Weight-adaptive joint mixed-precision quantization and pruning for neural network-based equalization in short-reach direct detection links.

Neural network (NN)-based equalizers have been widely applied for dealing with nonlinear impairments in intensity-modulated direct detection (IM/DD) systems due to their excellent performance. However, the computational complexity (CC) is a major concern that limits the real-time application of NN-based receivers. In this Letter, we propose, to our knowledge, a novel weight-adaptive joint mixed-precision quantization and pruning approach to reduce the CC of NN-based equalizers, where only integer arithmetic is taken into account instead of floating-point operations. The NN connections are either directly cutoff or represented by a proper number of quantization bits by weight partitioning, leading to a hybrid compressed sparse network that computes much faster and consumes less hardware resources. The proposed approach is verified in a 50-Gb/s 25-km pulse amplitude modulation (PAM)-4 IM/DD link using a directly modulated laser (DML) in the C-band. Compared with the traditional fully connected NN-based equalizer operated with standard floating-point arithmetic, about 80% memory can be saved at a minimum network size without degrading the system performance. Quantization is also shown to be more suitable to over-parameterized NN-based equalizers compared with NNs selected at a minimum size.

Pituitrin Injection before Hysteroscopic Curettage for Treating Type I Cesarean Scar Pregnancy in Comparison with Uterine Artery Embolization: A Retrospective Study.

Background: The effectiveness and safety of pituitrin injection coupled with hysteroscopy and suction curettage as treatment for type I cesarean scar pregnancy (CSP) have not been studied enough in the literature, by comparing it to uterine artery embolization (UAE) followed by suction curettage we aim to determine its efficacy. Materials and Methods: Data of 53 patients (the PIT group) with type I CSP treated with pituitrin injection combined with hysteroscopic suction curettage and 137 patients (the UAE group) with type I CSP treated with UAE followed by suction curettage were collected in retrospect. The clinical data were analyzed statistically to compare the efficacy and safety between the two groups. Results: The PIT group had a shorter duration of postoperative vaginal bleeding, postoperative hospitalization, and overall hospitalization length (P < 0.05). The PIT group had lower overall hospitalization costs and a lower rate of adverse events than the UAE group (P < 0.05). There was no significant difference between the two groups in terms of treatment success rate, the average length of operation, blood loss during the procedure, time when serum ß-hCG returned to normal range, and menstrual recovery time after hospital release (P > 0.05). Conclusion: UAE and pituitrin injection followed by hysteroscopic suction curettage are good choices for type I CSP treatment. However, pituitrin injection with hysteroscopic suction curettage outperforms UAE followed by suction curettage. Thus, pituitrin injection may be an option of high priority for type I CSP.

Transplantation of decellularized and lyophilized amniotic membrane inhibits endometrial fibrosis by regulating connective tissue growth factor and tissue inhibitor of matrix metalloproteinase-2.

Intrauterine adhesion (IUA) is a disease characterized by endometrial fibrosis caused by injury to the endometrium. In the present study, decellularized and lyophilized human amniotic membrane (DL-AM) material was transplanted in a rat model to explore the preventive effect against IUA. A total of 24 Sprague Dawley rats were randomly divided into an IUA (n=12) group and an IUA + DL-AM (n=12) group. To establish the model, the endometrium of the left uterus was scraped, while that of the right uterus was used as a control. In the IUA group, scraped uteri were sutured without any other treatment, whereas DL-AM was transplanted onto the scraped uteri in the IUA + DL-AM group. Uteri were resected for histological and immunohistochemical evaluation at 3, 7, 14 and 28 days after surgery. The results confirmed the development of IUA, which was accompanied by an increase in the rate of fibrotic area. Integral optical density (IOD) values of connective tissue growth factor (CTGF) were elevated in the IUA group, while matrix metalloproteinase-2 (MMP-2) decreased relative to the control group (P<0.05). After DL-AM transplantation, the IOD value of CTGF dropped, while MMP-2 increased compared with the IUA group (P<0.05). However, compared with that in the control group, the IOD value of CTGF was still higher, whereas MMP-2 was still lower in the IUA + DL-AM group (P<0.05). Furthermore, no evidence of endometrial regeneration was detected in both the IUA and IUA + DL-AM groups. Overall, these results indicated that in the rat model of IUA, transplantation of DL-AM had the potential to prevent the formation of fibrosis to a certain extent and may thus be an alternative strategy for managing the condition.

Cascade recurrent neural network-assisted nonlinear equalization for a 100 Gb/s PAM4 short-reach direct detection system.

We propose a novel, to the best of our knowledge, cascade recurrent neural network (RNN)-based nonlinear equalizer for a pulse amplitude modulation (PAM)4 short-reach direct detection system. A 100 Gb/s PAM4 link is experimentally demonstrated over 15 km standard single-mode fiber (SSMF), using a 16 GHz directly modulated laser (DML) in C-band. The link suffers from strong nonlinear impairments which is mainly induced by the mixture of linear channel effects with square-law detection, the DML frequency chirp, and the device nonlinearity. Experimental results show that the proposed cascade RNN-based equalizer outperforms other feedforward or non-cascade neural network (NN)-based equalizers owing to both its cascade and recurrent structure, showing the great potential to effectively tackle the nonlinear signal distortion. With the aid of a cascade RNN-based equalizer, a bit-error rate (BER) lower than the 7% hard-decision forward error correction (FEC) threshold can be achieved when the receiver power is larger than 5 dBm. Compared with traditional non-cascade NN-based equalizers, the training time could also be reduced by half with the help of the cascade structure.

Computational complexity comparison of feedforward/radial basis function/recurrent neural network-based equalizer for a 50-Gb/s PAM4 direct-detection optical link.

The computational complexity and system bit-error-rate (BER) performance of four types of neural-network-based nonlinear equalizers are analyzed for a 50-Gb/s pulse amplitude modulation (PAM)-4 direct-detection (DD) optical link. The four types are feedforward neural networks (F-NN), radial basis function neural networks (RBF-NN), auto-regressive recurrent neural networks (AR-RNN) and layer-recurrent neural networks (L-RNN). Numerical results show that, for a fixed BER threshold, the AR-RNN-based equalizers have the lowest computational complexity. Amongst all the nonlinear NN-based equalizers with the same number of inputs and hidden neurons, F-NN-based equalizers have the lowest computational complexity while the AR-RNN-based equalizers exhibit the best BER performance. Compared with F-NN or RNN, RBF-NN tends to require more hidden neurons with the increase of the number of inputs, making it not suitable for long fiber transmission distance. We also demonstrate that only a few tens of multiplications per symbol are needed for NN-based equalizers to guarantee a good BER performance. This relatively low computational complexity signifies that various NN-based equalizers can be potentially implemented in real time. More broadly, this paper provides guidelines for selecting a suitable NN-based equalizer based on BER and computational complexity requirements.

Ovarian steroid cell tumor, not otherwise specified, treated with surgery: a case report and review of literature.

Ovarian steroid cell tumors NOS are rare sex cord-stromal tumors. They account for less than 0.1% of ovarian tumors. We present a case of a 17-year-old girl with the complaint of amenorrhea. The serum testosterone level was 11.55 nmol/L (reference value, 0.35-2.6 nmol/L) and the serum value of dehydroepiandrosterone-sulfate (DHEA-S) was 5.9 µmol/L (reference value, 0.49-8.71 µmol/L). A computed tomography (CT) pelvic scan identified a solid, right ovarian tumor and detected no adrenal gland enlargement or additional tumors. We took a surgical excision and a wedge resection of the normal contralateral ovary. The histopathologic examination on microscopy confirmed it was a benign ovarian steroid cell tumor NOS. Four days postoperative, her sex hormones were declined to normal levels and her serum testosterone level was 2.37 nmol/L (reference value, 0.35-2.6 nmol/L) a month after surgery. Her serum testosterone level was in the normal range and there was no evidence of recurrence 6 months after surgery.