals Autofluorescence Imaging of the Skin Is an Objective Non-Invasive Technique for Diagnosing Pseudoxanthoma Elasticum

Diagnostics ◽  
2021 ◽  
Vol 11 (2) ◽  
pp. 260
Author(s):  
Klára Farkas ◽  
Szabolcs Bozsányi ◽  
Dóra Plázár ◽  
András Bánvölgyi ◽  
Luca Fésűs ◽  
...  
Keyword(s):  
Cost Effective ◽  
Pseudoxanthoma Elasticum ◽  
Skin Lesions ◽  
Cardiovascular Complications ◽  
Invasive Technique ◽  
The Novel ◽  
Autofluorescence Imaging ◽  
Non Invasive ◽  
In Vivo Diagnostics

Pseudoxanthoma elasticum (PXE) is a rare multisystemic autosomal recessive connective tissue disease. In most cases, skin manifestations of PXE are the first to develop, followed later by severe ocular and cardiovascular complications. In our present study, in addition to dermoscopy, we introduced novel techniques, autofluorescence (AF) and diffuse reflectance (DR) imaging for the assessment of affected skin sites of five PXE patients. PXE-affected skin areas in most skin sites showed a previously observed pattern upon dermoscopic examination. With the novel imaging, PXE-affected skin lesions displayed high AF intensity. During our measurements, significantly higher mean, minimum and maximum AF intensity values were found in areas of PXE-affected skin when compared to uninvolved skin. Conversely, images acquired with the use of 660 and 940 nm illumination showed no mentionable difference. Our results demonstrate that AF imaging may be used in the in vivo diagnostics and quantification of the severity of the skin lesions of PXE patients. In addition, it is a safe, fast and cost-effective diagnostic method. AF imaging may be also used to objectively monitor the efficacy of the possible novel therapeutic approaches of PXE in the future.

Sign in / Sign up

Export Citation Format

Share Document