Training set data expansion. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Structural plane recognition effect. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Structural plane classification. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Mixup data expansion. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Marking example. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Image_5_Monoallelic Heb/Tcf12 Deletion Reduces the Requirement for NOTCH1 Hyperactivation in T-Cell Acute Lymphoblastic Leukemia.jpeg by Diogo F. T. Veiga (12282320)

Neuroreceptor kinetics in rats repeatedly exposed to quinpirole as a model for OCD by Stijn Servaes (6433580)

“…Subsequently, sagittal slides were made of the CP in the right hemisphere and a staining was done with the D2R, mGluR5 and GLT-1 antibody to visualize the corresponding receptor.…”

Data_Sheet_1_Determining 5HT7R’s Involvement in Modifying the Antihyperalgesic Effects of Electroacupuncture on Rats With Recurrent Migraine.pdf by Lu Liu (171341)

“…<p>Electroacupuncture (EA) is widely used in clinical practice to relieve migraine pain. 5-HT₇ receptor (5-HT₇R) has been reported to play an excitatory role in neuronal systems and regulate hyperalgesic pain and neurogenic inflammation. 5-HT₇R could influence phosphorylation of protein kinase A (PKA)- or extracellular signal-regulated kinase₁_/₂ (ERK₁_/₂)-mediated signaling pathways, which mediate sensitization of nociceptive neurons via interacting with cyclic adenosine monophosphate (cAMP). …”

Net difference in premature mortality for PM_2.5 and NO₂ between models with varying resolution vs distance from primary road. by Alejandro Valencia (16007329)

The N-terminal domain is sufficient for CMPK2 antiviral function. by Joanna B. Pawlak (5210639)

Amygdala voxel correlations with ACC whose values <i>decrease</i> as symptoms of post-combat emotional distress increase. by Tom Brashers-Krug (762715)

“…Threshold for significance was a <i>z</i>-score < -5.39. Coronal slices through the brain at y = 4, 2, 0, -2, -4, -6, and -8 mm showing the amygdala bilaterally. …”

ACC voxel correlations with amygdala whose values <i>decrease</i> as symptoms of post-combat emotional distress increase. by Tom Brashers-Krug (762715)

“…Threshold for significance was a <i>z</i>-score < -5.39. Only values within the ACC ROI are shown. …”

Drill image dataset for training part II. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

U²-Net network structure diagram [8]. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

RSU-7 structure diagram [8]. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

Drill image dataset for training part I. by Qingjun Yu (1649473)

“…Based on the PyTorch deep learning framework, the initial U²-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. …”

<i>Nanos1</i>⁺ cells in the bulb decrease after X-ray irradiation or drug treatments. by Sosuke Fujita (17691623)

Decreased IL1B-induced NF-κB activity in chronic iAs-exposed HUC-1 cells by Pei-Fen Su (15123)

Effects of PANDER on the mRNA expression of <i>Dgat1</i>, <i>Dgat2</i>, <i>Cpt1</i>, and <i>Acox1</i> in the liver and adipose tissue of mice. by Xiaoqing Mo (2217691)

Impact of SARS‐CoV‐2 infection on T, B and NK cell numbers in patients compared with the control group (**** <i>P< 0</i>.<i>0001</i>). by Sam Alahyari (12645313)