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
To evaluate whether the knee adduction moment (KAM) could be reduced by a short instruction in the Draw-in (DI) maneuver in healthy adults, and whether knee joint function would improve with a longer ...DI gait intervention in patients with knee osteoarthritis (OA).
In Study 1, healthy adults received 10 minutes supervised instruction in DI gait in and then practiced the gait independently for 10 minutes. Three-dimensional motion analysis measurement was performed in each phase. In Study 2, patients with OA performed a 20-minute DI gait intervention daily for 6 weeks. At baseline and after 6 weeks, knee pain, the Knee injury and Osteoarthritis Outcome Score, the MOS 8 item Short-Form Health Survey, thoracic kyphosis angle, knee joint range of motion, knee extension muscle strength, hip abduction muscle strength, and activity level were evaluated.
In Study 1, the DI gait to decrease KAM could be learning following only 10 minutes of instruction and 10 minutes of self-practice in healthy adults. In Study 2, knee pain was reduced by 19 % and the thoracic kyphosis angle was reduced by 2.6° after 6 weeks. No significant changes in other parameters were detected, and the implementation rate was 86 ± 14 %.
In healthy adults, DI gait instruction for 10 minutes of instruction and 10 minutes of self-practice reduced the KAM. In patients with knee OA, 20 minutes of DI gait per day for 6 weeks may reduce knee pain and thoracic kyphosis.
•Using the draw-in maneuver while walking is an effective intervention for knee OA.•Gait retraining significantly improved knee pain and thoracic kyphosis.•Gait retraining had a high implementation rate.•Gate retraining using the draw-in maneuver can be mastered even by older adults.
Kimura's disease (KD) is known to be dominant among young Asian men, but it can also occur in middle- and advanced-aged people. The clinical characteristics of KD, especially by age, are not well ...known.
This study was performed to investigate the effects of age on the clinical characteristics of KD.
We conducted a case series study.
All case studies of patients diagnosed with KD were collected via a PubMed search of studies published until August 2018. The data were analyzed by age group.
In total, 215 studies were reviewed (238 patients; mean age of 36 years). The male:female ratio was 4:1 overall, 17:1 in patients aged <20 years, 4:1 in patients aged 20-39 years and 2:1 in patients aged ≥40 years (P = 0.01). The percentage of patients with pruritus was 15.4% overall, 3.8% in patients aged <20 years, 15.5% in patients aged 20-39 years and 21.7% in patients aged ≥40 years (P = 0.02). The time to diagnosis was 5.3 years overall, 3.2 years in patients aged <20 years, 4.7 years in patients aged 20-39 years and 7.1 years in patients aged ≥40 years (P < 0.01).
The proportion of female patients affected the incidence of pruritus, and the time to diagnosis increased as the patients' age increased. There were no significant age-related differences in region/race, complications, multiplicity, laterality, anatomical distribution, maximum size, eosinophil count, immunoglobulin E level, initial treatment, recurrence or outcomes. This may be useful information for the diagnosis of KD.
A search for dark matter using an underground single-phase liquid xenon detector was conducted at the Kamioka Observatory in Japan, particularly for Weakly Interacting Massive Particles (WIMPs). We ...have used 705.9 live days of data in a fiducial volume containing 97kg of liquid xenon at the center of the detector. The event rate in the fiducial volume after the data reduction was (4.2±0.2)×10−3day−1kg−1keVee−1 at 5keVee, with a signal efficiency of 20%. All the remaining events are consistent with our background evaluation, mostly of the “mis-reconstructed events” originated from 210Pb in the copper plates lining the detector's inner surface. The obtained upper limit on a spin-independent WIMP-nucleon cross section was 2.2×10−44cm2 for a WIMP mass of 60GeV/c2 at the 90% confidence level, which was the most stringent limit among results from single-phase liquid xenon detectors.
Mannose-binding lectin (MBL) is a collagen-like serum protein that mediates activation of the complement system and is of importance for host defence. Common variant alleles situated both in the ...promoter and structural region of the human MBL gene (MBL2) influence the stability and the serum concentration of the protein. Epidemiological studies have suggested that genetically determined variation in MBL serum concentration influences the susceptibility to and the course of different types of infections, autoimmune, metabolic and cardiovascular diseases, but this is still a subject of debate. The fact that these genetic variations are very frequent indicates a dual role for MBL in host defence. In this survey, we summarize the current molecular understanding of human MBL genetics.
Abstract
Chromium telluride compounds are promising ferromagnets for proximity coupling to magnetic topological insulators (MTIs) of the Cr-doped (Bi,Sb)
2
(Se,Te)
3
class of materials as they share ...the same elements, thus simplifying thin film growth, as well as due to their compatible crystal structure. Recently, it has been demonstrated that high quality (001)-oriented Cr
2
Te
3
thin films with perpendicular magnetic anisotropy can be grown on
c
-plane sapphire substrate. Here, we present a magnetic and soft x-ray absorption spectroscopy study of the chemical and magnetic properties of Cr
2
Te
3
thin films. X-ray magnetic circular dichroism (XMCD) measured at the Cr
L
2,3
edges gives information about the local electronic and magnetic structure of the Cr ions. We further demonstrate the overgrowth of Cr
2
Te
3
(001) thin films by high-quality Cr-doped Sb
2
Te
3
films. The magnetic properties of the layers have been characterized and our results provide a starting point for refining the physical models of the complex magnetic ordering in Cr
2
Te
3
thin films, and their integration into advanced MTI heterostructures for quantum device applications.