No abstract

No abstract

Visual object detection is a computer vision-based artificial intelligence (AI) technique which has many practical applications (e.g., fire hazard monitoring). However, due to privacy concerns and the high cost of transmitting video data, it is highly challenging to build object detection models on centrally stored large training datasets following the current approach. Federated learning (FL) is a promising approach to resolve this challenge. Nevertheless, there currently lacks an easy to use tool to enable computer vision application developers who are not experts in federated learning to conveniently leverage this technology and apply it in their systems. In this paper, we report FedVision - a machine learning engineering platform to support the development of federated learning powered computer vision applications. The platform has been deployed through a collaboration between WeBank and Extreme Vision to help customers develop computer vision-based safety monitoring solutions in smart city applications. Over four months of usage, it has achieved significant efficiency improvement and cost reduction while removing the need to transmit sensitive data for three major corporate customers. To the best of our knowledge, this is the first real application of FL in computer vision-based tasks.

Objective. The data scarcity problem in emotion recognition from electroencephalography (EEG) leads to difficulty in building an affective model with high accuracy using machine learning algorithms or deep neural networks. Inspired by emerging deep generative models, we propose three methods for augmenting EEG training data to enhance the performance of emotion recognition models. Approach. Our proposed methods are based on two deep generative models, variational autoencoder (VAE) and generative adversarial network (GAN), and two data augmentation ways, full and partial usage strategies. For the full usage strategy, all of the generated data are augmented to the training dataset without judging the quality of the generated data, while for the partial usage, only high-quality data are selected and appended to the training dataset. These three methods are called conditional Wasserstein GAN (cWGAN), selective VAE (sVAE), and selective WGAN (sWGAN). Main results. To evaluate the effectiveness of these proposed methods, we perform a systematic experimental study on two public EEG datasets for emotion recognition, namely, SEED and DEAP. We first generate realistic-like EEG training data in two forms: power spectral density and differential entropy. Then, we augment the original training datasets with a different number of generated realistic-like EEG data. Finally, we train support vector machines and deep neural networks with shortcut layers to build affective models using the original and augmented training datasets. The experimental results demonstrate that our proposed data augmentation methods based on generative models outperform the existing data augmentation approaches such as conditional VAE, Gaussian noise, and rotational data augmentation. We also observe that the number of generated data should be less than 10 times of the original training dataset to achieve the best performance. Significance. The augmented training datasets produced by our proposed sWGAN method significantly enhance the performance of EEG-based emotion recognition models.

No abstract

No abstract

Purpose To evaluate an integrin imaging approach based on single photon emission computed tomography (SPECT)/computed tomography (CT) by using technetium 99m (Tc)-dimeric cyclic arginine-glycine-aspartic acid (RGD) peptides with three polyethylene glycol spacers (3PRGD2) as the tracer to target the integrin αβ expression in lung cancer and lymph node metastasis. Materials and Methods With ethics committee approval and written informed consent, 65 patients (41 male, 24 female; mean age, 60 years ± 11 [standard deviation]) with suspicious lung lesions were recruited with informed consent. The patients underwent both Tc-3PRGD2 SPECT/CT and fluorine 18 (F) fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT within 1 week. Finally, 65 lung lesions in 53 patients were pathologically diagnosed as non-small cell lung cancer (NSCLC) and 14 lung lesions in 12 patients were benign. Per-region analysis of lymph nodes included 248 regions with metastasis and 56 negative regions. Twenty specimens from the removed lung lesions or lymph nodes were stained with integrin αβ, CD34, and Ki-67 to correlate with the image findings. Receiver operating characteristic curve, z statistics, McNemar test, and χ analysis were used to compare the diagnostic performance of the two imaging methods. Results Tc-3PRGD2 SPECT/CT was found to be more specific thanF-FDG PET/CT in the per-region diagnosis of lymph node metastasis (specificity, 94.6% vs 75.0%; P = .008) when the sensitivity of the two methods was comparable (88.3% vs 90.7%; P = .557). There was no significant difference between the two methods in the per-lesion diagnosis of lung tumor (z = 0.82, P = .410). The accumulation level of Tc-3PRGD2 was found in positive correlation with the integrin αβ expression (r = 0.84, P = .001) and microvessel density (r = 0.63, P = .011) in the tumors. Conclusion Tc-3PRGD2 SPECT/CT shows high specificity in the diagnosis of lymph node metastasis from NSCLC, which may benefit surgical decision making for the patients. RSNA, 2016.

Rationale:Langerhans cell histiocytosis (LCH) involves mainly the skin and bone and rarely the thyroid. Meanwhile, papillary thyroid carcinoma (PTC) is the most common subtype of thyroid cancer. Both LCH and PTC could make the thyroid enlarged and hypermetabolic. The coincidence of these 2 events in a patient is rare, and this paper aimed to report such case.Patient concerns:A 40-year-old man presented with polyuria and polydipsia for 5 years. The symptoms had been relieved well by drug therapy for >4 years, until the drugs could not control the symptoms anymore and an extensively enlarged thyroid gland was noticed.Diagnoses:Thyroid ultrasound showed a nodule with microcalcification in the upper right lobe, positron emission tomography/computer tomography scan demonstrated thyroid hypermetabolism, and fine needle aspiration (FNA) revealed PTC. Right lobectomy of the thyroid and cervical lymph node biopsy verified the diagnosis “LCH of the thyroid complicated by PTC.”Interventions:The ultrasound-guided FNA biopsy was performed prior to right lobectomy of the thyroid and cervical lymph node biopsy. Postoperative histopathological examination confirmed the diagnosis, after which the patient received adjuvant chemotherapy.Outcomes:After 5 cycles of adjuvant chemotherapy, the patient had been followed up for 2 years. LCH was controlled satisfactorily and there was no significant sign of recurrence or metastasis of PTC.Lessons:LCH of the thyroid complicated by PTC is rare. Thyroid involvement should always be considered in the differential diagnosis of LCH patients. Surgery for PTC followed by chemotherapy for LCH may be the suitable treatment.