MIXTURE of human expertise and deep learning-developing an explainable model for predicting pathological diagnosis and survival in patients with interstitial lung disease.

Interstitial pneumonia is a heterogeneous disease with a progressive course and poor prognosis, at times even worse than those in the main cancer types. Histopathological examination is crucial for its diagnosis and estimation of prognosis. However, the evaluation strongly depends on the experience of pathologists, and the reproducibility of diagnosis is low. Herein, we propose MIXTURE (huMan-In-the-loop eXplainable artificial intelligence Through the Use of REcurrent training), an original method to develop deep learning models for extracting pathologically significant findings based on an expert pathologist's perspective with a small annotation effort. The procedure of MIXTURE consists of three steps as follows. First, we created feature extractors for tiles from whole slide images using self-supervised learning. The similar looking tiles were clustered based on the output features and then pathologists integrated the pathologically synonymous clusters. Using the integrated clusters as labeled data, deep learning models to classify the tiles into pathological findings were created by transfer-learning the feature extractors. We developed three models for different magnifications. Using these extracted findings, our model was able to predict the diagnosis of usual interstitial pneumonia, a finding suggestive of progressive disease, with high accuracy (AUC 0.90 in validation set and AUC 0.86 in test set). This high accuracy could not be achieved without the integration of findings by pathologists. The patients predicted as UIP had poorer prognosis (5-year overall survival [OS]: 55.4%) than those predicted as non-UIP (OS: 95.2%). The Cox proportional hazards model for each microscopic finding and prognosis pointed out dense fibrosis, fibroblastic foci, elastosis, and lymphocyte aggregation as independent risk factors. We suggest that MIXTURE may serve as a model approach to different diseases evaluated by medical imaging, including pathology and radiology, and be the prototype for explainable artificial intelligence that can collaborate with humans.

Hyaluronic acid secretion by synoviocytes alters under cyclic compressive load in contracted collagen gels.

Knee osteoarthritis is a degenerative disease of diarthrodial joints. Biomechanical factors are considered as risk factors for the disease, the knee joint being normally subject to pressure. Some studies have examined the biomechanical environment of the knee joint in vitro. The aim of this study was to establish a culture model to mimic the knee joint environment. As a first step, synoviocytes induced contraction of three-dimensional collagen gels. Next, contracted collagen gels containing synoviocytes underwent cyclical compression ranging from 0 to 40 kPa at a frequency of 1.0 Hz for 1.5, 3, 6 and 12 h using the FX-4000C™ Flexercell(®) Compression Plus™ System. RNA in collagen gels was extracted immediately after compression and mRNA expression levels of HAS genes were analyzed by quantitative RT-PCR. Culture medium was collected 48 h after compression and analyzed by agarose gel electrophoresis and cellulose acetate electrophoresis. Synoviocytes in contracted collagen gels were stimulated by cyclic compressive load. Long-term compressive stimulation led to the production of higher molecular weight hyaluronic acid, whereas, short-term, compressive stimulation increased the total amount of hyaluronic acid. Furthermore, mRNA expression levels of both HAS-1 and HAS-2 were significantly higher than without compression. Taken together, using this gel culture system, synoviocytes synthesized higher molecular weight hyaluronic acid and produced large quantities of hyaluronic acid through up-regulation of HAS gene expression. Therefore, the contracted collagen gel model will be a useful in vitro three-dimensional model of the knee joint.

Construction of collagen gel scaffolds for mechanical stress analysis.

We designed a cyclic compression system using readily available six-well culture plates to investigate the influence of mechanical stress on skin-like structures. The effects of cyclic mechanical stress on protein expression by cells were easily examined, and hence, this system should be useful for further analysis of skin responses to mechanical stress.

Shark protein improves bone mineral density in ovariectomized rats and inhibits osteoclast differentiation.

OBJECTIVES: Fish proteins are potential sources of natural medicines and food additives. There are many studies being performed to develop underutilized fish proteins. Therefore, the aim of this study was to determine how shark protein functions as a dietary supplement for bone health. METHODS: Three groups of ovariectomized (OVX) rats were fed different diets containing 20% casein protein, 20% shark protein, or 20% cod protein for 4 wk. Bone mineral density of the right femur was measured by dual-energy x-ray absorptiometry and quantitative computed tomography. Furthermore, we prepared low-molecular-weight peptides from shark protein using protease for in vitro studies. Calcitriol was added to bone marrow cells and the receptor activator of the nuclear factor-κB ligand was added to RAW264 cells. After 7 d, the number of tartrate-resistant acid phosphatase-positive cells was counted. RESULTS: In the shark protein-fed group, bone mineral density of the femur epiphysis was higher than that of the casein protein-fed group. In particular, the shark protein-fed group showed an increase in bone mineral density, represented mainly by trabecular bone. Shark protein hydrolysates inhibited osteoclast formation in bone marrow cells and RAW264 cells. CONCLUSIONS: These results suggest that shark protein might suppress the bone loss caused by estrogen deficiency through the suppression of osteoclast formation.