Summary
Background
Application of deep‐learning technology to skin cancer classification can potentially improve the sensitivity and specificity of skin cancer screening, but the number of training ...images required for such a system is thought to be extremely large.
Objectives
To determine whether deep‐learning technology could be used to develop an efficient skin cancer classification system with a relatively small dataset of clinical images.
Methods
A deep convolutional neural network (DCNN) was trained using a dataset of 4867 clinical images obtained from 1842 patients diagnosed with skin tumours at the University of Tsukuba Hospital from 2003 to 2016. The images consisted of 14 diagnoses, including both malignant and benign conditions. Its performance was tested against 13 board‐certified dermatologists and nine dermatology trainees.
Results
The overall classification accuracy of the trained DCNN was 76·5%. The DCNN achieved 96·3% sensitivity (correctly classified malignant as malignant) and 89·5% specificity (correctly classified benign as benign). Although the accuracy of malignant or benign classification by the board‐certified dermatologists was statistically higher than that of the dermatology trainees (85·3% ± 3·7% and 74·4% ± 6·8%, P < 0·01), the DCNN achieved even greater accuracy, as high as 92·4% ± 2·1% (P < 0·001).
Conclusions
We have developed an efficient skin tumour classifier using a DCNN trained on a relatively small dataset. The DCNN classified images of skin tumours more accurately than board‐certified dermatologists. Collectively, the current system may have capabilities for screening purposes in general medical practice, particularly because it requires only a single clinical image for classification.
What's already known about this topic?
Several computer‐aided classification systems have been introduced that achieve high sensitivity for melanoma detection; however, low specificity was a trade‐off for high sensitivity.
The application of deep‐learning technology to skin cancer classification could potentially improve the sensitivity and specificity of skin cancer screening.
The number of training images required for such a system is thought to be extremely large, and compiling large datasets for rare skin conditions is difficult.
What does this study add?
Our deep convolutional neural network (DCNN), trained on only 4867 images from 1842 patients, classified images of skin tumours into 14 different diagnoses more accurately than board‐certified dermatologists.
The fluctuation range was only ± 3·2% by fivefold cross‐validation, showing the robustness of the system.
Our DCNN system requires only a single image and provides 96·3% sensitivity and 89·5% specificity in the detection of skin cancer; however, it should be validated in a prospective clinical study before use for screening purposes in general medical practice.
Linked Editorial: Janda and Soyer. Br J Dermatol 2019; 180:247–248.
Plain language summary available online
Respond to this article
As a generic property, all substances transfer heat through microscopic collisions of constituent particles
. A solid conducts heat through both transverse and longitudinal acoustic phonons, but a ...liquid employs only longitudinal vibrations
. As a result, a solid is usually thermally more conductive than a liquid. In canonical viewpoints, such a difference also serves as the dynamic signature distinguishing a solid from a liquid. Here, we report liquid-like thermal conduction observed in the crystalline AgCrSe
. The transverse acoustic phonons are completely suppressed by the ultrafast dynamic disorder while the longitudinal acoustic phonons are strongly scattered but survive, and are thus responsible for the intrinsically ultralow thermal conductivity. This scenario is applicable to a wide variety of layered compounds with heavy intercalants in the van der Waals gaps, manifesting a broad implication on suppressing thermal conduction. These microscopic insights might reshape the fundamental understanding on thermal transport properties of matter and open up a general opportunity to optimize performances of thermoelectrics.
The prognostic significance of various systemic inflammation-based markers has been explored in different cancers. These markers can be used to assist with decision-making in oncology clinics.
The ...aim of this study was to investigate the prognostic significance of three systemic inflammation-based factors: neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR) and modified Glasgow Prognostic Score (mGPS) in patients with advanced pancreatic cancer.
Data were collected retrospectively for advanced pancreatic cancer patients treated between 1 January 2008 and 31 December 2012 at the Royal Perth Hospital. The ratios were dichotomised as <5 versus ≥5 for NLR and <200 versus ≥200 for PLR. Modified Glasgow Prognostic Scores were scored as: mGPS '0' = both C-reactive protein (CRP) and albumin normal, mGPS '1' = elevated CRP < 10 mg/L and mGPS '2' = both elevated CRP > 10 mg/L and albumin < 35 g/L. Univariate and multivariate analyses were carried out.
Data were evaluable for 124 patients. Median survivals based on the three inflammation-based prognostic markers evaluated were: NLR <5 versus ≥5 = 8.5 months versus 2.6 months respectively (P = 0.0007; hazard ratio (HR) 1.81), PLR <200 versus ≥200 = 9.1 months versus 4 months respectively (P = 0.007; HR 1.64) and mGPS score 1, 2, 3 = 8.3 months, 9.6 months and 1.8 months respectively (P = 0.0004). Besides Eastern Cooperative Oncology Group performance status, NLR, PLR and mGPS were significant independent prognostic markers both on univariate as well as multivariate analysis.
Our findings suggest that the NLR, PLR and mGPS derived from routine blood tests can be used as clinically meaningful biomarkers to stratify advanced pancreatic cancer patients into different prognostic groups.
The IEEE 802.15.4-2015 standard defines a number of Medium Access Control (MAC) layer protocols for low power wireless communications, which are desirable for energy-constrained Internet of Things ...(IoT) devices. Originally defined in the IEEE 802.15.4e amendment, the Time Slotted Channel Hopping (TSCH) has recently been attracting attention from the research community due to its reduced contention (time scheduling) and robustness against fading (channel hopping). However, it requires a certain level of synchronization between the nodes, which can increase the energy consumption. In this work, we implement the Guard Beacon (GB) strategy, aiming at reducing the guard time usually implemented to compensate for imperfect synchronization. Moreover, besides presenting a realistic energy consumption model for a Contiki Operating System-based TSCH network, we show through analytical and practical results that, without the proposed scheme, the power consumption can be more than 13% higher.
Although many reports have demonstrated that ectopic pain develops in the orofacial region following tooth pulp inflammation, which often causes misdiagnosis and inappropriate treatment for patients ...with pulpitis, the precise mechanism remains unknown. In the present study, we hypothesized that the functional interaction between satellite glial cells and neurons mediated by interleukin 1β (IL-1β) in the trigeminal ganglion (TG) is involved in ectopic orofacial pain associated with tooth pulp inflammation. The digastric muscle electromyogram (D-EMG) activity elicited by capsaicin administration into the maxillary second molar tooth pulp was analyzed to evaluate the noxious reflex and was significantly increased in rats with inflammation of the maxillary first molar (M1) versus rats injected with saline. A significant increase in the expression of connexin43 (Cx43), a gap junction containing protein, was observed in activated satellite glial cells surrounding second molar–innervating neurons in the TG after M1 pulpitis. Daily administration of Gap26, a Cx43 mimetic peptide and inhibitor, in the TG significantly suppressed the enhancement of capsaicin-induced D-EMG activity and the percentage of Fluoro-Gold (FG)–labeled cells encircled by glial fibrillary acid protein–immunoreactive (IR) + Cx43-IR cells after M1 pulp inflammation (P < 0.01). The percentage of FG-labeled cells encircled by glial fibrillary acid protein–IR + IL-1β-IR cells, IL-1 type I receptor–IR cells labeled with FG, and TRPV1-IR cells labeled with FG significantly increased after M1 pulp inflammation (P < 0.01). Daily administration of IL-1ra, an IL-1 receptor antagonist, into the TG significantly reduced the enhancement of capsaicin-induced D-EMG activity and the percentage of TRPV1-IR neurons labeled with FG after M1 pulp inflammation (P < 0.01). The present findings suggest that satellite glial cell is activated in the TG via activated gap junctions composed of Cx43 following tooth pulp inflammation, which leads to the hyperactivation of remote neurons via IL-1β mechanisms and results in ectopic tooth pulp pain in the adjacent tooth.
In this study, NiFe2O4 nanoparticles were prepared as a thermophilic agent for magnetic thermotherapy. To improve biocompatibility, the particles were coated with polyethylene glycol, which is ...hydrophilic and nontoxic; X-ray diffraction measurements confirmed that the particle size could be controlled to 3, 6, 9, and 17 nm. Fourier-transform infrared spectroscopy confirmed PEG modification to the particles. Biocompatibility with cells were improved by coating with PEG of high molecular weight. The heat-generating effect of the particles was investigated by measuring the magnetic susceptibility and temperature rise in an AC magnetic field. Furthermore, a device using a zero-voltage switching circuit was developed to generate heat. Finally, in vitro experiments using human breast cancer cells were successfully performed to confirm the onset of the hyperthermia effect.
Leptin acts on its receptor (ObR) in the hypothalamus to inhibit food intake and energy expenditure. Leptin and ObR are also expressed in the gastrointestinal tract; however, the physiological ...significance of leptin signaling in the gut remains uncertain. Suppressor of cytokine signaling 3 (SOCS3) is a key negative feedback regulator of ObR-mediated signaling in the hypothalamus. We now show that gastrointestinal epithelial cell-specific SOCS3 conditional knockout (T3b-SOCS3 cKO) mice developed gastric tumors by enhancing leptin production and the ObRb/signal transducer and activator of transcription 3 (STAT3) signaling pathway. All T3b-SOCS3 cKO mice developed tumors in the stomach but not in the bowels by 2 months of age, even though the SOCS3 deletion occurred in both the epithelium of stomach and bowels. The tumors developed in the absence of the inflammatory response and all cKO mice died within 6 months. These tumors displayed pathology and molecular alterations, such as an increase in MUC2 (Mucin 2, oligomeric mucus/gel-forming) and TFF3 (trefoil factor 3), resembling human intestinal-type gastric tumors. Administration of antileptin antibody to T3b-SOCS3 cKO mice reduced hyperplasia of gastric mucosa, which is the step of the initiation of gastric tumor. These data suggest that SOCS3 is an antigastric tumor gene that suppresses leptin overexpression and ObRb/STAT3 hyperactivation, supporting the hypothesis that the leptin/ObRb/STAT3 axis accelerates tumorigenesis and that it may represent a new therapeutic target for the treatment of gastric cancer.