Mapping variation of extracellular matrix in human keloid scar by label-free multiphoton imaging and machine learning

Significance

Rapid diagnosis and analysis of human keloid scar tissues in an automated manner are essential for understanding pathogenesis and formulating treatment solutions.

Aim

Our aim is to resolve the features of the extracellular matrix in human keloid scar tissues automatically for accurate diagnosis with the aid of machine learning.

Approach

Multiphoton microscopy was utilized to acquire images of collagen and elastin fibers. Morphological features, histogram, and gray-level co-occurrence matrix-based texture features were obtained to produce a total of 28 features. The minimum redundancy maximum relevancy feature selection approach was implemented to rank these features and establish feature subsets, each of which was employed to build a machine learning model through the tree-based pipeline optimization tool (TPOT).

Results

The feature importance ranking was obtained, and 28 feature subsets were acquired by incremental feature selection. The subset with the top 23 features was identified as the most accurate. Then stochastic gradient descent classifier optimized by the TPOT was generated with an accuracy of 96.15% in classifying normal, scar, and adjacent tissues. The area under curve of the classification results (scar versus normal and adjacent, normal versus scar and adjacent, and adjacent versus normal and scar) was 1.0, 1.0, and 0.99, respectively.

Conclusions

The proposed approach has great potential for future dermatological clinical diagnosis and analysis and holds promise for the development of computer-aided systems to assist dermatologists in diagnosis and treatment.

Deep learning-based morphological feature analysis and the prognostic association study in colon adenocarcinoma histopathological images

Colorectal cancer (CRC) is now the third most common malignancy to cause mortality worldwide, and its prognosis is of great importance. Recent CRC prognostic prediction studies mainly focused on biomarkers, radiometric images, and end-to-end deep learning methods, while only a few works paid attention to exploring the relationship between the quantitative morphological features of patients' tissue slides and their prognosis. However, existing few works in this area suffered from the drawback of choosing the cells randomly from the whole slides, which contain the non-tumor region that lakes information about prognosis. In addition, the existing works, which tried to demonstrate their biological interpretability using patients' transcriptome data, failed to show the biological meaning closely related to cancer. In this study, we proposed and evaluated a prognostic model using morphological features of cells in the tumor region. The features were first extracted by the software CellProfiler from the tumor region selected by Eff-Unet deep learning model. Features from different regions were then averaged for each patient as their representative, and the Lasso-Cox model was used to select the prognosis-related features. The prognostic prediction model was at last constructed using the selected prognosis-related features and was evaluated through KM estimate and cross-validation. In terms of biological meaning, Gene Ontology (GO) enrichment analysis of the expressed genes that correlated with the prognostically significant features was performed to show the biological interpretability of our model.With the help of tumor segmentation, our model achieved better statistical significance and better biological interpretability compared to the results without tumor segmentation. Statistically, the Kaplan Meier (KM) estimate of our model showed that the model using features in the tumor region has a higher C-index, a lower p-value, and a better performance on cross-validation than the model without tumor segmentation. In addition, revealing the pathway of the immune escape and the spread of the tumor, the model with tumor segmentation demonstrated a biological meaning much more related to cancer immunobiology than the model without tumor segmentation. Our prognostic prediction model using quantitive morphological features from tumor regions was almost as good as the TNM tumor staging system as they had a close C-index, and our model can be combined with the TNM tumor stage system to make a better prognostic prediction. And to the best of our knowledge, the biological mechanisms in our study were the most relevant to the immune mechanism of cancer compared to the previous studies.

Placental morphological features of small for gestational age preterm neonates born to mothers with pregnancy-induced hypertension

Introduction

Small for gestational age (SGA) neonates are often born to mothers with pregnancy-induced hypertension (PIH). Here, we aimed to explore the morphometric characteristics of the placenta during the perinatal period associated with SGA risk in mothers with PIH and identify the risk factors related to SGA.

Methods

The medical records of 134 neonates born between 28- and 32-weeks' gestation to PIH mothers were retrospectively analyzed. Placental morphology and umbilical cord (UC) length were compared between the SGA and appropriate for gestational age (AGA) groups.

Results

The placenta of the SGA group had a shorter major (15.00 vs. 18.00 cm; z = -6.04, p < 0.01) and minor placenta axes (13.00 vs. 15.00 cm; z = -4.59, p < 0.01), lower weight (300.00 vs. 420.00 g; z = -7.21, p < 0.01), smaller volume (282.00 vs. 396.00 cm³; z = -5.00, p < 0.01), and smaller area (141.00 vs. 212.00 cm²; z = -5.96, p < 0.01) than the AGA group. The UC was significantly shorter (39.00 vs. 44.00 cm; z = -3.68, p < 0.01). Short placental major axis [p = 0.03; odds ratio (OR): 2.16; 95% confidence interval (CI): 1.84 - 2.63] and low placental weight (p < 0.01; OR: 2.68; 95% CI: 2.66 - 2.70) were independent risk factors for SGA in premature newborns of PIH mothers.

Discussion

A major axis shorter than 15.5 cm or placental weight lower than 347.50 g at birth was related to a greater risk of SGA infants born to PIH mothers. As a predictor in prenatal ultrasound, the major axis is more helpful for precise prenatal pre-evaluation of vulnerable SGA preterm neonates with PIH mothers.

Vitreomacular Interface Abnormalities in Myopic Foveoschisis: Correlation With Morphological Features and Outcome of Vitrectomy

Purpose: To compare the morphologic characteristics and response to surgery of myopic foveoschisis (MF) with different patterns of vitreomacular interface abnormalities (VMIAs).

Methods: In this observational case series, 158 eyes of 121 MF patients with epiretinal membrane (ERM) or vitreomacular traction (VMT) based on optical coherence tomography (OCT) were enrolled. All the eyes were divided into two groups by the pattern of VMIAs: ERM and VMT group. Sixty-one eyes underwent pars plana vitrectomy (PPV) and were followed up for at least 6 months. The morphologic characteristics based on OCT and the surgical outcome were evaluated.

Outcome: ERM and VMT were found in 47.47 and 52.53% of the cases, respectively. A higher rate of foveal detachment (61.4 vs. 26.7%; p < 0.001) and a higher rate of outer lamellar macular hole (45.8 vs. 21.3%; p = 0.001) were detected in the eyes with VMT compared with those with ERM. In contrast, a lower rate of inner lamellar macular hole (28.9 vs. 60.0%; p = 0.001) was detected in the eyes with VMT compared with those with ERM. The disruption of the external limiting membrane (ELM) was more common in the eyes with VMT than in those with ERM (45.8 vs. 21.3%; p = 0.001). PPV was performed in 61 eyes with a mean follow-up time of 23.55 ± 19.92 months. After surgery, anatomical resolution was achieved in 51 eyes (83.6%). At the final visit, the mean central foveal thickness (CFT) decreased significantly from 547.83 to 118.74 μm, and the mean LogMAR BCVA improved significantly from 0.92 to 0.57. The VMT group was associated with a higher proportion of eyes with visual acuity improvement postoperatively (p = 0.02) and had more a decrease of CFT (P = 0.007) compared with the ERM group.

Conclusion: In the eyes with MF, outer retinal lesions occurred more frequently in the eyes with VMT, whereas inner retinal lesions occurred more frequently in the eyes with ERM. Tangential force generated by ERM may act as a causative factor for the inner retinal lesions in MF, and inward-directed force resulting from VMT may act as a causative factor for outer retinal lesions in MF.

Rapid identification of chrysanthemum teas by computer vision and deep learning

Seven commercial Chinese chrysanthemum tea products were classified by computer vision combined with machine learning algorithms. Without the need of building any specific hardware, the image acquisition was achieved in two computer vision approaches. In the first approach, a series of multivariate classification models were built after morphological feature extraction of the image. The best prediction accuracies when classifying flowering stages and tea types were respectively 90% and 63%. In comparison, the deep neural network was applied directly on the raw image, yielded 96% and 89% correct identifications when classifying flowering stage and tea type, respectively. The model can be applied for rapid and automatic quality determination of teas and other related foods. The result indicated that computer vision, especially when combined with deep learning or other machine learning techniques can be a convenient and versatile method in the evaluation of food quality.

Small-Scale Morphological Features on a Solid Surface Processed by High-Pressure Abrasive Water Jet

Being subjected to a high-pressure abrasive water jet, solid samples will experience an essential variation of both internal stress and physical characteristics, which is closely associated with the kinetic energy attached to the abrasive particles involved in the jet stream. Here, experiments were performed, with particular emphasis being placed on the kinetic energy attenuation and turbulent features in the jet stream. At jet pressure of 260 MPa, mean velocity and root-mean-square (RMS) velocity on two jet-stream sections were acquired by utilizing the phase Doppler anemometry (PDA) technique. A jet-cutting experiment was then carried out with Al-Mg alloy samples being cut by an abrasive water jet. Morphological features and roughness on the cut surface were quantitatively examined through scanning electron microscopy (SEM) and optical profiling techniques. The results indicate that the high-pressure water jet is characterized by remarkably high mean flow velocities and distinct velocity fluctuations. Those irregular pits and grooves on the cut surfaces indicate both the energy attenuation and the development of radial velocity components in the jet stream. When the sample is positioned with different distances from the nozzle outlet, the obtained quantitative surface roughness varies accordingly. A descriptive model highlighting the behaviors of abrasive particles in jet-cutting process is established in light of the experimental results and correlation analysis.