Deep learning-based, computer-aided classifier developed with dermoscopic images shows comparable performance to 164 dermatologists in cutaneous disease diagnosis in the Chinese population / 中华医学杂志(英文版)

BACKGROUND@#Diagnoses of Skin diseases are frequently delayed in China due to lack of dermatologists. A deep learning-based diagnosis supporting system can facilitate pre-screening patients to prioritize dermatologists' efforts. We aimed to evaluate the classification sensitivity and specificity of deep learning models to classify skin tumors and psoriasis for Chinese population with a modest number of dermoscopic images.@*METHODS@#We developed a convolutional neural network (CNN) based on two datasets from a consecutive series of patients who underwent the dermoscopy in the clinic of the Department of Dermatology, Peking Union Medical College Hospital, between 2016 and 2018, prospectively. In order to evaluate the feasibility of the algorithm, we used two datasets. Dataset I consisted of 7192 dermoscopic images for a multi-class model to differentiate three most common skin tumors and other diseases. Dataset II consisted of 3115 dermoscopic images for a two-class model to classify psoriasis from other inflammatory diseases. We compared the performance of CNN with 164 dermatologists in a reader study with 130 dermoscopic images. The experts' consensus was used as the reference standard except for the cases of basal cell carcinoma (BCC), which were all confirmed by histopathology.@*RESULTS@#The accuracies of multi-class and two-class models were 81.49% ± 0.88% and 77.02% ± 1.81%, respectively. In the reader study, for the multi-class tasks, the diagnosis sensitivity and specificity of 164 dermatologists were 0.770 and 0.962 for BCC, 0.807 and 0.897 for melanocytic nevus, 0.624 and 0.976 for seborrheic keratosis, 0.939 and 0.875 for the "others" group, respectively; the diagnosis sensitivity and specificity of multi-class CNN were 0.800 and 1.000 for BCC, 0.800 and 0.840 for melanocytic nevus, 0.850 and 0.940 for seborrheic keratosis, 0.750 and 0.940 for the "others" group, respectively. For the two-class tasks, the sensitivity and specificity of dermatologists and CNN for classifying psoriasis were 0.872 and 0.838, 1.000 and 0.605, respectively. Both the dermatologists and CNN achieved at least moderate consistency with the reference standard, and there was no significant difference in Kappa coefficients between them (P > 0.05).@*CONCLUSIONS@#The performance of CNN developed with relatively modest number of dermoscopic images of skin tumors and psoriasis for Chinese population is comparable with 164 dermatologists. These two models could be used for screening in patients suspected with skin tumors and psoriasis respectively in primary care hospital.

Cone-beam computed tomography study of condyle position in high-angle adult female patients with different sagittal skeletal malocclusion / 口腔疾病防治

Objective @#To measure and compare condyle-related indexes(volume, surface area, anterior space, upper space and posterior space) in adult female patients with different sagittal malocclusions with the aim of providing a reference for clinical treatment. @*Methods @#The CBCT database of Qingdao Stomatology Hospital was searched to identify patients with high-angle adult skeletal malformations with the following classifications: classⅠ(n=30), class Ⅱ(n=30), and class Ⅲ(n=30). The CBCT images associated with the sample population were stored according to DICOM3.0 medical digital image communication standards. Mimics10.01 software was used to read and prepare three-dimensional reconstructions of the condylar, and the above-listed indicators were measured and statistically analyzed. @*Results @# There were no significant differences between the right and left sides among the three groups (P > 0.05). For parameters including condylar anterior joint space and post-joint space, the differences among the three groups were significant (P < 0.05). There was no significant difference in parameters related to condylar articular space between classⅠand class Ⅱ(P > 0.05). The differences between skeletal classesⅠand Ⅲ and between skeletal ClassesⅡ, and Ⅲ were statistically significant (P < 0.05). @*Conclusion@#Condylar position varies among high-angle adult woman with skeletal malocclusions with different sagittal skeletal patterns.

Ultrasonic microbubbles for glioma-targeted drug delivery / 药学学报

Ultrasonic microbubbles were used to open blood-brain barriers (BBB) with a reversed and limited behavior feature in the study, which could improve the brain-targeted delivery of anti-tumor drugs. The glioma rat model was prepared. Low-frequency ultrasound was combined with microbubbles to affect the permeability of BBB compared with the permeability of independently administered Evans blue (EB) crossing BBB. Time point and length of ultrasound were investigated whether they affect the permeability of BBB and the damage of brain tissue. The effect of the growth time of glioma on BBB permeability was explored. Only glioma had a very little impact on BBB permeability. However, ultrasonic microbubbles opened the BBB with the features of temporary, limited and reversed behavior and improved EB and magnetic resonance imaging contrast agent penetrating BBB. A length of 30 s ultrasound is appropriate for opening BBB and no damage of brain tissue. Drugs should be injected before ultrasound so that they enter into brain as BBB opening. Ultrasonic microbubbles can open BBB effectively and safely, which improve drugs penetrating BBB under proper time point and length.

Optimization of the preparation of sonogenic phospholipids-based microbubbles by using central composite experimental design and response surface methodology / 药学学报

Sonogenic microbubble agent is a newly developed drug targeting delivery system, which uses ultrasonic beam to enhance the delivery of drug and gene to targeted cells and tissues. In this paper, the preparation of sonogenic phospholipids-based microbubbles was optimized by using central composite experimental design (CCD) and response surface methodology (RSM). Hydrogenated egg phosphatidylcholine (EPC), Tween 80 and polyethylene glycol 1500 (PEG 1500) were important components affecting the concentration of 2 - 8 microm microbubbles in the preparation. The combined effects of these three factors were analyzed by CCD and optimized by RSM. Evaluation variable was the concentration of 2 - 8 microm microbubbles. Overall desirability was fitted to a second-order polynomial equation, through which three dimensional response surface graphs were produced. Optimal experimental conditions were selected from the stationary point of the response surfaces. The stability of the sonogenic phospholipids-based microbubbles by the optimal formulation was investigated by accelerated experiment. The contrast effect in vivo of the optimal formulation was investigated. Foreign market product SonoVue was used as the control. From the results, all the three factors had positive effects on the concentration of 2 - 8 microm microbubbles. The optimal condition in the preparation of phospholipids-based microbubbles was obtained as following: EPC 8.35 mg, Tween 80 21.68 mg and PEG 1500 201 mg. The mean value of the concentration of 2 - 8 microm microbubbles in rechecking experiment reached 8.60 x 10(9) x mL(-1). From the accelerated experiment, phospholipids-based microbubbles showed good physical stability. The intensity (relative unit) and duration of the contrast effect by the optimal formulation were 4.47 +/- 0.15 and (302 +/- 7) s respectively, which showed little difference with foreign market product SonoVue [4.28 +/- 0.13, (309 +/- 8) s]. The optimal formulation selected by CCD and RSM showed high microbubble concentration, good physical stability and effective sonogenic contrast effect.

Optimizing the operating variables that affect the transfection experiment of antisense oligodeoxyribonucleotide by gas-filled microbubbles / 药学学报

To optimize the operating variables that affect the transfection of antisense oligodeoxyribonucleotide (AS-ODNs) by insonated gas-filled lipid microbubbles, SF6-filled microbubbles were prepared by sonication-lyophilization method. An AS-ODNs sequence and a breast cancer cell line SK-BR-3 were used to define the various operating variables determining the transfection efficiency of SF6-filled microbubbles. Three levels of mixing speed, different durations of mixing and various delay time before ultrasound were examined, separately. Transfection efficiency was detected by fluorescence microscopy. Transfection results with and without incubation of AS-ODNs and microbubbles before mixing cells were compared. From the results, there is no significant difference between the transinfection efficiency with or without incubation of AS-ODNs and microbubbles before mixing cells. AS-ODNs transfection efficiency showed an increasing trend with mixing speed and mixing duration, but there is a negative relationship with delay time before ultrasound. The optimum parameters for AS-ODNs transfection by SF6-filled microbubbles were found at a mixing speed of 40-50 r x min(-1) for 30-60 s with less than 60 s delay before ultrasound. For a successful transfection, long time of incubation with gene is essential for normal nonviral vectors such as liposomes or cationic lipid-polymer hybrids, because these vectors depend on endocytosis and membrane fusion to realize transfection. Unlike liposomes and cationic lipid-polymer hybrids, gas-filled lipid microbubbles depend on sonorporation effect to realize transfection. Therefore, the incubation of gene and microbubbles before mixing cells may not be necessary. Ultrasound-mediated AS-ODNs transfection enhanced by gas-filled lipid microbubbles represents an effective avenue for gene transfer.

Recent advances in the applications of ultrasonic microbubbles as gene delivery systems / 药学学报

Ultrasonic beam can be focused on a particular tissue. As ultrasound contrast agents for medical ultrasound imaging, gas-filled microbubbles can also be used as drugs or gene carriers. Therefore, ultrasonic microbubbles become a topic of intense interest in drug delivery because they can be used as gene targeting delivery systems. Based on the relevant materials, the applications of ultrasonic microbubbles as gene delivery systems in various tissues in vivo were reviewed, such as cardiac, vascular, skeletal muscle and tumor. Their potential problems in future use were also discussed.